Web Search

Reflections on Being a Journal Editor-in-Chief

2011-12-30T16:22:00.001-05:00

Internet Research

I've been editor-in-chief of the journal, Internet Research for about a year now. So, I wanted to share some reflections about the job with you in case you are interested in being an editor, want to know about the process, or have to evaluate others who are editors.

My top three take-aways after one year are:

1. It's a lot of work (seriously, a lot of work!): I was unprepared for the amount of time that it takes! Although folks told me that being a journal editor is time consuming, part of me always thought "you just aren't as efficient with your time as I am!". .... No, they were right. It takes a lot of time. Given the number of new submissions, the reviewer reports, and the manuscript revisions that come in per day (plus locating reviewers and issue management), I easily spent a work day a week on just the editorship.

2. Your name gets in front of a lot of people: Given communication with the editorial board researchers, the manuscript authors, and manuscript reviewers, I'm virtually meeting hundreds of researchers from around the world. It really is an amazing experience and very enriching! Now, with the authors, it is not always pleasant, as the rejection rate is about 80+%, but it's still a a great opportunity to meet other researchers.

3. It's good to be the king (sometimes!): As editor, I make the final decisions on the accepts and the rejects. Although it seems obvious, once I started making these decisions day in and day out, it really struck me - "There is no appeal!" I don't 'like' your manuscript, I can reject it. ... On one hand, it's nice to have a hand in shaping a body of research. However, this authority has also put a lot of responsibility on me to be fair, polite, understanding ... and thick skinned!

Hope you found these thoughts useful!

Business engagement on Twitter: a path analysis

2011-09-30T23:31:00.000-04:00

Interested in using Twitter for online branding and customer engagement? ... The life cycle of tweet is 1.5 to 4 hours and only reaches consumers with a second-degree relationship to the brand.

In the paper, Business engagement on Twitter: a path analysis, we examined the potential influences of business engagement in online viral marketing at the consumer engagement level and investigated the trajectories of a business’ online viral message diffusion in the Twitter community.

Social media services like Twitter enable commercial businesses to participate actively in online viral advertising.

We used used path analysis to examine 164,478 tweets from 96,725 individual Twitter users with regards to nine brands during a 5-week study period. We operationalized business engagement as the amount of online word-of-mouth messages from brand and the number of consumers the brand follows.

We operationalized consumers’ engagement as the number of online branded messages from consumers both connecting to the brand and having no connection with the brand as well as the number of consumers following the brand.

We concluded that the business engagement on Twitter relates directly to consumers’ engagement with online viral communication.

In addition, retweeting, as an explicit way to show consumers’ response to business engagement, indicates that the influence only reaches consumers
with a second-degree relationship to the brand and that the life cycle of a tweet is generally 1.5 to 4 hours at most.

The lead author on the research paper was Mimi Zhang and our co-author was Abdur Chowdhury

For an analysis of tweets as viral marketing, see Twitter Power: Tweets as Electronic Word of Mouth

John Wiley Best JASIST Paper Award

2011-09-02T22:59:00.000-04:00

I’m quite honored that this year’s winner of the John Wiley Best JASIST Paper Award is my co-authored article, The Seventeen Theoretical Constructs of Information Searching and Information Retrieval, which I co-wrote with Soo Young Rieh of the University of Michigan.

This paper took me 5 years from idea concept, through research, through 4 rounds of review, to eventual publication. However, it was one of those ideas that I just knew was needed.

The information science field (along the field of computer science) is essentially empirical in nature and was lacking a clear statement of the field's core theoretical constructs. I believed these constructs needed to be explicitly stated.

Events seem to be proving that the need was there.

Within its first year of publication, the article was already on the syllabi of graduate courses in several of the information science and i-schools.

The American Society for Information Science and Technology awards the John Wiley Best JASIST Paper Award yearly. One paper is selected from the articles that are published in the Journal of the American Society for Information Science and Technology, the flagship journal for the society and for the field of information science.

Basic Web Analytics Process

2011-06-21T01:12:00.002-04:00

Although I need one several times, I couldn't find a good, straightforward, description of a web analytics process.

Most of the ones that I found are loaded with additional 'stuff' related more to a consulting view point than the basic web analytics process. Nothing wrong with the consulting 'value added', but from a teaching point of view, one wants the core elements.

So, I boiled the web analytics process down to 4 essential stages, shown below.


Web Analytics Process - 4 Steps

Collection of data: Exactly what it says, collection of the basic, elementary data. Typically, this data is counts of things.

Processing of data into information: At this stage, you typically take counts and make them ratios, although there are still some counts. You are taking your data and making it into information, specifically metrics.

Developing KPI: At this stage, you focus on those ratios (and counts) which are infused with business strategies. These are referred as Key Performance Indicators (KPI). Many times, KPIs deal with conversion aspects, but not always.

Formulating online strategy: These are the online goals, objectives, and standards for your organization or business. Usually, these strategies are related to making money, saving money, or increasing marketshare.

Each stage impacts (or can impact) the stage preceding or following it. So, sometimes the data that you *can* collect impacts your online strategy. Other times, your online strategy affects the data you collect.

This process is part of a presentation on web analytics. Part of the presentation is based on the book, Understanding User - Web Interaction via Web Analytics.

Web Analytics Webinar 17 June 1130-1230 EDT

2011-06-08T11:20:00.000-04:00

If you are interested in web analytics, tune in to this webinar on Friday 17 June 2011 at 1130 EDT.

Here is where you can register for the web analytics webinar.
(http://goo.gl/pHcLB)

Here's what we will be covering and some details.

Abstract: Web analytics can be a valuable tool in achieving overall organizational goals, if you understand the key elements and processes. Web analytics is measuring, collecting, analyzing, and reporting of internet data for understanding and optimizing online usage. From commercial businesses to academic libraries, many organizations are realizing the importance of using web analytics to better understand their customers (such as their goals, preferences, and online behaviors) and from this understanding, making better consumer relationship management decisions. In this webinar, we address the essential elements of web analytics, including key performance indicators, standard metrics, and the building of a linkage between the two. We will introduce tools, such as Google Analytics, to provide inexpensive web analytics reporting. We also review a process to link Web analytics as an informational foundation to strategic organizational decisions. Real life examples are presented from the commercial business and academic library domains.

Title: Web Analytics

Date: Friday, June 17, 2011

Time: 11:30 AM - 12:30 PM EDT

After registering you will receive a confirmation email containing information about joining the Webinar.

System Requirements
PC-based attendees
Required: Windows® 7, Vista, XP or 2003 Server

Macintosh®-based attendees
Required: Mac OS® X 10.4.11 (Tiger®) or newer

A thanks to ASIST for sponsoring.

Some of the webinar will be based on my book, Understanding user web interactions via web analytics. You might be able to get it free here, Understanding user web interactions via web analytics.

Howie Jacobson, Brett Tabke, and Jamie Murphy comment on the book, Understanding Sponsored Search

2011-05-25T16:02:00.003-04:00

I've written this book, Understanding Sponsored Search: Core Elements of Keyword Advertising, hitting the streets July 2011.

When you write a book like this, you have to send the manuscript to folks, hoping some of them will write a book endorsement for you.

Thought I would share the three book endorsements that I'm going to use.

The first, from Google AdWords guru, Howie Jacobson ...

"Jim Jansen proves that theory can make practice a whole lot more profitable. In this book, he explains how and why most search marketers miss their targets, and shows the reader exactly how to construct effective search marketing campaigns. While search marketing rules and features come and go, Jim's insights stay evergreen. If you've ever wondered what Human Information Processing Theory, Credence Goods, and George Carlin's Seven Words You Can't Say on Television can teach you about profiting from tiny online ads, Jim's your man. I'm going to quote from this book a lot, and people will think I'm a genius."

Howie Jacobson
Author, Google AdWords For Dummies
askHowie.com

The second from, Brett Tabke, who founded and runs the largest community of practice for online advertisers in the world ...

“With Understanding Sponsored Search, Jansen has delivered an insightful and well-researched exploration of fundamental search advertising principles that bridges the gap between academic study and everyday keyword buying, in an enjoyable book that pushes the boundaries of sponsored search. A definite must-read that ought to be required reading in the sponsored search industry.”

Brett Tabke
Founder and CEO of WebmasterWorld.com
Founder and CEO of PubCon

The third from Jamie Murphy, who is the academic lead of the Google Online Marketing Challenge, one of the largest in class competitions in the world and totally focused on keyword advertising ...

"Wow, finally Jim Jansen shares his insights and helps level the playing field! Professor Jansen and his students have consistently scored in the top 1% of the world's largest search marketing contest, the Google Online Marketing Challenge. If you are serious about keyword advertising, this book merits the number one position on your bookshelf."

Professor Jamie Murphy, Murdoch University
Lead Academic, The Google Online Marketing Challenge

A big thanks to Howie, Brett, and Jamie! Quite honored and comments much appreciated!

End-to-End Analysis of the Spam Value Chain

2011-05-21T08:47:00.007-04:00

Wanted to pass this on, as I believe it is a wonderful example of multi-disciplinary research. And, the best piece of anti-email spam work that I've seen.

Rather than leverage algorithmic attributes to stop spam email, the researchers actually track the entire value chain of 100 pieces of spam -- clicks, goods, money, etc. Good blend of technology, economics, business, finance, and legal.

Their goal was to locate 'bottlenecks' in the spam value chain. Super interesting.

Major finding: 95% of all spam transactions flow through *3* banks

Full paper: Click Trajectories: End-to-End Analysis of the Spam Value Chain
http://goo.gl/ce0Ea

Cliff Note version: Anatomy of a Spam Viagra Purchase
http://goo.gl/6bXRX

Great piece of research!

Some folks who are mentioned in Understanding Sponsored Search

2011-05-14T23:03:00.009-04:00

In the book, Understanding Sponsored Search, there are many folks who are mentioned by name. So, I want to give some of them a shout out!

Scott Banister, Jeffrey Brewer, and Bill Gross – credited with conceiving the idea of keyword advertising
John Battelle – the Database of Intentions
Sabeer Bhatia and Jack Smith – founded Hotmail, the first free internet email service, and conceived an amazing way to market it
Rick Boyce – credited with pioneering the idea of banner advertising
Sergey Brin – technology to enhances people’s lives and founded Google
Andrei Broder – three types of web queries (informational, navigational, transactional)
Nico Brooks – conceived of the 'conversion potential' in sponsored search
Clay Davis – provided me the idea for using ‘the 7 words you can’t say on television’
Rufus Evison - the hotel problem
Gord Hotchkiss – conceived the idea of the Google Triangle based on eyetracking studies
Brad Geddes – explains the three entities in web searching
Salar Kamangar and Eric Veach – developed Google AdWords, which provides ~90+% of Google's revenue
Marissa Mayer – the Google business model
Jamie Murphy, Daehee Park, and Lee Hunter – the Google Online Marketing Challenge
Larry Page – the concept of ads being relevant and non-annoying for web searching
Jan Pedersen – web searching behaviors
David Penniman – developed standard methods of log analysis
Pete Pirolli – information foraging theory
Marc Resnick – user reaction to sponsored links
Dan Rose – user intent in web searching
Tefko Saracevic – the concept of relevance
Eric Schmidt – Google being in the auction business
Amanda Spink – web searching behavior
Danny Sullivan – commentary on web searching
Mimi Zhang – search engine branding and viral marketing

There are MANY, MANY more who’s work is referenced in the book! (the book contains more than 350 references).

BTW, the next time you are at Amazon, go to Understanding Sponsored Search, and click the “I’d like to read this book on Kindle.” Thanks!

Teaching and Learning with Technology: Using Apps to Enhance Learning for University Students

2011-05-10T09:24:00.006-04:00

Penn State has been pretty ambitious in exploring ways to leverage technology to enhance learning and a large part of this endeavor is due to the Teaching and Learning with Technology (TLT) division at the university. They are at the forefront of enriching the educational experience for students via information technology.

I’ve been fortunate enough to have been selected as a TLT Fellow for 2011. Working with the TLT folks, we will be exploring ways to leverage the inherent aspects of smart phones (i.e., location aware, always on, socially networked) to improve learning in the classrooms. So, it ties in well with my prior research on information searching, but it also introduces a whole new avenue of research.

Honestly, there is already a lot on my plate with teaching and research; however, I just couldn’t pass up the opportunity to work with Cole Camplese, Allan Gyorke, Brad Kozlek, and the rest of the TLT folks. (Note: Cole is a definite ‘early adopter’ and teaches a very popular course at Penn State on disruptive technologies.)

Our end state is (hopefully!) a course specific app that we have developed and evaluated in a classroom setting. It’s going to be a challenging research project. But, there are several efforts ongoing in this area, including these apps from the search engine company, Wolfram Alpha, which they market as the Wolfram Course Assistant Apps.

Sabbatical Reflection #8: The Six Business Rules of Tenure

2011-05-05T11:10:00.007-04:00

Tenure is a big deal in academia.

Old Main at Penn State

(Note: Personally, I believe fixed term contracts of increasing length is a much better system for the professor, university, and students. For example, a 3 year contract for a professor, followed by a 6 year contract, followed by a 10 contract, etc.)

Although the tenure concept originated to provide academic freedom for professors, tenure today is more of a promotion. In most institutions, if you don’t make tenure, you are typically fired (a.k.a., asked to leave, contract not renewed, etc. -- basically fired).

Having went through the tenure process (with both successes and failures), and having some time to reflect upon the process, tenure is very analogous to running a start-up business.

As with running a winning business, there are some guiding principles to successfully getting tenure. I've identified six.

The Six Business Rules for getting tenure:

1. Product: Produce a Product! What would you think of a business that didn’t offer any product or service? … Crazy right? Tenure expectations are the same – you have to produce something!, be it research that leads to articles, grants that lead to articles/patents, teaching courses that leads to students who learned something, etc. The products vary from university to university but the deliverables are what counts for tenure. Too many academics get caught up in the process and get stuck there, never actually producing the product. And, when tenure decision time comes around, they have no product to show.

2. Marketing: Market your Product! What would you think of a business that set-up a shop but never stocked the shelves, never secured distribution channels, or never advertised? A set-up for failure! ... There are a lot of critical supporting activities that make a business successful. Same with tenure, the most critical supporting activity being marketing. That great research you did means nothing if no one knows about it – write a research paper and get it published. That great class you taught is wonderful for the small number of students in the course – but for tenure, share your experiences with other teachers by writing up a case study. Find ways to market your business’s products!

3. Advice: Seek Advice (and ignore most of it)! Building a business is *not* a process to learn OJT. Likewise with tenure. Seek out advice on the entire gamut of the tenure landscape, from publication outlets, to the tenure expectations at all levels (e.g., department, college, and university), to outside letter writers, etc. In getting information, be a sponge … and, then ignore most it! … At the end of the day, it is your business and other people’s advice and experiences many times just don’t apply. Plus, I’ve noticed that many times people didn’t follow the advice they dole out. However, collecting the information is extremely valuable in planning your tenure business.

4. Time: It Takes Time! It is the rare business that is successful overnight. Same for tenure. Can’t tell you the number of people who have asked for advice the year before the critical tenure review. It is just too late to do anything at this point, sadly. Example of what I mean by time: the typical article takes 2+ years from research planning to publication. For a 4 year tenure review, you need to start planning on how your packet will look at year 2! So, backward plan from your tenure date. You need to have most of the products done for tenure two years out! (i.e., tenure decision at year 6, have your packet mostly done at year 4).

5. Enthusiasm: Passion for Your Product! You ever talked to small business owners or folks at a start-up? Their enthusiasm for what they are doing is nearly overwhelming. Why? … ‘cause running a business is a whole hell of a lot of work, takes a lot of time, and requires a lot of energy. People give up a lot to run their own business. To survive, they have to love what they are doing. Ditto for tenure – academia is many times lonely, ego painful, time demanding, and energy taxing. To make it, without being miserable and to be successful, you have to love what you do. ... If you don’t love it … do yourself a favor. Go do something that you do love.

6. Relationships: Don’t be an Asshole! Most business owners that I know can be efficient to the point of being brisk, demanding, challenging, pushy, opinionated, etc. However, not too many that I know are just assholes. Why? … a lot of a successful business is about relationships. And, for a good relationship to exist, your heart has to be in the right place. A good portion of the tenure process is about other people, and those other people generally don’t like assholes. So, if you are an asshole …your tenure packet had better be rock solid 'cause people will be just looking for an excuse to send you packing. It’s much better to follow the golden rule in the tenure business.

Take away: Manage your tenure process as you would a start-up business! Produce a product, market it, seek advice, plan ahead, be passionate, and respect people.

Bonus Rule: Want to be successful in your field? ... Start early! I’ve always been curious about success in an academic field. So, I’ve observed my field (information science, information retrieval, information searching) to see what attributes are common among those who are *very successful* and, perhaps, identify attributes that are predictive of success.

What is the most reliable predictor that I see? … success came early!

Those who are most successful seem to have hit the market (i.e., immediate post PhD) at full speed – they had impactful publications of their own work in substantial numbers, and they produced these publications early in their careers.

This 'out of the gate' attribute seems more predictive of success than any other characteristic that I've noted (e.g., prestigious university, prestigious advisor, ‘hot’ research area, etc.)

Welcome your thoughts on the analogy between tenure and a business, as well as other 'rules' or 'principles' for the business of tenure.

Visit to SnapRetail

2011-05-03T16:43:00.010-04:00

SnapRetail logo banner

The folks at SnapRetail were kind enough to have me in for a visit. This Pittsburgh-based start-up has a really interesting business model, organization, and underlying technology.

Was hosted by Jill Berardi (VP of Strategy and Products). In addition to being with the company since the beginning, Jill Berardi has some fantastic experience in the marketing and advertising area.

Also had the opportunity to chat with Ted Teele (CEO) , Jeremy Hirsch (VP Marketing), Tyler Johnston (VP Business Development), Andy Fraley (CTO), and many other members of the SnapRetail team. Just some wonderful insights into the business aspects and culture of a tech start-up during our discussions!

The SnapRetail folks also made me feel really welcome with dozens of these personalized signs that were scattered like bread crumbs from the parking lot to their offices!

Welcome Jim Jansen to SnapRetail

SnapRetail is focused on the ecommerce space of linking retailers with vendors, offering SaaS for retailers in the gift, home and collegiate industries. The technology, all developed in-house, is really top-notch for creating and running marketing campaigns via email and on social media.

They appear to be onto something, as SnapRetail got $6 million in Series A funding, on top of an earlier $2.25 million funding round.

Really looking forward to see what SnapRetail does going forward!

As an additional bonus, I got to meet Ian Paterson, who was in the first graduating class from Penn State’s College of IST!

They’ve also got a good branding thing going themselves with things like Snappy Hour (a Q&A session on Facebook) and branded chairs like these.

SnapRetail branded chair

Understanding Sponsored Search (book cover)

2011-04-12T13:58:00.009-04:00

I'm liking the cover to my book, Understanding Sponsored Search: Core Elements of Keyword Advertising.

Doing everything I can to move the publication date up -- whenever the publisher sends me something to review or approve, I get it back to them the next day (if not sooner!).

The Table of Contents is:

1. A context for sponsored search
2. Modeling the process of sponsored search
3. Understanding the potential customers' intent for keyword selection
4. Sending signals to the potential customer with ads
5. Understanding consumer behavior for sponsored search
6. BAM: branding advertising and marketing for sponsored search
7. Sponsored search analytics
8. The serious game of bidding on keywords
9. Bringing it all together in a framework of sponsored search
10. The future of sponsored search.

BTW, can order Understanding Sponsored Search on Amazon.

Sabbatical Reflection #7: Comment on Confessions of an Advertising Man

2011-04-12T12:04:00.009-04:00

Confessions of an Advertising Man by David Ogilvy is a must read for any professional, especially those interested in marketing, branding, and advertising … which is just about everyone!

I read the book a couple of years ago, but there is one passage that has stuck with me, processing in the back of my mind and rising to the forefront on occasion.

The passage is this:

“If you prefer to spend all your time growing roses or playing with your children, I like you better, but do not complain that you are not being promoted fast enough. Managers promote the men who produce the most.” (page 172).
… Note: excuse the gender specific reference, the book was written in the 60s, when advertising was almost exclusively male.

The quote is in the chapter How to Rise to the Top of the Tree (Advice to the Young), pages 171 – 178.

This passage captured my attention the moment I read it. Although rather blunt, I find the straightforwardness refreshing. It is really about setting priorities, acting on them, and then being adult enough to accept the outcome of those decisions. Note, that Ogilvy doesn't make a judgment on which decision is the best. He implies that both have value (i.e, even if I don't promote you, I might like you better than someone I do promote based on the decisions you made).

The passage compares non-work decisions with work decisions, but it can just as easily be applied to decisions within a professional context, which is why I'm commenting on it.

In academia, it can pertain to “decisions on where to publish (or not to publish)”, “service roles that one takes on (or doesn’t)”, “the courses that one teaches (or doesn’t)”, “grants that ones applies for (or doesn’t)”, “writing blog postings (or not)”, etc. Each of these decisions has some fairly predictable outcomes. If you’ve thought them though, you should be adult enough to accept the consequences.

The passage really got me thinking about my work priorities and the consequences from them. It’s caused me to alter some of my professional decisions … and hopefully accept the possible consequences of them.

Anyway, Confessions of an Advertising Man is a great book, and I recommend it to anyone who does any writing, and certainly to anyone in marketing, advertising, counseling, branding, or sales. I drew on the concepts when writing my book, Understanding Sponsored Search.

The Convenience versus Privacy trade-off in web Search

2011-03-12T17:26:00.007-05:00

Web search engines offer a service (i.e., search) that has a inherent tension between convenience and privacy for the searcher.

Although you may think of Google as the exemplar example of a web search engines, social networking sites, such as Facebook and Twitter, will be increasing their searching services. Simultaneously, search engines such as Google are increasing their social networking attributes. So, the line between ‘search engine’ and ‘social networking site’ is blurring. With the combined social and searching service, the tension between convenience and privacy will be even more pronounced.

Additionally, the aspect of web searching is inherently linked with web browsing, which is traversing websites on the internet. Web browsing has a related and implicit tie to the convenience and privacy tension of web search.

By web search, I’m referring to this macro-searching view rather the micro-searching view of simply entering a query into a text box.

Convenience is anything that is intended to save resources (e.g., time, energy) or frustration. In the online context, convenience refers to the services websites offered to users or to the businesses that operate these sites.

Privacy is the condition of being left alone, out of public view, or in control of information that is known about you . In the online context, privacy can be a nuanced concept, incorporating aspects of control, anonymity, and obscurity.

Given the way that web search is architectured, as convenience is increased, privacy is decreased.
The situation is complicated for the search engines (again, mainly Google, Facebook, and Twitter but also Yahoo! and Bing to some extend) in that they compete internationally. Depending on the culture and society, the acceptable expectations for convenience and privacy differ. Additionally, non-US based search engines face similar (and in some cases additional) pressures in resolving the convenience – privacy tension. Examples of these are clear with Baidu in China, Yandex in Russia, and Exalead in France.

It seems that the convenience – privacy tension will continual for some time. However, there seems to be a ‘space’ that provides an acceptable level of privacy along with worthwhile convenience for most stakeholders.

Good convenience and reasonable privacy: An example within this acceptable range is the use of referral keywords. When you enter a query into a search engine and click on a result link, the search engine passes something called a referral URL to the website that you visit. This referral URL has the address of the search engine (e.g., www.google.com) and the query terms that you entered. The website owners use this information to optimize their websites to provide the visitor a better user experience. Other than the search terms, no other personal information is passed. Most folks seem okay with this practice (of course, there are always exceptions).

Convenience trumps privacy: There are times when the convenience issues outweigh any privacy concerns . For example, some web surfers have turned off cookies in their browsers out of privacy concerns. However, there are websites (i.e., online banks, mutual funds, insurance companies, etc.) that have considerable security responsibilities (i.e., do you want someone else accessing your bank account or retirement fund or heath insurance profile?). Since these sites can’t rely on standard cookies, they have adopted techniques such as flash cookies, which collect far more information about the user’s computer than standard cookies ever could..

Privacy trumps convenience: Conversely, there are times when supposedly convenience benefits have been implemented, causing an outrage among users over privacy concerns. An example of this is when Facebook automatically began broadcasting profile updates to friends in a user’s networks (i.e., Lisa is ‘now single’, Tom is ‘now unemployed’, etc.). Many users were outraged, as this feature was implemented without any prior notification or opt out feature.

The convenience – privacy tension may be a good thing, as it causes a continual evaluation of our online context. I was quite surprised at how often I had written about the convenience – privacy tension:

The Brand Effect in Keyword Advertising

2011-03-04T21:39:00.008-05:00

Nearly every PPC practitioner worth her salt knows to separate branded terms from non-brand terms in any keyword advertising campaign.

Among the many reasons is the expectation that branded terms have better overall performance than non-branded terms.

However, this assertion has never been analyzed in an academic study using a large data set from an actual keyword advertising campaign, until now.

In the soon to be published research article, The Brand Effect of Queries and Advertisements in Keyword Advertising, my co-authors (Kate Sobel and Mimi Zhang ) and I analyze the effect of branded terms in queries, branded terms in the ad copy, and the interaction between the two.

We use a sponsored search campaign that ran on all three major search engines comprised of about 2.5 million daily records, including the keyphrase and advertisement copy.

We analyzed the brand effect on number of clicks, cost per click, sales revenue generated, number of orders, number of items ordered, and return on advertising cost, as well as impressions triggered by these keyphrases

Findings?

A combination of a branded phrase and a branded ad generated 15 times higher sales revenue than any other category (i.e., combination of phrase or ad, either branded or not). The presence or absence of a branded term in either the query or ad was significant in all metrics.

We also tested for effect size (i.e., the data set was so large that even a small difference would be significant, so this tests for practical effects). Accounting for the size of the data, the only large effect was in those records with both branded queries and ads versus non-branded queries and ads, although they were across board small effects for most categories.

Interestingly, the second highest performing categories (accounting for the difference in cost) in most cases was the nonbranded query and branded ad.

Percentage of Overall Cost and Revenue by Brand Focus Categories

Based on these finding:
- if there is a branded term in the query, it pays to have a branded term in the ad
- if there is no branded term in the query, it still pays to have a branded term in the ad
- it probably doesn't pay, on average, to bid on competitors branded phrase (Note: in a separate research study, Mark Rosso and I examined this bidding strategy; it doesn't happen much. Now I know why. This strategy doesn't generate revenue, See: Brand Names as Keywords in Sponsored Search Advertising)

The full results are shown in The Brand Effect of Queries and Advertisements in Keyword Advertising

Warning: a lot of tables and numbers.

Sabbatical Reflection #6: Assuming editor-in-chief duties for an academic journal

2011-03-03T06:49:00.005-05:00

My sabbatical started in mid-May 2010, and I’m using part of the time to take stock of and reflect on where my professional life had been, where it’s at, and where to go from here.

With that reflection, I’ve decided to take on the job as editor of an academic journal, Internet Research.

The publication of scientific literature is at the core of the academic discipline. As such, I believe there is a great responsibility on journal editors for the maintenance of standards while also being open to novel ideas, approaches, and methods.

The publication of research articles also have implications for individual researchers in a variety of fields for tenure, promotion, and reputation. As such, I approach the editor’s role with a great deal of respect and seriousness. It’s an area where I believe I can make a positive impact.

What about the journal?

I see Internet Research publishing research centered in an interdisciplinary area of the internet research. A journal can’t be all things to all people. There has to be a ‘core’ something to the journal that is meaningful to the community that it serves.

This interdisciplinary research area is a good operating space for Internet Research. Related themes of interest are:

Commerce – a revolution in how industry communicates and transacts business on the web, especially keyword advertising, content advertising, and mobile advertising
Organization – the seemingly impossible task of matching humans with the information they seek, including information searching and information seeking
Governance – dealing with questions of rights, obligations, jurisdiction, and freedom in the realm of cyberspace
Infrastructure – the systems and standards that made the explosive growth of the internet possible
Interfacing – examining the roles that presentation and interaction play in our acceptance of, adoption, and integration into the online world.
Linking – a core concept that has since morphed into the ubiquitous web, including the use of social media

The journal has a solid grounding now. The rejection rate is 71%, meaning 29 out every 100 submissions get published, but there is certainly a lot of room for improvement, including:

Impact factor: rising the impact factor (currently at 0.844) above 1
Review Process: tightening the review process to have a rapid turn around time with quality comments and editorial guidance
Publication time: shortening the time from submission to online publication making it competitive with conferences

So, I'm looking forward to this!

Can You Love! (or hate) a Search Engine just on Looks?

2011-02-22T08:18:00.018-05:00

Branding is a critical element of marketing and advertising.

However, the technology area has paid little attention to branding and the effect that branding has on the results obtained from user studies, from system evaluations, and on technology acceptance. This is true of many web technologies.

I had conducted a study investigating the effect of brand on the evaluation of search engine results, Search engine brand and the effect on user perception of searching performance. In the study, I took the results from one search engine and ‘placed’ them in the branding elements of another. Findings? ... A positive search engine brand was worth about 10%-15% for users when evaluating the usefulness of the results.

So, just the branding elements (i.e., name, logo, colors of results page, etc.) could effect searcher react to the information presented by the search engine.

This got me wondering if branding elements could affect the user’s perception of the search engine itself. In other words, regardless of the underlying technology or quality of service, does just the branding elements cause an affective (i.e., emotional) reaction in a potential searcher.

In an experiment to investigate this question, we had 200 participants record their reactions to 10 search engine logos, including a logo for a search engine that did not exist. This non-existent search engine provided a validity check and served as a baseline for comparison. Here is a sample question from the survey:

Sample from Survey Instrument

In the analysis, we controlled for whether or not a participant had seen or used the search engine before.

Results?

The reactions to the logos ranged from “love” to what can only be described as "disgust" or "contempt".

Google elicited a string of positive comments from the participants, to the point of several uses of the term ‘love.’ Most of the other search engines, including Microsoft, had primarily negative terms associated with them, although AOL, Ask, and Yahoo! had a mix of both positive and negative comments.

Bottom line:

branding elements can have an effect on system evaluations, separate from the technology. Therefore, system evaluations would control for this response.
branding aspects, such as name and logo, are critical attributes for taking a technology to market, perhaps as important as the technology itself initially.
it would seem reasonable that marketing efforts, experience with the technology use over time, etc. could change the perception of a brand, either positively or negatively

Here are the ten search engine logos and sample reactions that capture the essence and trends of participant comments.GoogleYahoo!

Ask

AOL

DogpileMahaloMSN

AllTheWebA9

AI2RS

You can read the entire research study, Investigating Brand Knowledge of Web Search Engines: User Reactions to Search Engine Logos. This research was conducted with my co-authors, Lu Zhang and Anna Mattila.

Is the Buying Funnel of Any Value for Keyword Advertising?

2011-02-21T08:01:00.010-05:00

The buying funnel is one of the most widely used ... and also one of the most widely criticized ... business models around.

This got me thinking "what value does it have for keyword advertisers?"

To investigate this question, my co-author and I used a seven million record dataset from a multi-million dollar keyword advertising campaign of a major US national online and brick & mortar retailer to investigate the value of the buying funnel.

Findings ...

As a model for consumer behavior, the buying funnel is not of much value.
As a paradigm for making sense of keyterms, the buying funnel is a nice organizing metaphor from broad to narrow.

The buying funnel has been used in advertising since at least 1898, when introduced by E. St. Elmo Lewis.

The buying funnel model is typically depicted with a four stages, with each stage relating to the cognitive phase of the consumer who is contemplating on whether or not to make a purchase. Four stages of the buying funnel are:

Awareness: when a customer realizes that there is a product that can solve his/her problem
Research: After a consumer realizes that a product can address a problem, the customer finds a specific product line and become more knowledgeable about this type of product or service.
Decision: is when a consumer is deciding between different brands of a specific product by forming choice set.
Purchase: The final stage of the buying funnel. This stage is when a consumer knows what specific product and brand that they intend to purchase, and they are typically doing a price, convenience to order, or similar comparison before executing the purchase.

We took each keyphrase in our dataset and classified it into one of the stages of the buying funnel, and then statistically analyzed each grouping of keyphrase using impressions, clicks, cost per click, sales revenue, orders, and items ordered for the entire campaign.

Here is some overview of the dataset to give you an idea of the scope of the keyword campaign.

As you can see, a $56M in sales campaign (with millions of impressions and clicks) and $8.5M in advertising spent. So, a robust campaign by any measure.

Major findings?

there were statistically significant differences among all critical keyword advertising attributes for all stages of the buying funnel
the buying funnel is a workable model for classifying the focus (i.e., broad to narrow) of online ecommerce queries
a ‘funnel’ may not be the best shape for describing the search process as it is employed by the aggregate set of online consumers
consumers appear to adopt a satisficing behavior when engaging in online searching, in accordance with the principle of least effort

Rather than a funnel, the online searching processing is more of a multi-path process.

You can read the entire study, Bidding on the Buying Funnel for Sponsored Search and Keyword Advertising, co-authored with Simone Schuster

Visit to Evri: Semantic news aggregation

2011-02-11T19:10:00.007-05:00

Had a great visit to Evri, which has taken an interesting spin on news aggregation.

Evri uses advance topic identification software, developed internally, to channel real-time streams of news on topics of interest, which one can then personalize.

In addition to web information sources, such as blogs and Twitter, Evri gets professional feeds such as those from the AP and Getty Images. So, some high quality information streams, along with their crawls of real time web sources.

Evri provides options to search for specific news topics, and it then shows the semantic graph of related items.

There are also widgets for blogs, iphone, and tablets.

All free, btw!

I like it so much, changing my homepage to Evri to give it a try.

Big thanks to Raman Chandrasekar for organizing the visit! And to Kris Koperski and Carsten Tusk for a fantastic lunch conversation!

During the visit, got the opportunity to presentation research on ecommerce information sharing on social media sites.

Visit to SEOmoz: a SaaS Model for SEO

2011-02-11T00:22:00.006-05:00

The folks at SEOmoz were kind enough to host me for a visit at their company headquarters in Seattle, Washington.

My primary research is on the search engine marketing (SEM) side, so a visit to a search engine optimization (SEO) firm was really interesting for me. It’s a side of the online marketing business that I don’t get to interact with much.

SEOmoz is unique in that they do not do SEO consulting. Their primarily business model is providing SEO applications using a software as a service (SaaS) model. There are free versions of the tool and a premium version for a subscription fee.

I was really interested in the subscription model, as I’d author a Pew report on purchasing online content, which looked at the purchase of software online and subscription models for digital content.

Although formerly in the consulting area, the SaaS model is now the core business for SEOmoz, with a wide customer base. Interestingly, this whole line grew out of a small side activity of the company, which was a blog where the company would share information, tips, and tools.

SEOmoz stills offers an array of SEO learning material, including The Beginner's Guide to Search Engine Optimization. This ebook is an in depth tutorial on both search engines and the fundamental strategies to make your websites search engine friendly.

Got a nice tour of the operation from Sarah Bird. Then, went to lunch with Adam Feldstein, Jamie Steven, and Gillian Muessig, where we had a wide ranging conversation from SEO practices, to social media, to shortcomings of current advertising-marketing degrees, to getting girls interested in technology.

Wonderful visit!

Plus, I got this cool t-shirt!

Final interesting note, SEOmoz was founded by the mother – son team, of Gillian and Rand Fishkin.

The book, Understanding Sponsored Search, …. Submitted!

2011-01-31T14:27:00.004-05:00

I submitted the manuscript of Understanding Sponsored Search: A Coverage of the Core Elements of Keyword Advertising to the kind folks at Cambridge University Press.

Happy to get it, generally, off of my lap onto to someone else’s.

Some thanks are in order.

Brooke Randel, independent copy writer, who performed an expert proofing and content review of the entire book. Brooke was a former student of mine in the GOMCHA course, who has now moved on to better things! We’ll continue to hear more good things about her.

Rianna Jansen, my daughter, who did the final copy editing prior to submission. She corrected the many mistakes of grammar that I interjected into the manuscript with my continual fiddling up to the last moment. Her efforts significantly improved the readability of the final content.

Thanks also to the subject matter experts who reviewed individual chapters. Nearly every chapter was reviewed by both an academic and a practitioner.

The chapter reviewers were:
• Brad Geddes, CertifiedKnowledge.org
• Daehee Park, Acxiom
• Dietmar Wolfram, University of Wisconsin Milwaukee
• Don Turnbull, Independent consultant and researcher
• G. Allen Westra, Upper Iowa University
• George Michie, Rimm-Kaufamn Group
• Lu Zhang, Penn State University
• Nico Brooks, Two Octobers
• Sebastien Lahaie, Yahoo! Research
• Shelby Thayer, Penn State University
• Theresa Clarke, James Madison University

Still some fiddling to do, I’m sure! But, glad to be basically done.

Sabbatical Reflection #5: Should I Write a Book?

2011-01-22T08:40:00.007-05:00

A project that many academics do on sabbatical is to write a book.

When I was planning my sabbatical, I intentionally decided NOT to write one. It just seemed liked such a typical thing to do, plus I wasn’t really excited about it.

I had already been involved with three books, having written (Understanding User – Web Interactions via Web Analytics), co-written (Web Search: Public Searching of the Web), and co-edited (Handbook of Research on Web log Analysis), so it just didn’t seem like the energizing project that I was looking for as part of my sabbatical.

However, sometimes one makes a plan just so fate … or whatever … can mess with it.

In mid-May 2010, a couple of opportunities came up to submit a book proposal, and I decided to go with Cambridge University Press. The theme for the book was a theoretical exploration of sponsored search (a.k.a., keyword advertising), a topic that I had been considering for some time.

One advantage of ‘thinking’ about a topic for awhile (about 2 years!) was that the words just came out. I wrote a 50 page proposal in one sitting. By July, the book proposal was submitted. The reviewers were on the ball, and they got comments back by the end of the month. The contract was signed by mid-August 2010, with a submission date of 31 January 2011. I devoted a set time each week for book writing. By December 2010, the majority of the content was written, with January being set aside for external content reviewing and copy editing.

A fairly rapid schedule for an academic book!

In hindsight, I’m super glad I did it. The theory of sponsored search is a topic that I am interested in, have done a range of research on, and have had time to reflect upon. My first research journal paper on sponsored search was in published in 2006, An examination of searcher's perceptions of non-sponsored and sponsored links during ecommerce Web searching. So, I started conducting research in the area in 2004, with a range of projects in the field since then.

More importantly, writing the book gave me the opportunity to craft a coherent view of sponsored search, synthesizing a variety of research projects and readings together into one package. So, although written for a target market segment, the book was very cathartic for me.

So, should I write a book on sabbatical? … Didn’t plan on it. But, did it.

How much are real time search queries worth?

2011-01-17T21:41:00.011-05:00

About $33 million dollars.

Real time search is the querying of real time content on the web. Examples of real time content are tweets, comments on blog posting, and newsfeeds from social media sites.

Real time content is entering the main stream. Although only 8 percent of the US internet population is on Twitter, people are becoming accustomed to receiving real time content directly from major news events, sometimes within minutes of these events occurring. Many more people are accustomed to receiving tweets and other real time content on sites like Facebook and LinkedIn.

Several companies are investigating whether or not there are commercial opportunities in providing real-time search services.

Both the major search engines and new search firms are entering the marketplace providing technologies for searching real time content. For example, Google and Bing provide access to Twitter tweets.

The question is, how much is this search traffic worth?

To date, there had been no published economic valuation of this search traffic.

Determining the economic value of this search traffic was one goal of the research project reported in the paper, Real Time Search on the Web: Queries, Topics, and Economic Value.

We estimated the economic value of the real time search traffic using the Google AdWords keyword advertising platform.

For this research, we took the unique queries submitted to Collecta, a real time search engine, and leveraged the Google AdWords platform to determine a rough valuation of real time search traffic if it had occurred on Google. To establish value, we used the cost it would have taken to advertise for these queries on Google for one day.

The estimated economic value of the real time traffic was $33,023,320. This $33 million is the average of the high and low estimates for the queries that returned a value in AdWords.

Search results retrieved in response to this real time search traffic would generate an estimated 6,492,552 clicks at an average CPC of $1.14, with specific figures for the traffic shown in following table.

The next table shows the top twenty-five keywords based on average cost per day.

$33 million is a very rough estimate of the economic value; however, it provides a starting point for estimating the value of real time searching content.

Real time content providers, such as Twitter and Facebook, are exploring ways to advertise within the content stream, which has its own unique challenges.

However, the topics of searching on real time search engines is inline with that of searching on traditional search engines, as the top queries in the table above indicate. Therefore, the keyword advertising model of sponsored search may port to these real time search platforms.

Co-authors on the paper were Zhe Liu, Courtney Weaver, Gerry Campbell, and Matt Gregg.

A draft version of the research paper, Real Time Search on the Web: Queries, Topics, and Economic Value, is posted.

Searching as Learning!

2011-01-13T13:40:00.009-05:00

One aim of web search analysis is to model the deeper underlying motivations of the searcher (i.e., ‘why is the searcher actually searching?’, ‘what does the person want to accomplish?’). These motivations are not explicitly apparent from the query terms themselves, but knowing the underlying motivations can greatly inform search systems.

One idea that I’ve had on the back burner for some time is the concept that much searching is a learning behavior.

This conceptualization makes sense for many types of search, most notably for those searches that attempt to address some information asymmetry (i.e., the person searching doesn’t have the information they need and someone else does).

The issue was how to research this linkage between searching and learning?

We used Anderson and Krathwohl’s taxonomy (i.e., think updated Bloom’s Taxonomy) and developed searching scenarios for a laboratory experiment.

We investigated queries per session, average query length, unique terms, number of topics, session duration, result pages viewed, and systems used.

What did we find?

For nearly all search characteristics, the data looked like the following.

The numbers for the learning episodes of Applying and Analyzing where higher than those of the other four categories of Remembering, Understanding, Evaluating, and Creating.

We proposed a searching effort – learning taxonomy framework.

I found the results really interesting, it that we could use searching behaviors to algorithmically determine complex underlying cognitive needs and motivations of searchers.

One can tie this searching-as-learning modeling directly to searching systems like Wolfram Alpha and Google Squared.

The research paper, Using the taxonomy of cognitive learning to model online searching, reports the findings of the complete research linking searching to learning.

My co-authors on the research paper were Danielle Booth and Brian Smith

What do we know from a single search query?

2011-01-07T12:54:00.008-05:00

You submit a single query to a search engine.

From that one query, what can we find out about you?

It turns out, quite a lot.

In the paper, To What Degree Can Log Data Profile a Web Searcher?, we illustrate how to derive a wide spectrum of personal information from a single query. We provide this framework of user profiling.

Framework for Profiling a Searcher Using Query Log Data

To illustrate our point, we propose a hypothetical query from an unknown user that is submitted to a web search engine. From this one query about an academic society’s conference, we derive:

the likelihood of the searcher’s occupation and field
the gender probability of the searcher
the searcher’s education level
the geographical location of the searcher
possibilities on where the person works
the searcher's motivation for conducting the search, including whether or not the searcher is planning to travel

Important privacy questions aside, this was a fun project to work on, and the hypothetical query was more of a demonstration of what is possible with search log data.

However, from this introductory example, the paper goes on to outline how to determine:

geographical location
locality intent of the query
topic of the query
query reformulation patterns
user intent
commercial intent
purchase intent
gender of the searcher
whether or not the user will click on a result
searcher identity

The take away is that online companies have gotten, and will continue to get, better at deriving knowledge from our online trace data.

Co-researchers on the project, all students or former students, were Mimi Zhang, Danielle Booth, Daehee Park, Ying Zhang, Ashish Kathuria, and Pat Bonner.

Is technology changing the way we think? … Seriously doubt it

2011-01-01T23:40:00.006-05:00

Occasionally, the claim is made that some aspect of technology is changing the way we think. It’s computers, the Internet, or Google or whatever.

Photo: Paul Sakuma/The Associated Press

Typically, the undercurrent is that the change is not good. Usually along the lines that technology is dumbing us down or making us lose the ability for serious thought or taking away our concentration or in some way making us stupid.

Now the view that technology is making us dumb is not universal, as this article from Pew Internet, Does Google Make Us Stupid? reports. Or this really nice article, Is the Web helping us evolve?, by David Brin.

I believe that technology is a wonderful tool and is certainly making us more knowledgeable.

But, I have serious doubts that it is substantially changing the way we think, either in a positive or negative way.

I just find it hard to believe that cognitive thought patterns, developed over hundreds of thousands of years through a natural selection process, are going to seriously change.

There are at least three broad arguments supporting the position that technology does not change the way we think:

Practical Argument: Tools don’t change the way we think, they just change the way we ‘think’ about things: In this sense, ‘think’ is not an ingrained cognitive process, but more of a habit or metaphor for viewing the world.

Classic examples are the clock and compass. While it is true that we form habits with the tools we use, these habits are quickly unlearned. Concerning the clock, ditch your watch while on vacation. Your view of the day changes nearly immediately. As for the compass, as it is not in common use anymore, our referring to the world around us via NSEW directions has also stopped. Most people can't even find north anymore.

We can list many other tools that enforced habits that disappeared once we stop using the technology (i.e., the shoelace, the levered ice cube tray, the rotary phone, etc.)
Philosophical Argument: If something is so human as to be an inherent trait, how would humans ever notice it? In other words, if something is really essential to being human, no human can identify it because that person is also human.

Related to this is the ‘explain your own actions’ test. When ever a researcher comes up with a theory about human behavior, try applying the theory to explain what that researcher is doing, to see if the theory still makes sense.

So, when I read about someone saying, for example, ‘the Internet is making us lose our ability for concentrated thought?’, my first question is along the lines of ‘then how did you write this 300 page book?’
Biological Argument: If the way we think is changing, then we should see some change in the structure of our brains. We aren’t seeing any.

The last major biological change to the human brain occurred about 20,000 years ago when humans started leaving hunter-gather communities and settling down into farming communities (i.e., we domesticated ourselves).

So, although it makes good drama to say so, no one has made a good case that technology is seriously changing the way we think.

I’ve spent about $10 this month on digital content. … About average

2011-01-01T19:54:00.012-05:00

65% of internet users have paid for online content from the Pew Internet & American Life Project hit the street, presenting a run down of:

the percentage of internet users who have paid for online content (a.k.a., digital media)
the type of online content purchased
how this content was paid for
how much was spent on this content

As the author of the report, I was surprised when the percentage of internet users who purchased online content (65%) was nearly the same as the percentage who had purchased tangible products online (67%).

In fact, I found it really surprising, and didn’t believe it.

If you had asked me “do you pay for online content?”, I would have responded with a big ‘almost never’.

In the survey, though, instead of asking this single question, we asked about purchasing 16 specific types of online content. When I went though this list, it was a really eye opener.

Movies? … Well, I do have my Netflix’s subscription.
Television shows? … occasional purchase from iTunes.
Music? … routine purchases from iTunes

As I went through this list, I discovered that I had purchased more online content than I realized. In addition to movies, tv shows, and music … apps (check), software (check), reports or articles (check). In fact, a lot of folks purchased more than one type of content in a given month.

In the end, in a typical month, I spend about $10. This was what most internet users ended up spending.

It will be interesting to see what the percentages and dollar amounts are in the future.

Certainly, with e-readers, smart phones, and tablet PCs, one would expect the newspaper and magazine online content to take off. However, as Victor Godinez mentions in this article, Two-thirds of Web users have bought digital content, study finds, they need to get the subscription and payment model worked out.

I would expect that several business models will develop for online content providers, including:

advertising (e.g., Google)
subscription (e.g., NetFlix)
premium (e.g., Rivals.com)
individual file download (e.g., iTunes)
B2B but free to consumer (e.g., Google and AP)

A couple of interesting aspects with the business models:

This linkage between a tangible product and online content (i.e., ipod and iTunes) might be what works for newspapers and magazines (i.e., tablet and some subscription service).
The B2B agreement but free to the consumer is also interesting. Most of my national newspaper reading is related to my job. My employer, like many, provides print papers free, but not electronic. Would be great to see this employee benefit transition to the online world.

What can you learn from 34,514 social networkers talking about products?

2010-12-16T12:08:00.004-05:00

Classifying Ecommerce Information Sharing Behaviour by Youths on Social Networking Sites reports findings from a research project conducted with Kate Sobel and Geoff Cook.

Using a survey of 34,514 respondents from myYearbook.com, the is an analysis of information seeking and sharing behaviors in the ecommerce domain on four social networking sites (Facebook, MySpace, myYearbook, and Twitter).

Using k-means clustering analysis, we find clusters within this demographic based on levels of being connected (i.e., number of social media sites) on and being engaged (i.e., frequency of status updates) with social networking services.

Many respondents have accounts on multiple platforms.

We then compared the groups based on seeking ecommerce information in statuses, receiving ecommerce information in statuses, acting on ecommerce advice in statuses, taking sponsored polls, and fanning a company.

Specifically, we found the following clusters of users:

• High Connected (HC) – High Engagement (HE) cluster, with accounts on many platforms and very active on all of them
• High Connected (HC) – Skewed Engagement (SE) cluster, with accounts on many platforms but a skewed engagement based on the particular platform, updating usually one social media platform much more frequently than the others
• Medium Connected (MC) – High Engagement (HE) cluster, with accounts on some platforms but updating all of these frequently
• Medium Connected (MC) – Skewed Engagement (SE) cluster, with accounts on some platforms and updating particular ones more than others
• Low Connected (LC) – High Engagement (HE) cluster, with accounts on few platforms but updating them quite frequently
• Low Connected (LC) – Low Engagement (LE) cluster, with accounts on few platforms and rarely updating them

We can also examined this at more specific levels, based on our sample, finding for levels of connection:
• the majority of youth, 56%, are categorized in a low level of being connected, with only one social network site profile
• only 14% are highly connected with four social network site accounts
• 30% have multiple accounts but less than four

Concerning levels of engagement, we find:
• a significant majority of this demographic (74%) is highly engaged, updating their status at least multiple times in a given week
• nearly 7% of this demographic has a skewed level of engagement, meaning that they update their status frequently on at least one SNS but seldom or never update on other social network sites
• only 19% of this overall youth grouping has a low engagement with social network sites. So, these users are very active, regardless of the number of social network site platforms, providing a rich data source for analytics and communication analysis

We then compared clusters on seeking ecommerce information in statuses, receiving ecommerce information in statuses, acting on ecommerce advice in statuses, taking sponsored polls, and fanning a company. There were significant differences among clusters for each, along with some platform differences (i.e., differ behaviors if on MySpace, Facebook, or Twitter).

While one may expect a correlation between seeking - sharing ecommerce advice and being more active in the social media area, one important aspect of this is the concept of acting on this information.

There has been a lingering question in the area of social media marketing on whether or not it is a fruitful endeavor and whether or not users are going to be receptive to any form of advertising. These results shown that users of these platforms are not only receptive to viral modes of ecommerce information but that these people act upon this advice.

Full findings are addressed in the paper, Classifying Ecommerce Information Sharing Behaviour by Youths on Social Networking Sites, in the Journal of Information Science.

Sabbatical Reflection #4:Three months at Pew Internet

2010-12-15T21:14:00.013-05:00

As part of my sabbatical, I signed up to do a gig with the Pew Internet & American Life Project, one of seven projects of the Pew Research Center, which is funded by the Pew Charity Trust.

The overall mission of the Pew Research Center is to be a ‘fact tank’, providing facts, demographics, and population opinions on issues of the day. Their target audience is primarily policy makers, but their reports and findings are freely available to anyone.

During the three month assignment, I spent two days a week at the Pew Research Center, interacting with research staff from both Pew Internet and the other projects. It was a wonderful experience, thanks to the many great folks working there!

Pew Internet has an amazing collection of data dating from 2000 onward. All the Pew Internet data (note, data!) is online as well reports as that summarize some of the major findings. The data is nearly all surveys from national random samples that are generally in the 95% confidence interval with a margin of error +/- 3 percent. The data is also weighted to reflect any known demographic discrepancies. So, this is not convenience sample data. It’s really is a gold mine of research opportunities that academics have just not tapped into yet.

During my time there, I did reports for Pew Internet based on this survey data, including one on ecommerce researching and purchasing and one on the use of the internet by higher income households.

Really a good learning experience for me, as the Pew writing style is geared toward that of newspaper writing, with the main point up front. It has helped me with my academic presentation.

It was also just a good experience spending time in Washington, DC and at the Pew Research Center, which is housed in a just beautiful building. Here is a couple of pictures of the lobby.

Pew Research Center has the entire 7th floor of the building, including a kick b*tt conference room.

The facilities are quite nice, including this industrial strength latte machine in the café, which I put to good use during my time there! :-)

I took daughter #1 to my office one day, and she made me this office wall art, which I just think is just great!

An added benefit was getting to stroll around Union Station, which I find just beautiful!, each week while waiting for my train.

And, of course, I got SO use to seeing this …

My three months regularly visiting Pew Internet were just great! Super folks working there! And, a special thanks to Lee Rainie, who directs a great operation.

Understanding Sponsored Search

2010-12-14T14:31:00.002-05:00

Happy the way my book is coming along, and I'm fairly close to being on schedule even!

Understanding Sponsored Search is a theoretical view of keyword advertising, and an project on this aspect of web search that I've been 'working' … at least mentally :-) … for some time.

In addition to the preface and extensive glossary, I'm working 10 chapters addressing the core areas of keyword advertising:

Chapter 1 A Context for Sponsored Search
Chapter 2 Modeling the Process of Sponsored Search
Chapter 3 Understanding the Potential Customer’s Intent for Keyword Selection
Chapter 4 Sending Signals to the Potential Customer with Ads
Chapter 5 Understanding Consumer Behavior for Sponsored Search
Chapter 6 BAM: Branding Advertising and Marketing for Sponsored Search
Chapter 7 Sponsored Search Analytics
Chapter 8 The Serious Game of Bidding on Keywords
Chapter 9 Bringing It All Together in a Framework of Sponsored Search
Chapter 10 The Future of Sponsored Search

There's a view that theory and advertising don't go together. My goal with the book is to illustrate the theory underlying many keyword advertising practices.

Honestly, it's been a daunting task but things are coming together quite nicely.

Addressing the Digital Divide: Forget broadband. Give everyone a smartphone and free apps!

2010-12-01T12:23:00.006-05:00

The digital divide is an aspect of internet usage that receives a lot of attention from many researchers and practitioners.

The digital divide is the disparity in engagement with the internet among different demographics. Many demographic variables come into to play based on the focus of the researcher, including cultural, social, and economic.

For this last variable, economic, one remedy for addressing the divide (note the assumption is that its to everyone’s benefit to have access, which I tend to agree with) is to provide broadband access to larger numbers of people.

In fact, the Pew Internet & American Life Project, a project of the Pew Research Center, has produced several studies documenting that those with broadband connections access the affordances (i.e., academic jargon for the benefits of a technology) of the internet more than those with just dial-up connections.

While I believe that access to broadband is good, another report, also from Pew, focusing on income and internet usage has an interesting finding that makes me think the focus on broadband to address the digital divide may be:

'a day late and a dollar short'

The finding that I’m referring to is this:

Note that cellphone usage is nearly on par among the three top income groups, and much better than broadband at the lowest income level.

In other words, the market seems to have spoken … forget the computer, the common technology that offers internet access is the cellphone (combined with apps and email clients).

Now, not all cellphones are the same (you can get disposable ones to very expensive ones), but the trend is better, smaller, and more internet access features.

As more and more accessing the internet is done via cellphones and related mobile devices, internet behavior is changing. I’ve noted that the cellphone – app combination has changed my searching behavior. I’m sure it is happening to others also in many areas besides searching.

And, if I wasn’t in the business of writing papers and crunching data (I’m a professor at a research university), I probably wouldn’t need a computer. And, even with my computer–bound profession, I haven’t owned a desktop computer in years, getting by just fine with highend laptops.

So, while broadband may be nice, it seems that cellphones (specifically smartphones with free to inexpensive apps) may be a quicker, more accepted, and certainly cheaper approach to addressing the digital divide than broadband.

The income divide: How those making >$150,000 per year use the internet

2010-11-26T09:02:00.007-05:00

Most of the report, Use of the internet in higher-income households, examines utilization of the internet by four income categories, from <$30,000 to >$75,000.

However, on page 11 of the report, there is an analysis of internet activities by households making more than $150,000 per year compared to those who make less than $150,000. Those making more than $150,000 per year are approximately the top 5% of US households according to the Current Population Survey (CPS) for 2009.

The results are really eye opening, with 96% of this high income group utilizing the internet and email (compared to 66% and 71% respective for the rest of the US population).

Even when comparing those households with internet access, those in the higher income groups participate in a range of online activities in far higher percentages, with many activities having double digit differences.

What I find most striking about these differences is that, for those of us engaged with the internet, we assume that “of course, everyone uses the web!”.

What the findings from this report show is that there is a significant income divide, with those in top of income bracket fully engaged with the internet, but those in the lower income brackets significantly less so.

Read the full 13-page report, Use of the internet in higher-income households.

Double Digit Differences in Use of Internet Among Income Groups

2010-11-26T08:07:00.018-05:00

The report, Use of the internet in higher-income households, examines utilization of the internet by four income categories.

Although, one might expect that those with money would use the internet more than those with less wealth, the differences by income for many activities are quite striking.

For many online activities, there are double digit differences.

Some activities are commerce related, where we would expect the usage to be correlated with income; however, many activities were not commerce related at all.

For example, accessing online news.

And, accessing online medical information.

And, note, that these percentages are for the internet population not the general population. So, we are comparing households in each income group that already have internet access.

Why should we care about this disparity in internet usage by income?

The internet offers affordances that make a person or family more effective and/or efficient in areas such as shopping, getting news, accessing heath care information, enhancing education, etc.

Therefore, there is great value in understanding how various income segments are levering these internet affordances in their daily lives to better measure the impact and reach of the benefits on being online.

It also sheds light on possible obstacles (technical, cultural, educational, etc.) that might be hindering use of the internet, along with aspects that might be facilitating usage.

From the results of this report, there is a disparity in the reach and impact of internet affordances, with the more well-off being the primarily beneficiaries.

Read the full 13 page report, Use of the internet in higher-income households.

Only 52% have used internet for travel reservations?

2010-11-21T23:00:00.009-05:00

I started working at the Pew Research Center’s Internet & American Life Project, and the subject of my first report was online product research.

Basically, 58% of Americans have researched a product or service online (78% of American internet users). Yes, about 25% of Americans *do not* use the internet for one reason or another (I still find this percentage surprising.)

And, on any given day, 21% of Americans internet users have looked for product or service information.

What I really found surprising though, was that 52% of American (74% of American internet users) have booked travel reservations online.

I found this number really low, as in *only* 52% of Americans! Even airline tickets? Next to books, airline tickets seem like one of those products made for internet shopping.

However, the data was inline with other reports, such as the Travelers’ Use of the Internet 2009, so it appeared that the number was accurate.

So, I looked into it.

Guess, what? Americans don’t fly that much. In a given year, only about 42% of Americans travel by air.

11% of Americans say they have never flown in a plane.

In fact, about 25% of Americans don’t travel more than 50 miles from their home in a given year.

So, if you take the percentage of folks who don’t use the internet (25%), those who don’t travel much (25%), you end up with about 50% of the folks who would use of the internet for travel products and services.

Go figure.

Visit to the search engine marketing company IMPAQT

2010-11-11T10:24:00.008-05:00

I had the opportunity to visit IMPAQT, a full service search engine marketing agency headquartered in Pittsburgh, Pennsylvania.

IMPAQT is a really amazing company. Founded in 1999:

48% of their work is for Fortune 500 clients
92% of their clients they’ve retained for seven years
They currently provide services for 102 different brands
They've served 500 brands since their founding

IMPAQT has also been a great sponsored of GOMCHA, providing funding for the IMPAQT Innovation Award, presented to top teams in the local Penn State GOMCHA Challenge, along with providing some great guest speakers for the courses.

During my visit, I gave a talk on the use of sponsored search analytics. IMPAQT is well known for their use of data analysis, so hopefully the talk went over well. Had some really good discussions and questions during the talk.

During the presentation, I incorporated data from the SEMPO Annual Survey showing the expected growth of search engine marketing ad spend.

The whole visit was great, including getting to chat with two of my former students who now work at IMPAQT, Alyssa Miller and Laurel Upton.

Just a super visit and a big thanks to IMPAQT for inviting me!

Visit to the Yahoo! Research Lab New York City

2010-11-10T15:44:00.013-05:00

I visited the Yahoo! Research Lab in New York City, interacted with several folks there individually, and presented to the group on Using Sponsored Search Analytics for Predictive Modeling and Forecasting.

I used some data from the Pew Research Center’s Internet & American Life Project on Online Shopping in my presentation.

It was a wonderful visit, with some great research conversations and exchanges on a variety of topics, including sponsored search.

Dave Pennock hosted me. Dave has done a lot of work in sponsored search auctions, with many of his papers available on his webpage.

Another Yahoo! researcher, Sébastien Lahaie is leading an effort examining user behavior in sponsored search.

While there, Andrei Broder was in the office for a speaking engagement in the city on computational advertising, which is a computer science spin on keyword and content advertising.

So, all in all, a super visit!

Yahoo! Research in NYC is in a great place, located in midtown Manhattan overlooking Bryant Park, which is right behind the New York Public Library. Really fantastic place.

And, I really liked the office space, with plenty of the Yahoo! purple!

The Future of Search

2010-11-07T09:06:00.009-05:00

As part of a research project funded by Microsoft, I participated in an interview on The Future of Search.

I was send the final research findings and some of the conclusions are interesting.

Certainly, some of the predictions are 'way off' the mark; however, a synthesis of opinions from 54 folks from a variety of fields on the topic of where search is going can be insightful to get one's own thoughts going.

For a short synopsis of the major findings:
• Value through Structure: Search will increasingly structure a wider range of data
• Community Participation Will Play A Greater Role In Search Quality: Human participation includes a variety of activities that filter or augment information, making it more useful
• Personalized Interaction: To deal with user frustration and difficulty in finding relevant search results, we will see a shift towards providing personalized results that leverage user history and
• An Expansion Of Search Activities On Mobile Devices: Mobile search will grow faster than PC search, but both will grow.
• A Growth in Specialized Vertical Search Sites: Vertical sites like Yelp, Kayak and Amazon have become more popular, have brand recognition, and maintain user data that cannot easily be accessed by search engines
• Search Engine Monetization Model Will Change … [I thought this was by far the weakest section. From the list of participants, it appears that this sector was unrepresented in the sample, which is ironic, as the search engine business model has shaped the web as we know it.]

In addition to aggregate results, the report has quotes from individuals, which is quite insightful for seeing how particular folks are viewing the search field.

Aside from me, some commentaries are from:

Oren Etzioni, University of Washington
Rob Hayes, First Round Capital
Timothy Chang, Nor West Ventures
Rob Coneybeer, Shasta Venture
Jake Seid, Lightspeed
Jeff Beard, Localeze
Dave Patterson, University of California, Berkeley
Lukas Biewald, Founder, Crowdflower
Nathan Eagle, MIT
Jerry Michaliski, Sociate
Michael Cafarella, University of Michigan
Jens Lapinski aiHit Ltd
Victor Zue, MIT
Chris Burges, Microsoft
Mark Torrance, Rocketfuel
Brad Feld, Foundry
Louis Gray, Palladin
Rajesh Lalwani, Owner, Blogworks India
Ben Finkel, Fluther
Susan Dumais, Microsoft
Eyan Adar, Univ. of Michigan
Siva Kumar, The Find
Michael Dearing, Harrison Metal
Oren Etzioni, Academia
Andrew Yates, Silobreaker

I’ve posted the full report here, Industry Signals on the Future of Search.

Classifying the user intent of web queries using k-means clustering

2010-11-01T11:16:00.010-04:00

An active research area in web search is the identification of user intent (i.e., what type of content is a searcher looking for).

The article, Classifying the user intent of web queries using k-means clustering, reports on research using 130,000 web search engine queries that are categorized as informational, navigational, or transactional based on a variety of query traits using a k-means clustering approach.

The research findings show that more than 75 percent of web queries (clustered into eight classifications) are informational in nature, with about 12 percent each for navigational and transactional. Results also show that web queries fall into eight clusters, six primarily informational, and one each of primarily transactional and navigational.

Some examples of queries classified as informational include “flowering climbing vines” (cluster 6) and “hazards on football fields” (cluster 2). Both of the queries are looking for information about specific topics. They do not seem to be geared towards navigating to a particular site or completing a transaction. Instead, the user appears to wants to find web sites that will provide more information about these two topics.

Navigational queries, on the other hand, include queries where the user is trying to get to a particular web site or class of sites. For instance, two queries classified as navigational (cluster 7) were “university of north carolina” and “image search engines”. One of these queries was attempting to get to a site concerning a university while the other wanted a site where the user could search for images on the web. In both of these cases, the users were trying to locate specific types of sites instead of trying to get information about a subject as the informational queries were.

Transactional queries (cluster 4) were queries attempting to complete a transaction. Examples of this type of query were “download hp simple backup” and “buy camping gear”. For the first query the user appears to have intended to want to find a web site where they could play or download a software program, while in the second query, the user appears to want to purchase camping gear. So, this query has a transactional goal instead of a purely navigational or informational goal.

I believe the key aspect of the research is that researchers are getting closer to more automated and accurate methods of user intent identification, where earlier research was more qualitative analysis of smaller datasets.

Comments on the 2010 Washington Ideas Forum

2010-10-01T14:21:00.004-04:00

I attended the 2010 Washington Ideas Forum, hosted by The Atlantic, The Aspen Institute, and the Newseum.

It was an super interesting and fast paced series of presentations. Most of the headliners were interviewed by a reporter in front of the audience. So, some good interaction.

My overall take on the reoccurring themes (followed by my reflection):
- much more taking responsibility for the issues in the country and world than I believe most Americans assume from the Washington folks (as a citizen this made me feel good)
- the military continually came up as the one governmental organization that can be counted on to actually accomplish something (this really bothered me, as a lot of problems aren’t military, which also came up)
- there was only one comment about the failure of the press (which I found somewhat troubling as they have some serious problems and the conference was held in the Newseum)

The event was streamed live, and video snippets are online. So, here are some bullet reactions from me.

David Rohde (the journalist held captive in Pakistan for 7 months)
- we’re sending Pakistan billions and they aren’t doing much to help us
- there is a lack of education in the Taliban fighters which is a great aid in their propaganda efforts to their own
- Wow! Great sense that this dude has some on the ground experience.

Richard Holbrooke (senior advisor to Obama) with Christiane Amanpour
- Amanpour is a great interviewer. Sharp, crisp, and with pressure, but friendly
- Holbrooke commenting on the Afghanistan and Pakistan situation, he said complicated. I would use a more messy term
- Holbrooke quoting Moa Zetung “With two good [i.e., ruthless] men, I can take any village” .. Go in, kill the largest landowner, kill all the officials, marginalize the good people.

Spike Lee with Douglas Brinkley
- Lee: commenting on Bush “a lot of people are dead because of what he failed to do"
- really liked Spike! Spoke a lot from the heart.

Tony and Heather Podesta (a couple of lobbyists) with Ronald Brownstein
- Tony Podesta: lobbyist are most often asked to change the general direction of legislation rather than something very specific

Kenneth Feinberg (ran the Sept 11 fund, Virginia Tech fund, and now the Oil Spill) with Walter Isaacson
- Feinberg: there were about 11 families that could not take money from the Sept 11 fund due to grief; they couldn't sign. Even after great pressure, there were still a couple families that didn't take the money
- really liked this interview, empirical and moralistic at the same time

Governor Haley Barbour with Jeffrey Goldberg
- this was by far the funniest interview. Both Barbour and Goldberg were really insightful and really entertaining
- Barbour: “if the Republicans don’t end up with 30 governorships, I'll be disappointed”
- Barbour: gave interesting insights on Republican gains in New England, especially at state level

Eric Schmidt with James Bennet
- Schmidt: a lot of praise for the research coming out of universities, noting that the military funding made the modern research university and continues to
- Schmidt: from looking at a lot of data about what people do, most people are very positive
- Schmidt: mentioned consent at least three times on data collection

Senator Jim Webb with John King
- I’m fan of Senator Webb
- obviously he is not happy with the way Washington is run
- Webb: at the beginning of the Obama administration, a focus on health care when the voters cared about jobs and the economy was a mistake

Political Panel: Chuck Todd, Rob Brownstein, Charlie Cook, and Margaret Carlson, with John King
- this panel was great, and really liked Chuck Todd’s insights and expressions
- Todd: commenting on comments about the current collection of candidates, mainly from the Tea Party. “As my mother says. We sent the competent people to Washington. Look at what a good job they did!”
- Todd: the Obama administration is not prepared for the change coming post-mid term elections

Sabbatical Reflection #3: At the Pew Research Center

2010-09-15T18:56:00.007-04:00

One of the purposes of a sabbatical is to break old professional patterns and do something new and different. A change of pace can be really energizing, with some positive carry-over when one returns to the ‘normal’ job.

I am in academia, an associate professor at the College of Information Sciences and Technology. When I had the opportunity to do a sabbatical, I wanted to do something that was outside of academia. So, I intentionally decided not to go to another university.

A lot of professors write books during their sabbaticals. Although writing a book is fine (I’m actually writing a book during my sabbatical, Understanding Sponsored Search), this just didn’t seem ‘different’ enough for me.

After exploring various options, it seemed that the Pew Internet and American Life Project, a project within the Pew Research Center was a perfect place for me. It has a great fit with my prior and current research, and I was, of course, familiar with the reports and research the project staff produced.

After some coordination, and thanks to Lee Raine, I have the opportunity to spend my sabbatical with the team of excellent researchers and staff on the project. Within the first week, we had narrowed some possible research topics for me to pursue. The data sets from the Pew Internet surveys are quite rich. So, I am really excited about the research possibilities.

The Pew Research Center is funded by the Pew Charitable Trust, which was founded by Joseph N. Pew Jr and is one of the most well funded ‘fact tanks’ around. The Pew Research Center’s mission to provide data about topics of interest and not to take a side on questions of policy.

In addition to the research and the co-workers, I really like the building. Here are a couple of pictures.

Understanding Sponsored Search: A Coverage of the Core Elements of Keyword Advertising

2010-09-10T13:31:00.010-04:00

One of the projects that I’m working is writing a book that articulates the theoretical and methodological foundations of sponsored search. It is a book that, I believe, is needed in both the academic and practitioner areas.

Sponsored search (a.k.a., keyword advertising) is a technology, a business process, an information searching concept, a marketing effort, a cognitive experience, and an advertising model that has had pronounced impact on millions and millions of people and businesses, and therefore, impacted our society and culture.

One could rightfully say that the Web would look very different today if it was not for the revenue streams provided to the search engines from sponsored search and that many businesses would look different if not for the revenue generated via sponsored search campaigns.

Without sponsored search, the search engines would be different. People would use the web differently. Ecommerce would be different. Providing online information would be different. So, the effects of sponsored search have really been dramatic. And, it only got started in 1998!

There are a lot of good "how to" books on different aspects of keyword advertising, including Search Engine Optimization: Your Visual Blueprint for Effective Internet Marketing, Advanced Google AdWords, and Customers Now: Profiting From the New Frontier of Content Based Internet Advertising. (Note: There are other good ones, I just selected three that I'm familiar with.)

So, here are a lot of 'how' books. But, there is no 'why' book?

Why is sponsored search the way it is?

Why do some techniques work and others don’t?

Why has sponsored search been so successful?

Drawing on the academic fields of information science, consumer behavior, and advertising, along with aspects of computer science, cognitive science, marketing, and statistics, Understanding Sponsored Search addresses the ‘why’.

Target date for final manuscript: 31 December 2010. Really got to get busy!

Significant trends and technologies for Web Searching

2010-08-30T12:33:00.011-04:00

Was reading the State of Search Engine Marketing Report 2010, and this particular chart caught my eye, with some interesting findings for both academics and practitioners.

The question was “In the context of your search marketing efforts, how would you describe the impact of the following trends and technologies?”

First, 75% of the respondents stated that “personalization of search results” was significant or highly significant. With the use of search engine tool bars and algorithmic changes for result rankings, different individuals are increasingly seeing a more personalized search listing. The implications are that search engine marketing is going to continue the move from the mass advertising approach even more than it has already.

Second, 65% commented that the mobile internet was significant or highly significant. Not a big surprise here, as the search engines are reporting some really unbelievable percentages of searches generated via mobile phones.

Third, real time search was rated significant or highly significant by 61% of the respondents. Hopefully, academic folks (if they read this report) will begin focusing on this really innovative aspect of web search. It’s an area that is wide open for research implications.

Disclaimer: I serve on the research committee for SEMPO

Two Teams in Final 15 of GOMCHA 2010!

2010-07-20T20:18:00.012-04:00

The results of the Google Online Marketing Challenge 2010 are posted.

In the 2010 Google Online Marketing Challenge (GOMCHA) 2010, there were 3,000+ registered teams from more than 70 countries, which is about 33% more teams than in 2009 and double the number of teams from the inaugural year in 2008.

Very happy (and proud!) to announce that *two* Penn State teams from the keyword advertising course that I taught this year placed in the Final 15!

The Final 15 are the top 15 campaigns worldwide!, which is a wonderful accomplishment for the students! Really proud of them and proud of all the teams from the course, as they all did extremely well.

The two teams placing in the Final 15 are:

Team Awesome (Erica N., Chrissy C., and Stef A.)

Team AdVenture (Garrett M., Danielle W., Andrew G. )

The teams were really multi-disciplinary with Team Awesome a mix of advertising and marketing students and Team AdVenture a mix of technology and advertising students.

Just to put it into perspective what these students did:

3,000+ student teams from 70 countries completed the Challenge.
at the conclusion of the Challenge's advertising campaign, Google did an algorithmic evaluation of the campaign results, narrowing the field to 150 teams (0.05 of the 3,000).
then, Google employees did a manual review of these campaigns, trimming the 150 to 15 (0.005 of 3,000)

To have two teams from Penn State place in the top 15 teams is really astonishing and speaks really highly of the Penn State students.

Here is a blog posting of the end of course GOMCHA 2010 celebration that we hold at Penn State, where we conduct our own Challenge.

A big thanks to IMPAQT, which sponsors our local competition with the $1,000 IMPAQT Innovation Awards, and the Smeal College of Business and the College of Information Sciences and Technology for providing support for the course.

Finally, we thank Google, for sponsoring GOMCHA, and Jamie Murphy, who is the academic lead for the project. Really a fantastic learning opportunity.

Again, congrats to Teams Awesome and AdVenture!, and all the students from 70+ countries who competed.

Sabbatical Reflection #2: Accepting my SRTE Scores

2010-06-28T08:36:00.017-04:00

I’m now on sabbatical, and I’m using part of the time to take stock of where my professional academic life has been, where it’s at, and where it’s going. So, I’m doing a complete evaluation of my research, teaching, and service.

In the teaching area, one thing that I’ve decided to is to accept the typical range of my SRTE scores (i.e., Student Rating of Teaching Effectiveness scores) and ditch any desire of getting them higher.

At most US universities, professors in each course each semester are evaluated by students who took the course during that semester from that professor. These student evaluations are typically used in promotions and annual assessments.

At Penn State, the student evaluations are on a 7 point Likert scale (1- the worse to 7 – the best). There are a dozen or so questions, but there are only two that matter.

Rate the overall quality of this course
Rate the overall quality of the instructor

My SRTEs scores are generally in the high 5’s to low 6’s for most undergraduate courses. Sometimes a little lower. Sometimes a little higher. Generally, though, they are in this range.

However, I would have liked my scores to be in the high 6’s. I wouldn’t call it a goal, but just something that would be nice and make me happy.

However, upon some personal reflection and analysis, I’m going to ditch this desire. I don’t believe it is reflective of my personality or teaching style. (Note: This is not to say that I can’t or shouldn’t improve my teaching. I’m seriously looking into this. Just saying that I’m ditching any intend of higher SRTE scores.)

I’ve analyzed my SRTE scores over the past few years and from sitting on faculty committees reviewing the packets of other professors, I’ve had the opportunity to analyze the SRTE scores of others. As a result, I have arrived at some findings that indicate my SRTE scores are probably going to remain in the range they are currently at.

Why do I say this?

Well, here are my SRTEs from two courses from the Spring 2010 semester in response to the question: Rate the overall quality of the instructor.

Course 1 - Overall SRTE Evaluation: 6.18

Rating - 1 - 2 - 3 - 4 - 5 - 6 - 7
Count - 0 - 0 - 0 - 5 - 5 - 12 - 23
Percentage - 0% - 0% - 0% - 11% - 11% - 27% - 51%

Course 2 - Overall SRTE Evaluation: 5.86

Rating - 1 - 2 - 3 - 4 - 5 - 6 - 7
Count - 0 - 1 - 0 - 4 - 5 - 14 - 12
Percentage - 0% - 3% - 0% - 11% - 14% - 39% - 33%

This distribution appears in nearly every evaluation period for me. There are about 75% (or so) of the students rate me a 6 or 7, but there are about 10% (or so) of the students who rate me 4 (average) or lower. Can’t get high 6s with the lower outliers.

I also have identified trends from reviewing packets of professors who did get SRTE scores in the high 6s. These trends are (note: trends, so there are always exceptions):
- the instructors were generally ‘nice’ people
- the courses were generally at the junior level
- the courses were generally dealing with a ‘fun’ subject matter
- the response rate (i.e., number of SRTEs/number of students enrolled) was generally about 60%

These are some additional reasons that I’ve decided to let my wish of higher SRTEs go.

Why?

I’m not nice in the classroom, but rather demanding and straightforward in evaluation. I typically teach graduating seniors, usually in their final semester. The subjects I teach aren’t necessarily fun, and I have a mandatory attendance policy in my classes, so even the students that really don’t want to be there have to show up.

Actually, I believe this decision will actually improve my teaching. With this distraction out of the way, I can channel my full energies on improving effectiveness on the gamut of my teaching duties.

Plus, it's important to accept who you are.

Note: Several studies have pointed out that SRTEs have little to no correlation with teaching effectiveness. Here is an interesting blog posting on the subject of SRTE scores and teaching outcomes, which discusses an academic research paper on the subject.

Enhanced Messaging for the Emergency Response Sector

2010-06-24T10:53:00.007-04:00

I am involved with this interesting research project, funded by the National Science Foundation, Enhanced Messaging for the Emergency Response Sector (EMERSE).

The goal of the research project is to develop and deploy, in collaboration with an NGO coordination body (NetHope), a reusable text message-based infrastructure that classifies and aggregates multi-lingual tweets and text messages so that they can be easily subscribed by NGO's, survivors in relief areas, and their friends and families. We are using data collected about Haiti relief operations as our sample set.

It is an interesting project because it addresses many traditional information searching and information retrieval issues, social networking aspects at the organizational level, and the emerging importance of social media platforms, such as Twitter, for information searching and sharing.

Sabbatical Reflection #1: I’m Done with Academic Conferences

2010-06-10T14:03:00.004-04:00

My sabbatical de facto started mid-May 2010, and my idea is to use part of the time to take stock of and reflect on where my professional life had been, where it’s at, and where to go from here.

So, I plan to do a complete evaluation of my research, teaching, and service for both efforts and goals. In fact, I’ve already started.

In the research and service area, one thing that I’ve already decide to do is near totally ***cut out academic conferences***. I am only going to go to one conference a year, and this is just (primarily) to socialize.

I have commented on what I think of academic conferences before (see Academic Conferences as Research Outlets, Publishing at Conferences versus in Journals, and Why Am I Getting Better Conversations in the Blogosphere than I Do at Academic Conferences?), and others also have less than stellar opinions of academic conferences (see Time for Computer Science to Grow Up and A must read for computer scientists: Lance is right: Time to grow up).

Here is the problem -- in a cost – benefit analysis, the cost of academic conferences greatly out weighs the benefits. The costs including the travel expense, registration (increasingly expensive), time away from family, missing classes, and lost research time. Certainly, there are some benefits (meeting some new people and being introduced to some interesting ideas via discussions with folks).

However, after one or two conferences a year, I believe there is diminishing return on these benefits. Additionally, many academics travel in packs, going to the same conferences, meeting the same people, and hearing the same ideas. So, I’m not seeing the new ideas! Where are the opportunities to meet the new people?

So, I just don’t see a string of conferences as being worth it. One (maybe two) conferences a year will do it for most academics, including (I believe) those academics seeking tenure letter writers. (Spend the most of your time getting good research published!)

Now, this is step one for me. The next step is discover ways to have academic conferences better serve the academic community.

I’ve begun brain storming approaches to make conferences more beneficial for those in attendance. How can we set-up a conference to introduce *new* ideas? What can we do to increase individual interaction? What can we do to keep costs in both time and money down? This step is still a work in progress.

Disclaimer: I am on the editorial boards of eight journals (Journal of the American Society for Information Science and Technology, Information Processing & Management, Information Research, Future Internet, International Journal of Internet Science, Journal of Internet Research, Journal of Emerging Trends in Computing and Information Sciences, and (until recently) Library and Information Science). Academic journals have their problems also and need to change. More on that later.

Is the Era of the Web Search Engine Over?

2010-06-05T17:08:00.007-04:00

I’m a history buff, and, as history buffs know, one can segment history by which country is the alpha dog at a given time (i.e., think Pax Romana).

As an analogy, we can take this same perspective and say that the last decade+ (say 1998 to 2010) has been the Era of the Web Search Engine on the Internet. The search engines were the technology, virtual place, and service that nearly everyone went to or used to find information. The search engines shaped the Web, online information usage, and ecommerce. It reached a point, for me, where I really depended on them in my daily life.

With the rise of social media and app technology, is the era of the search engine as *the* place to which one turns to search for information coming to an end?

As a person who really (really!) enjoys technology and understands it, I am always skeptical of claims that such and such technology will “change the way we live”. I typically gauge these claims by the 100-10-1 rule. 100 systems will be built. 10 will get some buzz. 1 will actually have some impact on people.

Although a computer scientist, I really evaluate technology by the way people use it. No other technology claim matters.

However, the combination of social media - app technology really has captured my attention as having a changing impact on online information searching.

So, I decided to evaluate for one day how I now search for information on the Web.

How do I search today?

In my one day data collection period, I did the following:
1. When to Twitter, Facebook, and Linkedin (which I have bookmarked) to check status updates from those in my social network. Skimmed most status messages, commented on two, and read 5 articles that these folks sent, stored 3 links as background material for a course that teach.
2. Checked two standing searches of mine on Twitter, ‘GOMCHA’ and ‘ASIST2010’. Both are short term standing queries that will last a few months in duration.
3. Use the YellowBook app on my iphone to location a restaurant, which then took me to Google maps for directions.
4. Did 5 navigational searches (all for my own homepage). I was working on a research manuscript and needed to access some papers on my Website. Searching for my name is quicker than typing in the URL. Used Google.
5. Did 4 known item searches, all were titles of research papers. I was working on some reviews and wanted to provide references. Used Google Scholar.
6. Received three Email Alerts from Google (two variations on my name and one concerns a consulting project that I am working on). These are standing queries that I have with Google. Two are fairly permanent and one is of short duration.
7. Did one information search on Google to get a list of restaurants in a section of my hometown. I was providing a comment to a friend’s Facebook status message asking for a restaurant recommendation and wanted to spell the restaurant’s name correctly. Also, checked out the Yelp reviews to see what others thought.
8. Did a search on Google for a person’s name along with the term ‘LinkedIn’, as this person started following me on Twitter, and I wanted to see who they were. Immediately went to LinkedIn to check out their profile.
9. Read two news articles on Yahoo!, which is my homepage.

This was a shocker for me!

Now, I am a person who is really is on the Web nearly 24-7. A year ago, I can safely say that I did 20-30 informational searches in a given day using Google and occasionally Yahoo! on a variety of subjects.

Today, one year later, my use of the search engines is almost secondary to (or perhaps integrated with) getting my task done, used as a navigational tool, memory aid, fact checking, or an add-on to an app.

I believe that the social media – app combination might be the 1 out of a 100.

Search Query as Trademarked Terms

2010-06-02T13:18:00.010-04:00

Completed a research project with Mark Rosso concerning the use of trademarked terms in keyword advertising.

The issue is this: some advertisers don't like it when other advertisers use their trademarked terms to trigger ads. (Note, this is not the use of a trademarked term *in* an ad. That is pretty much a no-no for competitors.).

The advertisers say it is trademark infringement because the search engines are selling their brand terms. The search engines say they do not sell keywords; they sell ads.

This issue has resulted in a variety of complaints, including lawsuits against the search engines, mainly Google. For a nice run down of legal issues and lawsuits pending, see Technology & Marketing Law Blog by Eric Goldman and Common Issues Facing E-Commerce Businesses by Jonathan Frieden.

Rather than a legal perspective, in our research, we investigated the issue from an empirical information science view by seeing how often this type of thing happens and what the effect is when it does happen.

For our research, we submitted 100 search queries consisting solely of one of the 100 top global brand names to three major search engines, Google, Yahoo!, and Microsoft.

We then analyzed 8,345 results from the search engine results pages.

Findings? It doesn't happen that much, and the impact appears minimal.

Only 4% of sponsored ads were triggered by competitors’ trademarked terms. When it does happen, the results are pretty much what consumers are use to seeing, so there doesn't appear to be many negative consequences. For example, see what Bing does algorithmically, which appears to be pretty much like using a trademarked term to trigger an ad.

Thus, competitive use of trademarked terms to trigger online ads does not appear to be a widespread phenomenon and is similar to the query suggestion feature that many search engines employ.

The full report will be published in the journal, Communications of the Association for Information Systems. However, preliminary results appear in:

Rosso, M. and Jansen, B.J., (2010) Smart Marketing or Bait & Switch? Competitors’ Brands as Keywords in Online Advertising. 4th Workshop on Information Credibility on the Web (WICOW 2010). World Wide Web Conference (WWW 2010), Raleigh, NC. 26-30 April.

How did the status message get this important?

2010-05-14T16:41:00.006-04:00

It 's sometimes hard to believe that the status message, once a simple “I’m away” notice has become the hottest thing on the Web!

In a co-authored paper with Abdur Chowdhury of Twitter and Geoff Cook of myYearbook.com, we explore the history of the status message and how it has become so important on the Web today, including its impact on real time search.

We present data showing that people really do use the status message for information sharing, including sharing information about products and services (using data from Twitter), and that these status messages from others within one’s social network have an impact on people’s behavior (using data from myYearbook.com).

The material is presented in The Ubiquitous and Increasingly Significant Status Message.

Identifying the Seventeen Theoretical Constructs of IS and IR

2010-05-09T22:41:00.006-04:00

Information search and information retrieval are two very practical and generally empirical fields that are synergistic.

However, it is important to take stock of where the fields are headed so that one can truly understand the implications of one’s research. It is truly amazing that, as academics, we reflect little on our own research field within the broader context.

I did take stock of and reflect on the two fields of information searching (IS) and information retrieval (IR) in a co-authored paper with Soo Young Rieh entitled, The Seventeen Theoretical Constructs of Information Searching and Information Retrieval.

With 181 references, this paper presents the 17 underlying theoretical constructs of the fields of IS and IR. These constructs are drivers for most of the research in these two fields.

In the paper, we organize these constructs in a framework that compares and contrasts the two fields. We believe that an explicit presentation of these constructs will enable a more in-depth understanding of the fields of information searching and information retrieval, along with the relationship between the two.

The 17 constructs are:

Multiple Definitions of Information: Information is a fundamental concept with a range of definitions.
Hierarchical Relationship of Information: There is a hierarchical relationship among data, information, knowledge, and wisdom from the perspective of an information searcher.
Perceived Benefit of Information: An information system will tend not to be used when it is more painful and troublesome for a user to have information than for the user not to have it (Mooers, 1996).
Relevance: Relevance is a key performance indicator of searching or retrieval.
Information Representation: Information can be represented algorithmically by the sum of its attributes.
Information Ranking: Information that addresses an information need as expressed by a query can be ranked in order of metrics.
Document Similarity: If a document is relevant to a given query, than similar documents will also be relevant.
Uncertainty Principle: A user engaged in Information searching is in the process of attempting to resolve some uncertainty in tacit knowledge (Kuhlthau, 1993).
Principle of Least Effort: A user of an information system will adopt a course of action that the user perceives will involve the expenditure of the least effort to locate the desired content (Zipf, 1949).
Searching as an Iterative Process: Searching may involve multiple processes of interacting with information systems until the underlying information need is satisfied or met.
Interaction: Information searching and information retrieval involve interactions between a user and a system or information.
Information Provision: Providing information to a user is beneficial in order that a user can accomplish a certain task.
Preference of Channel: People have preferences of media and technologies when attempting to obtain information.
Information Obtainability: Information is used in direct proportion to how easy that information is to obtain (Summit, 1993).
Query: A user’s information need is represented as a question and then transformed into a query that an information retrieval system accepts.
Neutrality of Technology: Information retrieval systems present unbiased content.
Memex Vision: Technology is the solution for making information available to people (Bush, 1945).

References
Bush, V. (1945). As We May Think. Atlantic Monthly, 176, 101-108

Kuhlthau, C. (1993). A Principle of Uncertainty for Information Seeking. Journal of Documentation, 49, 339-355.

Mooers, C.N. (1996). Mooers’ Law; or why some retrieval systems are used and others are not. Bulletin of the American Society for Information Science, 23(1), 22-23.

Summit, R.K. (1993). The Year 2000: dreams and nightmares. Searcher, 1, 16-17.

Zipf, G.K. (1949). Human Behavior and the principle of least effort. Cambridge, MA: Addison-Wesley Press

Jansen, B.J, & Rieh, S. (2010). The seventeen theoretical constructs of information searching and information retrieval Journal of the American Society for Information Science and Technology DOI: 10.1002/asi.21358

Google Online Marketing Challenge Celebration at Penn State for 2010

2010-04-24T21:41:00.036-04:00

The Google Online Marketing Challenge 2010 has come to an end at Penn State, with a great field of teams comprised of students from ten different majors.

We marked the completion of the Challenge with a celebration dinner and award ceremony at Rotolli, both reflecting on the experiences of the Challenge and presenting the $1,000 IMPAQT Innovation Awards, which IMPAQT (a PA-based search engine marketing firm) sponsors. Representing IMPAQT at the celebration was Zach Simon, who presented the awards.

We had 16 teams compete this year, with most teams having a mix of advertising, marketing, and technology students (one has to do some recruiting to get this mix.). I was really proud of all 16 teams, as they ran some outstanding campaigns, used some advanced search engine marketing techniques, and did some really creative things with their advertising efforts.

The Challenge is a big thing at Penn State, with PSU GOMCHA t-shirts, the celebration dinner, and the IMPAQT Innovation Awards. Additionally, we had several outstanding guest speakers during the Challenge this year, including:

A thanks to all the guest speakers for sharing their insights on keyword advertising with the students. It’s wonderful to get this practitioner view of pay per click advertising.

Also, thanks to Brad Geddes of bg Thoery, who donated copies of his book, Advanced Google AdWords (autographed!), which we used as prizes for some of the teams that did really well.

The 16 teams we had competing in the 2010 Challenge were, in no particular order, are:

Team JAB (Bill N., Abby S., Jackie D.) 1st Place

Team Awesome (Erica N., Stef A., Chrissy C.) 2nd Place

Team Click Here (Ting S., Rachel A., Craig L.) 3rd Place

Team Blitz (Sophie G., Pat M., Morgan T.) 1st Honorable Mention

Team AdVenture (Garrett M., Danielle W., Andrew G. ) 2nd Honorable Mention

Team Just Bananas (Chrissy A., Trevor C., Sarah O.)

(no picture)
Team MEL Interactive (Lauren M., Mariya S., Emily S.)

Lady Google: Caught in an Ad Romance (Jordan A., Hannah Q., Benn S.)

Six Pack Ads (Jill M., Pete S., Kristen B. )

Team Spain (Roxanna S., Tyler F., Lauren S.)

Lucky Number Seven (Greg T., Heather F., Rich H.)

Team Boarders (Cory T., Stef L., Megan G.)

Dream Team (Maurice Z., Jamie S., Sam C.)

Googlites (Kate S., Lauren A., Karen L.)

Blonde Throttle (Marissa N., Laurel U., Alicia S. not pictured )

Team Encantado (Emily W., Kaitlin K., Scott M. not pictured)

Also, a big thanks to the teaching assistants, Brooke Randel, Mitch Rukat, and Krishna Makadia.

Their efforts really helped make the course a great learning environment. Brooke and Mitch were members of the Penn State GOMCHA 2009 winning teams.

We also thank the Smeal College of Business and the College of Information Sciences and Technology for providing support for the course.

Penn State has done pretty well in the prior GOMCHAs. In the 2008, a Penn State team placed first in the Americas region in GOMCHA. In 2009, there were three Penn State teams in the top 15 in GOMCHA.

Finally, we thank Google, for sponsoring GOMCHA, and Jamie Murphy, who is the academic lead for the project. Really a fantastic learning opportunity.

Wrapping up the Google Online Marketing Challenge 2010 at Penn State

2010-04-21T19:13:00.006-04:00

The Penn State Google Online Marketing Challenge 2010 teams are wrapping up their campaigns and reports. We are preparing for our local Victory Celebration and Awards Ceremony (see this entry on last year’s celebration, Google Online Marketing Challenge 2009).

The competition was quite tough this year, with several strongly performing teams, which I'm really happy about.

We are holding our celebration dinner this year on Wednesday 28 April at Rotelli, where we will honor some of our highest performing teams with the IMPAQT Innovation Award, sponsored by IMPAQT, a Pittsburgh based search engine marketing firm.

IMPAQT has been a good friend of GOMCHA at Penn State. Representing IMPAQT this year will be Zach Simon, Marketing Coordinator at IMPAQT.

Looking forward to an exciting evening!

Using the University Entrepreneurship Competition as a Class Project

2010-03-31T11:59:00.006-04:00

The undergraduates in our college have to take a capstone course that brings together major portion of their college and general education course into an evaluation environment. I have been teaching this capstone course for several years, experimenting with a variety of projects.

Ever since participating in the Google Online Marketing Challenge, I’ve been interested in leveraging competitions for the classroom. These types of competitions are a measure of how the students (and instruction) stack up against external metrics.

I also believe that these competitions are excellent resume fodder for the students, provide feedback for the institution (e.g., we can say we’ll good but these are external metrics to use for assessment), and may be good vehicles for teaching – industry collaboration.

So, for the first project in the capstone course, I had each student team enter the Penn State IdeaPitch competition. It required a much more business focus than what the college’s technology students were use to;however, I just related all the stuff to project management and IT systems.

I believe the experiment turned out really well.

One of the team got 1st and another team got 3rd place in the competition, Team of IST students wins Penn State IdeaPitch competition.

One thing that really made the class interesting was a set of wonderful guest speakers, which were:
Mark Cramer, Mark Cramer of Surf Canyon Guest Speaker for Penn State IST Course
Jason Calacanis, Jason Calacanis's presentation to Penn State students
David Rusenko, David Rusenko from Weebly Presents to Penn State IST students
Alex Mann, Alex Mann of Trendrr Speaks to Penn State Students

I am going to experiment with future courses to see if this competition approach will work. So far, it has been extremely successful.

Howie Jacobson speaks to Penn State GOMCHA students

2010-03-23T21:42:00.008-04:00

Howie Jacobson, author of AdWords for Dummies, spoke to Penn State GOMCHA students, giving a really astute presentation on keyword advertising.

Howie breezed through the mechanics of keyword adverting and Google AdWords, and he instead focused on understanding the intent of the customer and the context (my term, not Howie’s) that the customer is in.

Howie called it the "mind of the searcher". And, he related the importance that individual keywords have in discerning this user intent.

He really advocated developing scenarios of the customer that went well beyond “what are they searching” and into how they are feeling, why they are searching for this, and how the product would make them feel or what it would make them do. He referred to this as making an affinity for the product in the searcher.

This is really doing a persona for a target customer, but Howie pitched it in such an interesting way that it really made an impact on the students.

He also had a couple of other really key suggestions, including:
- the DNA of keyword advertising is to get into the mind of the searcher
- there is great value into coming to keyword advertising as a beginner (because you aren’t 'too into' the client’s business)
- for a keyword advertising campaign, select and optimize for the one keyword (I realized that this is a really good exercise to help focus on what’s important)

Howie is updating his AdWords for Dummies book with free chapters

Courtney Kriebel of Razorfish Speaks to Penn State GOMCHA Course

2010-03-16T21:02:00.003-04:00

Courtney Kriebel of Razorfish spoke to Penn State students in the Google Online Marketing Challenge course (GOMCHA course). Visit the GOMCHA Facebook Fan Page.

Courtney’s presentation was really interesting and insightful, with some great advice, especially concerning the content network, including:
• Don’t use the same keywords for content and paid search
• Create different copy from the search network
• Use the site and category exclusion tool once the campaign starts running

Always excellent to get advice from one actually doing keyword advertising! Plus, her presentation was really professional!

I was on a panel sponsored by the Penn State Advertising Club with Courtney, Alison Lentz (also of Razorfish), and Mandy Mazzseo (at the time from Acronym Media).

Penn State GOMCHA 2010 t-shirts

2010-03-10T13:11:00.004-05:00

Each semester that I teach the Google Online Marketing Challenge course, we get a t-shirt for each student. I find it a great bonding experience, something novel, and adds to the course esprit de corps, as Penn State GOMCHA alumni can attest.

Here is our t-shirt design this year:

The art was designed by Karen Lee and the shirts were printed by Spotted Lizard.

Steve Donlin of GSI Commerce Speaks to Penn State GOMCHA Course

2010-03-03T23:28:00.005-05:00

Steve Donlin of GSI spoke to Penn State students in the Google Online Marketing Challenge course (GOMCHA course). Although a professor can provide the content and structure for the course, nothing beats current advice and insights from ‘the field’.

Steve gave really good presentation, providing some really interesting insights. What really made an impact was Steve relating the size in terms of money, keywords, and ads of these large campaigns. Makes a really big impact relative to what the GOMCHA students are working on. Really insightful for me also!

Headquartered in King-of-Prussia Pennsylvania, GSI has a really interesting history and 4,000 employees as of 2008. Since 2007, GSI has been on a buying spree, including acquiring Pepperjam, a PA company. See a history of GSI Commerce.

Steve works for True Action, which is GSI’s full-service digital agency.

Ian Howells and Francis Shovlin of Pepperjam Speak to Penn State GOMCHA Course

2010-03-02T12:41:00.005-05:00

Ian Howells and Francis Shovlin from Pepperjam spoke to Penn State students in the Google Online Marketing Challenge course (GOMCHA course). The 50-student course is a nice mix of students from the technology, advertising, and business colleges at the university.

Ian and Francis provided some wonderful insights for the students concerning GOMCHA. Some highlights that stuck with me were:
- best to geo-target by metro area
- when day parting, decrease bids when store is not open
- the importance of A/B testing (the biggest impact is with headline changes)
- collect one week of data before doing any changes.

Kris Jones, the President of Pepperjam, was also an advisory board member for an economic develop research project that I did using keyword advertising to assist SMBs in central PA.

I visited the Pepperjam and give a research presentation on search engine branding

I also reveiwed Kris’s book, Search Engine Optimization: Your Visual Blueprint for Effective Internet Marketing. You can read my review of Search Engine Optimization.

You can follow Pepperjam on Twitter @pjamnetwork (note: focuses on their affiliate network).

Hey, Google is an Advertising Company

2010-02-26T10:38:00.004-05:00

Wow, you can read the numbers from the Google financial reports, but seeing it as a graph really drives it home. Google is an advertising company!, and a darn good one!

More details, at CHART OF THE DAY: In Case You Had Any Doubts About Where Google's Revenue Comes From

Tayna Ferrell from IMPAQT Speaks to Students in Penn State GOMCHA Course

2010-02-26T00:13:00.005-05:00

Tayna Ferrell from IMPAQT spoke to students in my Google Online Marketing Challenge course (GOMCHA course), where is gave the students some really wonderful advice on structuring keyword advertising campaigns and answered some student questions concerning GOMCHA.

It was really one of the most informative presentations on keyword advertising that I’ve ever heard.

Tanya also presented insights into the role of a search engine marketer, as shown in this slide.

IMPAQT has been a great sponsor of the Penn State GOMCHA course, providing funds for the IMPAQT Innovation Awards, which is our local competition.

Here is a really good student blog posting about our Penn State GOMCHA celebration ceremony, where IMPAQT representatives present the IMPAQT Innovation Awards.

The CEO of IMPAQT was also an advisory board member for an economic develop research project that I did using keyword advertising.

I previously visited IMPAQT and give a research presentation on search engine branding.

Mandy Mazzeo of Acronym Media speaks to Penn Sate GOMCHA Students

2010-02-24T12:14:00.006-05:00

Mandy Mazzeo from Acronym Media spoke to students in my Google Online Marketing Challenge course (GOMCHA course), providing some great insights into her experiences with GOMCHA, as well as her working experiences.

Mandy was in my GOMCHA course last year. She was one of those students who had a little attitude but in a good way (if you’ve been teaching for awhile, you know what I'm talking about). So, she added some unique perspectives to the course and, it became apparent that she was insightful about online marketing.

For example, Mandy formed her team before the course started, so her team entered with a team structure already in place. They did some really innovative things when preparing their marketing strategy, including conducting a customer survey to see what ad factors would appeal most to their target demographic (i.e., their client was hungrybuffs.com, which has a really challenging Website for SEM'ing).

The team’s insight and hard work paid off, finishing in the top 15 in the world in the 2009 Challenge.

Mandy parlayed her keyword advertising experience, landing a job at Acronym Media in Manhattan working on a multi-million dollar account. Nice job and nice location!

You can follow Mandy on Twitter at @amazzing

Alex Mann of Trendrr Speaks to Penn State Students

2010-02-17T07:15:00.007-05:00

Alex Mann of Trendrr presented to students in a course that I'm teaching, where the student teams have to do a start-up pitch for our university's start-up competition.

Alex was the winner of the 2008 IdeaPitch competition, so he was able to supply some nuts and bolts advice to the students concerning the pitch. Several of the teams immediately implemented some of Alex's suggestions.

Alex’s story is also insightful. Here is a quote from Alex’s blog: I’ve previously started two businesses of my own, and failed gracefully. Failure is always more constructive than success; it’s also less fun.

David Rusenko from Weebly Presents to Penn State IST students

2010-02-14T22:13:00.009-05:00

David Rusenko of Weebly presented to students in a course that I'm teaching, where the student teams have to do a start-up pitch for our university's start-up competition.

I have blogged about Weebly before, Perceptions of Weebly (a Web start-up) for Designing WebPages, Trip to Google Award Ceremony, the Google Online Marketing Challenge, and Genetic Algorithms in Online Advertising?

David gave a fantastic talk, focusing in on the phrase from 'concept of the idea' to 'moving into the first Weebly headquarters'.

It was really insightful, as the concept for Weebly came from a course project that David did. From there, Weebly made Time magazine's 50 Best Websites 2007. That and a lot of hard work, contributed to Weebly’s success.

It was really interesting, for me, to see a 20 some year old dude, talking about angel funding, venture capital, risk mitigation, book keeping, and cooperative agreements. And, David really connected with the students.

David Rusenko is one of the founders of Weebly, along with Chris Fanini and Dan Veltr.

Jason Calacanis's presentation to Penn State students

2010-02-12T00:29:00.008-05:00

Jason Calacanis gave a great presentation! to students in a course that I'm teaching, where the student teams have to do a start-up pitch for our university's start-up competition.

You can see the whole of Jason Calacanis's presentation, but one aspect that really stuck with me, was when Jason talked about his father. Actually, two fathers. One that ran a start-up, and one that worked for the 'man'. Although the same physical person, two totally different persons.

Really insightful on the impact that one's context has on an individual.

Another thing that stuck with me was that you just need to be 6 months to a year ahead of everyone else.

Jason talked and answered questions for about 40 minutes, and the students (and I!) loved it! Entertaining and interesting, and worth the watch!

Real Time Search User Behavior

2010-02-11T12:19:00.010-05:00

Real time search is an increasingly important area of information seeking on the Web.

In results reported in this co-authored research paper, Real Time Search User Behavior, we analyze 1,005,296 user interactions with a real time search engine over a 190 day period. Co-authors are Gerry Campbell and Matthew Gregg, both of Collecta.

We investigate aggregate usage of the search engine, such as number of users, queries, and terms. We also investigate the structure of queries and terms submitted by these users.

Results are:

Real time search has similarities to traditional Web searching
that 60% of the traffic comes from the engine’s application program interface, indicating that real time search is heavily leveraged by other applications.
Of the queries, 30% were unique (used only once in the entire dataset)
The most frequent query accounted for 0.003% of the query set
Less than 8% of the terms were unique
The most frequently used terms accounted for only 0.03% of the total terms
Concerning search topics, the most used terms dealt with technology, entertainment, and politics, reflecting both the temporal nature of the queries and, perhaps, an early adopter user-based
Sexual queries were quite low, relative to traditional Web search
Searchers of real time content often repeat queries overtime, perhaps indicating long term interest in a topic

As real time content increasingly enters the main stream, an understanding of user searching interactions will be increasingly important.

Classifying Web Queries by Topic and User Intent

2010-02-09T22:13:00.005-05:00

I’ve been researching methods to automatically classify Web queries, actually enrich them, in order to improve searching effectiveness. Here is a research paper, Classifying Web Queries by Topic and User Intent, that explores classifying queries by topic and user intent. My co-author on the research was Danielle Booth.

We took a 20,000 plus Web query data set sectioned by topic, and we manually classified each query using a three-level hierarchy of user intent (i.e., informational, navigational, and transactional).

We noted significant differences in user intent across topics. Results show that user intent (informational, navigational, and transactional) varies by topic (15 to 24 percent depending on the category).

Business queries are clearly mostly navigational (51.9%), as are Holidays queries at 50.8%, and Organizational at 72.1%.
For transactional topics, pornography had far and way the highest percentage of transactional queries (62.3%). Other topics with high percentages of transactional queries were Computing (27.7%), Games (24.8), and Shopping (35.0%).
The general classifications scheme of informational, navigational, and transactional queries vary across topics.

We then used this manually classified data set to classify searches in a Web search engine query stream automatically, using an exact match followed by n-gram approach.

An approach such as this has the advantage of being implementable in real time for query classification of Web searches.

The implications are that a search engine can improve retrieval performance by more effectively identifying the intent underlying user queries. Starting with some basic log data, one can really enrich a search query.

Gen X and Y’s Attitudes on Using Social Media Platforms for Opinion Sharing

2010-02-06T21:07:00.007-05:00

If you are interested in opinion sharing on social media platforms, here is a co-authored research paper that examines opinion sharing of 34,514 survey respondents (so we have a nice sample) between the ages of 13 – 24 (we've targeted an important demographic), Gen X and Y’s Attitudes on Using Social Media Platforms for Opinion Sharing.

We examine differences among in levels of engagement (i.e., number profiles on social media platforms, willingness to seek opinions via status messages, and willingness to act on opinions received in state messages (so, there are some important implications for businesses).

Our findings show:

there is a difference in attitudes among platforms (Facebook, MySpace, my Yearbook, and Twitter), with Facebook-only users being the most private
there is a positive correlation between profiles on social media platforms and willing to share and receive information.

Co-authors were Kate Sobel, of Penn State, and Geoff Cook, of myYearbook

Mark Cramer of Surf Canyon Guest Speaker for Penn State IST Course

2010-02-02T18:03:00.006-05:00

I teach a course where the students, in teams, develop start-up ideas for IdeaPitch, the Penn State entrepreneurship competition. I really believe that many (if not most) technology programs ignore the business aspects that tech folks must work in and need to understand. This course helps address this short coming.

Bringing a “where the rubber meets the road” perspective, we had a great guest speaker, Mark Cramer, of Surf Canyon. He really connected with the students during his talk and brought a real world perspective to the concept of starting a company.

What was great about Mark’s talk was what he focused on, which was the mental and affective aspects of starting a company. Students are fairly use to hearing “starting a company is a lot of work”, but Mark highlighted the dichotomies that an entrepreneur faces.

Here is copy of Mark’s slide, which captures the essence of his presentation.

On behalf of all the students (55 of them!), a big thanks to Mark for sharing his time and insights!

Top Search Marketing and Social Media Research Projects of 2009

2010-01-21T09:35:00.004-05:00

WordStream released their Top Search Marketing and Social Media Research Projects of 2009.

Certainly pleased that two of my co-authored research papers were selected. WordStream selected research papers “Investigating Customer Click-Through Behavior with Integrated Sponsored and Nonsponsored Results” and "Twitter Power: Tweets as Electronic Word of Mouth" for inclusion on the list. Co-authors include Amanda Spink, Abdur Chowdurhy, Mimi Zhang, and Kate Sobel.

All the papers listed are worth the read, though. Some of the work, I am quite familiar with.

I have been citing the two papers from Anindya Ghose and Sha Yang in research that I am doing on a 7 million record data set of a keyword adverting campaign covering 4 calendar years. The first paper, Gender Demographic Targeting in Sponsored Search, is already available.

I follow the personalization work by Jaime Teevan, which is really interesting. I’ve worked in the personalization area. It is a quite hard topic!

Finally, The BuyerSphere Project by Gord Hotchkiss of Enquiro, is one of the first research investigating the B2B area.

Effect of Gender Demographic Targeting in Keyword Advertising

2010-01-16T05:48:00.005-05:00

Co-authored a paper, Gender Demographic Targeting in Sponsored Search, in which we examine nearly 7,000,000 daily records spanning 33 consecutive months of a search engine marketing campaign from a major US retailer.

We look at all the major key word advertising metrics, specifically impressions, clicks, cost-per-click, sales revenue, orders, and items sold.

We used the Microsoft adCenter Labs Demographics Prediction Tool to classify the key phrase. The distribution of gender queries in the dataset was:

Implications are:

Findings from our analysis show that the gender-orientation of the key phrase is a significant determinant in predicting behaviors and performance, with statistically different consumer behaviors for all attributes as the probability of a male or female keyword phrase changes.

However, gender neutral phrases perform the best overall, calling into question the benefits of demographic targeting.

Specific findings are:

increased gender-oriented key phrases are more expensive and do not generate increase sales revenue

gender neutral key phrases generated nearly twice as many impressions as any other category of key phrases and nearly two and half times as many clicks

female-oriented key phrases generated about 10% to 20%

more impressions and approximately 40% to 127% more clicks than the male oriented key phrases

female-oriented key phrases generated 70% to 100% more mean sales revenue than the analogous

male-oriented key phrases, with nearly the same mean number of orders

our findings do support prior work indicating differences in the way that males and females gather and process information

There is very little research in the sponsored search keyword advertising area, and this research is certainly the largest in terms of data used.

So, Professor, do you want to win the 2010 Google Online Marketing Challenge?

2009-12-06T19:01:00.011-05:00

In January 2010, Google will kick off the 3rd Google Online Marketing Challenge (GOMCHA).

GOMCHA is one of the largest, in-class, worldwide, student competitions around. It combines real world instruction with real clients using a real advertising platform. Additionally, GOMCHA has done a lot to introduce (and prod) academia to begin educational courses in SEM.

I have participated in GOMCHA in 2008 and 2009, being fortunate enough to have some wonderful and gifted students in both challenges. One student team won the America’s region in 2008, and three student teams placed in the top 15 in the world in 2009.

Folks ask me “what is the secret to winning?” … Well, I am going to tell you the secret!

Initially, let me tell you how to lose. (The items listed are all things professors have told me or that I’ve read on listservs, blogs, or Web postings from professors about the Challenge).

I am too busy to actually use AdWords myself.: If this is true, you’re also too busy to teach this course. Save everyone some time and go do something else. You *must* use AdWords yourself to be successful.
It seems like something that the students can teach themselves?: If you think this, you have no clue what you’re doing and don’t understand SEM. There are folks who engage in this space 24/7 and are still learning. Again, save everyone some time and go do something else. You must **teach** the students to be successful.
I just use the Google AdWords Learning Center material.: The AdWords Learning is a nice starting point, and it is good place for the students to learn how to push a button or flip a switch on AdWords. (BTW, this is something that the students can learn themselves!). However, if you just use this material, the students don’t need you. Again, go do something else. You must **leverage** outside resources to understand how to do well in keyword advertising.

Now, the secret three-step process to fielding top-notch teams!

Step one, you need a good foundation in SEM and AdWords. Two good resources to do this are (note, I have *no* financial interest in either):

AdWords for Dummies by the esteemed Howie Jacobson
Full Disclosure: Cracking the Google Online Marketing Challenge by the very talented Daehee Park

Step two, you need to organize your course. Here is a really good article on how to do this (the lead author is the outstanding, Theresa Flaherty, who is also a two-time winner in the GOMCHA challenges):

Flaherty, T. B., Jansen, B. J., Hofacker, C., Murphy, J. (2009). Insights on the Google Online Marketing Challenge and Its Successful Classroom Implementation. Journal of Online Learning and Teaching, 5(2), 446-457.

Step three, you need some worthwhile content for each lesson. To give you a jump start, I have posted all my slides from the 2009 Challenge to Slideshare. The topic and presentations (in the order that I presented them) are:

And, that's the secret … (domain knowledge, course organization, and lesson substance), … which are the three essential ingredients for any successful course. Not a big secret at all.

Good luck in GOMCHA 2010!

Searching for Salvation! and Publishing Multidisciplinary Research?

2009-12-01T17:08:00.009-05:00

This posting has two themes, (1) reporting research findings and (2) a discussion of the perils of publishing multidisciplinary research. However, you will see that they are related.

The Research Results
I’ve been trying to push the boundaries of my academic research, and one product is a study on how people search for religious information on the Web, with results reported in this co-authored paper, Searching for salvation: An analysis of US religious searching on the World Wide Web published in the journal, Religion.

It is a longitudinal study using several search logs over about a decade, employing transaction log analysis of about five million queries.

The major research findings are:

The religious topics being searched are not cult but are typically mainstream. (Why important? There is this view that the Web is used for a lot of non-traditional religious searching and cults.)
The Web attributes of religious searching (i.e., query length, session length, etc.) mirrors that of searching for many other topics. So, people searching for religious information are just as capable as their non-religious brethren. (Why important? There is this view that religious folks are less technically suave than non-religious folk.)
The major search engines do not perform well for religious queries, which is why many religious information sites are exchanged via blogs and listservs. (Why important? Since many religious sites are localized geographically in nature, they face the same issues that many small-to-medium sized businesses have in getting ranked highly by the search engines.)

Now the Multidisciplinary Publishing
The biggest issue in the research effort was getting darn thing published! It was a multidisciplinary research piece, using data and methods unfamiliar to religious researchers, although drawing on religious studies theories and models, such as secularization theory.

A big shout out to Michael Stausberg, co-editor of the journal, Religion, who took a chance on the article. However, the article went through five (count them, 5!) rounds of review, with a new reviewer added each round, before the article was finally accepted. And, acceptance was still contingent on changes in the final version.

We had tried three other outlets previously, with some interesting outcomes that, I believe, highlight the issues when trying to publish multidisciplinary research.

The first journal that we shipped it to (a journal that scientifically studies religion) charged us $50 and took 8 months for the first review. However, we got a positive R&R. So, we revised the manuscript and resubmitted it. It was under review for another 6 months, after which the editor rejected the paper without comment. The editor then ignored our emails and telephone calls for the next three months.
The second journal (a journal that studies American sociology) to which we submitted it also charged us $50. Shortly after submission, we got a letter back saying that we needed to use more rigorous methods to publish in their outlet. They did return our check though.
The third journal (a journal that publishes reviews of religious research) made us pay $30 to join their academic organization. They reviewed the manuscript, and then rejected it as not focusing on religious studies. They didn’t refund our money.

What did I learn from this? ... you can do multidisciplinary research, but it is not worth the time and effort to publish it that way. Publish in your own domain.

Visit to SEM / SEO firm, Bloom Marketing

2009-11-17T16:39:00.003-05:00

When I was in Vancouver attending the ASIST 09 conference, I had opportunity to visit Bloom Marketing, specifically chatting with Carlos Obregon. We had a nice cup of java in the coffee shop right near his office, chatting for a couple of hours.

I have visited several large SEM and SEO firms, which always provides some good insights into the state of affairs in the industry. However, engaging with Carlos was interesting for me, as he is involved in the entire stream of business processes for his firm, from dealing with clients, implementing ad campaigns, Web analytics, interfacing with agencies, and all the headaches and joys that come from running a SMB. Just a great conversation!

Really got some great snippets of information concerning the most workable SEM tools, some of which I am going to try to incorporate into my Google Online Marketing Challenge Course.

In addition, we had a fascinating discussion of the positive effect of having results in both the organic and sponsored listings on the SERP page.

SEM firms have long stated that this has a positive effect (i.e., an advertiser gets more combined clicks than if they appeared in either the organic or sponsored listings separately). Honestly, I’ve always found the idea a little questionable, as I couldn’t see a reason for it. However, Carlos and I discussed the concept of the blockade.

An advertising blockade is a traditional advertising strategy, where a firm buys up all the advertising space in a medium at a given point in time. It’s most successful implementation was with the advertising campaign for the Ford Mustang, but President Obama did a modern take on it during the 2008 Presidential Campaign. For some reason, this was the first time that I really made that, now obvious, connection.

An ad blockade is really what is happening when the same content provider appears on both organic and sponsored listings, given searchers’ habit of adhering to the results ranking on the SERP.

In addition to these insights, got some copies of case studies that Bloom Marketing had done, showing return on investments for PPC campaigns strategies. Going to rework them slightly and use in my Google Online Marketing Challenge Course.

Searcher Profiling and Prediction of Future Actions on Web Search Engines

2009-11-11T17:25:00.007-05:00

Attended the ASIST 2009 Conference (basically an academic affair), where I gave two presentations on current research projects. One presentation was on a project investigating how rich of a profile we could develop on a person based on data typically stored in search logs. Answer … we can determine quite a lot!

The other presentation was results investigating the use of data in logs to predict future behavior using n-grams. This is a state-based approach we used to predict future query reformulation. However, it a component of other research using time series analysis and neural networks investigating predicting future click through (e.g., Time Series Analysis of a Web Search Engine Transaction Log and Identification of factors predicting clickthrough in Web searching using neural network analysis).

The papers addressing the conference presentations are To What Degree Can Log Data Profile a Web Searcher? and Patterns of Query Modification during Web Searching.

Presentations from ASIST09 are here: User Profiling - http://tinyurl.com/yb8zbgj and Query Reformulation - http://tinyurl.com/yb9atxo

Visit to Enquiro Search Solutions, Inc.

2009-11-07T18:27:00.002-05:00

Spent a day at Enquiro, which is a search engine marketing and optimization firm based in the lovely town of Kelowna, British Colombia. ... Being serious, the place was beautiful. ... Gord Hotchkiss is the president and CEO, Enquiro.

Enquiro is an interesting company, in that not only do they do SEM and SEO stuff, but they also have a research division, and have produced many Enquiro research papers in the search engine marking area, including the classic, Inside the Mind of the Searcher, which is still a standard in the area. Lately, Enquiro has been focusing on the B2B area, with Gord’s upcoming book, The BuyerShere Project, which is an outgrowth of The BuyerShere Project series.

Although long day (I left the hotel in Vancouver at 6 am and got back at 10 pm), but it was a lot of fun with interesting exchanges and conversations. I gave a talk on Profiling the Web Searcher Using Log Data, along with brief introductions to a *couple* of other research projects that I am invovled with.

Always interesting for academics to present to practitioners, because (in Web search at least), many times the practitioners are ahead of those in academia. (Note: I am talking real practitioners, not the researchers from industry who typically come to academic conferences. In my experience, these industry researchers are just academics, even though academia treats them like practitioners.) The Enquiro folks are engaged in this SEM/SEO/B2B space! Therefore, the exchanges and discussions were educational for me. I picked up three insights that will really help me focus some of my research projects.

I also participated in the Enquiro foosball tournament. Although not as polished as I once was, I held my own.

A big thanks to the entire Enquiro team!

The “Sex” Of Search Queries & Personalization

2009-11-01T19:14:00.002-05:00

I did an extensive interview with Search Engine Land, The “Sex” Of Search Queries & Personalization, in which I discussed research where we took queries from a search engine marketing campaign and classified them based on gender probability using Microsoft’s demographic tool, which will classify a query by it’s probability of being male or female. We did not attempt to classify the searcher as male or female. Rather, we took it as whether or not the particular query fit a gender stereotype - did the query have a kind of a male, for example, feel to it or stereotype implications?

This research (combined with other findings) is really making me question the role of personalization. Maybe a task-focused personalization approach is better than focusing on a person.

In this portion of the interview, I highlight the gender classification issue, personalization, human complexity, task personalization, local search, mobile devices, mobile, search, and push-pull information.

The first part of the interview was published in The Truths & Myths Of The Search Buying Funnel.

The full interview transcript is here, Talking Search with Jim Jansen at Penn State.

Why Am I Getting Better Conversations in the Blogosphere than I Do at Academic Conferences?

2009-10-28T23:11:00.003-04:00

Check out this commentary on one of my research papers, Brands, and putting twitter word of mouth in context by Adrian Chan. When I read parts of this posting, I thought “Wow, I wish that I had wrote that in the original paper!”

My next thought was, “Why don’t I get exchanges like this at academic conferences?”

Honestly, I don’t go to a lot of academic conferences. I believe the cost (in time) greatly exceeds the benefits (can’t really think of many). And, as I state in this posting about conferences, Academic Conferences as Research Outlets, they favor a certain type of academic (no kids, good travel budget, teaching grad students). However, the main problem is that they just aren’t really meaningful conduits for research discussions.

However, if I had meaningful academic exchanges, like in the Brands, and putting twitter word of mouth in context blog posting, I would consider going to a lot more.

Companion Website for Understanding Web – User Interactions Via Web Analytics

2009-10-28T14:03:00.003-04:00

In August 2009, I published a lecture (i.e., short book), Understanding User-Web Interactions via Web Analytics, as part of the Synthesis Lectures on Information Concepts, Retrieval, and Services. In the lecture, I review and provide an overview of many components of Web Analytics, including foundational elements and implementation processes.

Naturally, from the moment of publication, there is additional information that I wish that I had included or topics that I wanted to addressed. Therefore, I have started a supplemental Website as a companion for the lecture, which is already populated with additional information, Supplementary material for Understanding User - Web Interactions Via Web Analytics.

Here is a list of universities that access to the lecture, which is the based for the organization and the material on the Websites.

Pictures from Undergraduate Research Day At Harrisburg

2009-10-21T22:19:00.000-04:00

Undergrad Researchers Present at State Capital

2009-10-19T08:59:00.003-04:00

Four Penn State undergraduate researchers get the opportunity to present their research at the Pennsylvania State Capital, along with other undergraduates from across the state. The four researchers are Arielle Amchin (Smeal College of Business and College of Communications), Pete Smith, Mitchell Rukat, and Paul Rinaldi (College of Information Sciences and Technology).

The research project, Search Engine Marketing for Economic and Workforce Deployment for SMBs in Pennsylvania, is an 8-month long economic and workforce deployment effort to leverage keyword marketing involving four small to medium size businesses (SMBs) in Pennsylvania. The project uses search engine keyword advertising to expand the potential customer base for these businesses and to increase their ability to compete in the online marketplace. The students leading this research project leverage knowledge gained in a keyword marketing course, tying the classroom to the real world. In addition to development and implementation, there is a training component to transition the advertising campaign from the students to the SMB at the conclusion of the project. The educational component helps ensure long term viability of the program. This research venture is funded by a $25,000 Penn State Outreach Grant, and three of the SMBs are members of the Penn State Technical Assistance Program, tying the university to the local community.

The four researchers for this project are all graduates of a keyword advertising course, in which they participated in the Google Online Marketing Challenge.

Follow-up Comments on Twitter and Micro-communications

2009-10-06T23:24:00.007-04:00

In an academic research paper, Twitter and electronic word of mouth, my co-authors and I present findings concerning electronic word of mouth branding and marketing using Twitter as a micro-communication platform.

Since then, I have had the pleasure of reading several commentaries and 'analysis of the analysis'. What is great is that there have been some really super questions, comments, and insights that I had not considered or did not elaborate on in the paper.

Here are three posting that I found really interesting.

First, there is this one from Social Media Today, 20% of tweets about brands. What’s cool, is that the author brings up that he/she made 4 comments about brands just since the weekend. And, if 20% of tweets are about brands, that translates to about 600K tweets. Also, there is this neat sign about what is Twitter good for.

Second, there is this posting from We are Social, 19% of tweets mention brands. The author comments on the aspect of listening and responding in micro-communication, and the flow of conversation.

Finally, there is this one from emarketer, Information Ahead of Opinion on Twitter, where they tie in research about customer service from companies.

As all the articles noted, a lot of the tweets about brands were positive. A lot of blogs and review site postings are typically negative. As a posted on Natalia Hatalska’s blog posting on the Twitter study asked could some of these be from companies? Yes, they could.

But, I believe that the ease of tweeting makes it easier for positive sentiment to be expressed. With blogs and review sites, one really has to be fired up to comment, so these posts are typically folks that are upset.

Maybe Twitter will raise our level of online conversation?

The Truths & Myths Of The Search Buying Funnel

2009-10-02T13:38:00.003-04:00

Here is a posting from Search Engine Land’s Just Behave column, entitled The Truths & Myths Of The Search Buying Funnel, which is an interview by (actually a discussion with) Gord Hotchkiss.

Although not a research interview per say, Gord and I touch on a lot of heavy duty research concepts, including, of course, the buying funnel, but also the database of intentions, the AIDA model, satisficing, and the principle of least effort, just to name a few.

We also get into the issue of searching as task and searching as entertainment, and the different searching behaviors that result.

A conversation with Gord is always good and informative, with good questions and plenty of push back. Working with members of his team from Enquiro, he has lead several interesting and groundbreaking research studies involving eye tracking and Web searching.

For Penn State Students: Keyword Advertising Couse Spring 10

2009-09-13T08:10:00.003-04:00

I am offering an online advertising/marketing/technology course, IST402 Google Online Marketing Challenge, in Spring 2010

Course description: IST 402 - Google Online Marketing Challenge

Indispensable for locating information, Web search engines also facilitate e-commerce transactions, with many organizations using search engines for online marketing, branding, and advertising. This course examines the online marketing environment via a real world competition using the Google AdWords platform. Students will participate in the Google Online Marketing Challenge (http://www.google.com/onlinechallenge/). Students will gain practical, real world online marketing experience while getting all the excitement of competing with teams of other students from all over the world.

The times for the course: T / R 4:15 - 5:30. The course will be capped at 48 students (firm cap).

Students interested in registering need to send me their name, resume, and Penn State id number (need id to get them registered, controls are on the course).

Here is a press release about the course last semester http://live.psu.edu/story/40781 and a couple of student blog postings, http://www.dennissiller.com/blog/2009/4/29/google-online-marketing-challenge-2009.html and http://onwardstate.com/2009/06/03/class-watch-google-marketing-ist-402/

Non-profits Project for Keyword Advertising Course

2009-09-02T11:02:00.004-04:00

Thanks to a Google grants program, that provides $30,000 in online advertising support for non-profits, 55 technology and advertising students at Penn State are getting the opportunity to design and implement keyword advertising campaigns for eleven non-profit organizations.

The students, working in teams of 4 or 5, will implement search engine marketing campaigns for these non-profit organizations, along with doing some search engine optimization work on the organizational Websites, especially the landing pages.

Really excited about this opportunity, as the student projects we typically do are business or government-related.

Doing work in the non-profit sector can really be an eye opener for students as they see the unique challenges that these organizations face in the resource, marketing, and technology areas.

The 11 non-profits working with the students are:

Organization (URL)
Arthritis Foundation, Western Pennsylvania Chapter (http://www.arthritis.org/chapters/western-pennsylvania/)
Centre Volunteers in Medicine (http://www.cvim.net/)
CentrePeace (http://www.centrepeace.org)
ClearWater Conservancy (http://clearwaterconservancy.org)
GivingPoint (http://www.mygivingpoint.org)
Philadelphia Animal Welfare Society (PAWS) (http://www.phillypaws.org/)
Planned Parenthood Southeastern Pennsylvania (http://www.plannedparenthood.org/ppsp/)
Southeastern Pennsylvania Chapter of the American Red Cross (http://www.redcross-philly.org/)
THON, Penn State IFC/Panhellenic Dance Marathon (http://thon.org/)
West Philadelphia Alliance For Children (http://www.wepac.org/)
World Affairs Council of Philadelphia (http://www.wacphila.org)

It's also good for me as an educator, because I feel like the course and material that I teach can actually have some immediate and measureable impact on the local community. In addition, the Google Grant funding continues after the course, so this really is a real-world project.

Feeling good about it! Good group of students also.

Cognitively Modeling Web Searching as Learning

2009-08-30T07:21:00.004-04:00

One approach to improve the performance of information retrieval systems is to understand better the underlying need of a searcher. What is motivating the search? This can take many forms, including identifying topic or content needs.

Another approach is to model the search process itself. What type of cognitive process is the searcher engaged in?

Many researchers have explored decision making or problem solving to model the search process. However, from my research, these models did not stand up to empirical testing. There was no correlation between problem solving ability and searching performance or behavior. This caused me to look for other approaches, leading me to a learning model.

The results of the research using a version of Bloom’s Taxonomy to model online searching is presented in the co-authored paper (with Danielle Booth and Brian Smith), Using the taxonomy of cognitive learning to model online searching.

One really interesting inference from the research findings is that, based on the searching behaviors, we could identify the cognitive level of learning in the model. Therefore, there are implementable implications from the research.

Using a laboratory approach, we studied the searching characteristics of 72 participants engaged in 426 searching tasks. We classified the searching tasks according to Anderson and Krathwohl’s taxonomy of the cognitive learning domain.

Findings shows that applying and analyzing, the middle two of the six categories, generally take the most searching effort in terms of queries per session, topics searched per session, and total time searching.

The lowest two learning categories, remembering and understanding, exhibit searching characteristics similar to the highest order learning categories of evaluating and creating.

The results indicate that the view of Web searchers having simple information needs may be incorrect. Instead, searchers applied simple searching expressions and tactics in support of higher-level information needs.

Overall, these research results indicate that a learning theory may better describe the information searching process than more commonly used paradigms of decision making or problem solving.

Branding and its effect on user perception of search engine performance

2009-08-27T19:11:00.002-04:00

In a short conference presentation, some co-author (Mimi Zhang) and I presented some research on the effect of branding on user evaluation of search engine results.

I call it branding, but one could also view it as affective aspects that effect the evaluation of search results.

Conference papers are nice for quick and incremental reports of research findings, but they really don’t provide the space (and time) for in-depth and detailed analysis.

The journal version of our branding research has come out, Brand and its effect on user perception of search engine performance, and it really provides an analysis of search engine branding from a lot of perspectives, with co-author, Carsten Schultz.

In addition to the empirical findings (e.g., brand gives 10% lift, lack of brand has a 10% hit, effect variation of brand across domains, etc.), we present a 4 stage model of branding from selection of search engine to the evaluation of the landing page.

The power of Twitter, the effect of Tweets

2009-08-25T14:57:00.009-04:00

Here is some interesting academic research concerning electronic word of mouth branding and marketing, reported in the research paper on Twitter and electronic word of mouth

In the research paper, we report the research results investigating micro-communication as a form of electronic word-of-mouth (eWOM) for sharing consumer opinions concerning brands.

We analyzed more than 150,000 microblog postings containing branding comments, sentiments, and opinions. We investigated the overall structure of these tweets, the types of expressions, and the movement in positive or negative sentiment. We also analyzed the range, frequency, timing, and content of tweets from a corporate account.

The research findings show that 19% of tweets contain a mention of a brand. Of the branding tweets, nearly 20% contained some expression of brand sentiments.

Of these, more than 50% were positive and 33% were critical of the company or product.

In analyzing the tweets for structure and composition, the linguistic structure of tweets approximate the linguistic patterns of natural language expressions.

Overall, the practical research implications are that micro-communication is an effective online tool for customer word of mouth communications and is – or soon will be – an essential part of an integrated marketing strategy.

You can read the press release on the eWOM Twitter paper.

The Latest Search Engine Research that Every Marketer Needs to Know

2009-08-11T19:43:00.002-04:00

Hey!

I did a radio show with RSS Ray, The Latest Search Engine Research that Every Marketer Needs to Know. The podcast is about 30 minutes, available on iTunes also.

However, here are four questions that I was asked and some more coherent responses.

1. How has the way we search the Internet changed in the past few years?

Looking at the general purpose search engines, like Google, Bing, and Ask! From a technology point, there hasn’t been much change. The last major technology innovations to Web Search was Page Rank and sponsored search. The interfaces are nearly identical to those of 10 years ago. Google in 1999 would feel very similar to Google 2009. The crawling algorithms are the pretty much the same. Now, ranking has improved with some innovative uses of anchor text and other factors. Topic identification and query suggestions has improved. And, there is some interesting work going on in video and multimedia. Mobile search may eventually take off. But, generally, from a technical stand point, there hasn’t been much innovation but a lot of improvement in effectiveness and efficiency.

How searchers interact with search engines has also held pretty constant also. Users enter typically submit short queries, a two or three terms, and sessions are short, two or three queries. The trend has slowly been increasing on both fronts, but generally the same.

What has changed is the context, the scale, and the variety.

In terms of context, the direct availability of content accessible on the Web is nearly ubiquitous. Web search engines provide access to textual and multimedia content in a wide variety of settings including both home and work, as well as in mobile situations.

Second, there is the number of searchers attempting to access this content via Web search engines. The scale of topics submitted by these users is surely unparalleled in pre-Web end user searching.

Third, the variety of content, users, and systems is certainly unique. This combined diversity on the Web in both content and users is extreme.

In addition to satisfying information problems, modern Web search engines are navigational tools to take users to specific uniform resource locators (URLs) or to aid in browsing.

People use search engines as applications to conduct ecommerce transactions, such as with sponsored search or Google’s payment system. The increase in main line ecommerce use of the Web is certainly a major change, not only from the user side but with professional search engine marketing firms, search engine optimization firms, and the affiliate networks. These have really changed the game in how commercial content is found, accessed, and presented to Web users.

Search engines provide access to content collections of images, songs, and videos rather than directly addressing an information need with a specific object.

Search engines provide access to transactional services such as maps, online auctions, driving directions, or even other search engines.

Search engines perform social networking functions, as with Yahoo! Answers. Web search engines are spell checkers, thesauruses, and dictionaries.

They are games, such as Google Whacking or vanity searching.

Modern Web search engines are adding an increasing diverse range of features. Providers are placing more and highly varied content and services on the Web. In response, people are employing search engines in new, novel, and increasing diverse ways.

2. What are the big search engines currently doing to improve the search experience?

Well, in the US, when you talk ‘big’, you are only taking two now, Google (65% share) and Bing (with about a 30% share). Maybe three if we include Ask, which is typically around 4 %. All the rest split the remaining 1 %. … and, honestly, spoken persons from these companies could probably answer it better.

From my analysis though, some of the most exciting work is the behavioral targeting. By behavioral targeting, I mean the collection of techniques that leverage past and current user experience to enhance future (both near and far) searching experience. There is a lot of beneficial work going online advertising targeting, query reformulation, topic identification, and user behavior prediction.

I have been doing work on predicting click through using n-grams, neural network and time series analysis. We are trying now to develop a formula for an individual user given a search history. We are see some success. I would expect the search engine companies are pursuing similar studies. They have better data than I have and more of it. So, we should see some good progress from them.

Another area that the search engine companies are doing some good work, is in the area of spam prevention and detection, both on the organic side and on the sponsored side. This stuff is not exciting but it is something that probably has the most direct and beneficial lift to the

3. Google has been seeing some major competition from Bing recently, do you think Google has anything to be worried about?

Well, major competition might be a stretch. Bing is floating around 8%, up a little from pre-Bing days.

Some of the early reports on numbers are coming into okay, the slight uptick in traffic, the reports of the Bing converts on the sponsored search side are coming in positive, the advertising budget has been reported at $100M, so that has to have some impact. The Yahoo! deal bought them an additional 20% marketshare. My daughter likes their font and the cool pictures they have as background. So, some okay number and reactions.

However, my daughter, although she likes the interface, doesn’t use it. She uses Google. I have tried it a couple of times, just to play again. It’s okay. Probably in the ballpark with Google. But, it doesn’t seem like a game changer to me, and I’m a person that kept using Yahoo! , along with Google, right up to the end.

The Google brand is just too strong. The habit is just too engrained. Bing or some other competitor is going to have to bring a totally disruptive technology to the Web search that users just can’t ignore, 10% better results at lease. This is going to be hard to do, ‘cause Google has a great culture of technology innovation. Alternatively, Google could do something to cause searchers to loss trust and confidence.

I’m not thinking that the Bing advertising campaign, the first decision engine is work either. A lot of search is not a decision process. When I’ve put search as problem solving and decision making to the test, it doesn’t correlate. It appears to me that searching is better viewed as a learning process. I believe this fits with Bing name anyway. Bing! Search, learn!

The bottom line though, should Google be worried about Bing? Yes! The landscape is littered with companies that did not take a threat from Microsoft seriously.

4. How do you think personalized search results will change the way business optimize their websites?

Yes, it will. Personalized search reorders search results based on your history of past searches, giving more weight to topics that interest you. Personalized search results has been around for awhile Google experimented with these at least 5 years. However, with Google accounts and desktop search bars more common, the past data is there now.

It will make SEO efforts more difficult for sure and getting brand awareness will be a bigger challenge.

However, people change. Current habits are not locked in concrete.

It actually may a be bonus in that unlikely customers will be coming to their websites. This will allow attentive and efforts to be focused on those visits that are truly potential customers.

5. Given current trends, what is the best way for search marketers to improve their overall performance?

Nothing has changed here from basis advertising. Get the right message in front of the right person at the right time.

Academic – Industry – Community Keyword Advertising Project

2009-08-04T16:46:00.002-04:00

Several Penn State students earned top honors in the 2009 Google Online Marketing Challenge. Of the 10 Penn State teams, 3 were in the top 15 world-wide (out of 2,107) and the other 7 were in the top 50 in the Americas region (about 900 teams).

What is really great though was the industry – academic – community partnership. I’m always looking for ways to make my research and teaching more immediately relevant to my local community. The Google Online Marketing Challenge integrates subjects that I teach, the students who take those courses and my research programs in a way that directly benefits local business through economic and workforce development in Pennsylvania.

The Challenge also gave me the opportunity to work with university level programs. Several of the student/client partnerships were made possible by collaboration with the Pennsylvania Technical Assistance Program (PennTAP), a unit of Penn State Outreach.

Penn State Outreach provided a keyword advertising seed grant, which I am using to bring keyword advertising expertise to assist small businesses in central Pennsylvania.

Understanding User-Web Interactions via Web Analytics

2009-07-30T13:01:00.001-04:00

For those interested in Web analytics, including search log analysis, this lecture (i.e., short booklet of about 100 pages) might be useful, Understanding User-Web Interactions via Web Analytics.

In the lecture, I present an overview of the Web analytics process, beginning with theoretical and methodological foundations to highlight the strengths and weaknesses of the approach.

I touch on the academic side of things from transaction log analysis work in library and information science, but the main focus of the lecture is on the practitioner side things. Also address supplementary data collection techniques including surveys and laboratory studies.

The aim of the lecture to provide the background and methodology for understanding Web analytics in order to improve Web systems, increase customer satisfaction, and target revenue through effective analysis of user–Website interactions.

The lecture is part of the Synthesis Lectures on Information Concepts, Retrieval, and Services series edited by Gary Marchionini.

2009 Google Online Marketing Challenge Winners

2009-07-24T09:46:00.001-04:00

The finalists and winners of the 2009 Google Online Marketing Challenge were announced! A big congrats to the winners!

The Global Winner comes from Deakin University, Australia. The APAC Winner comes from the International College of Management Sydney (Australia). The AMERICAS Winner comes from James Madison University (US). The EMEA Winner comes from the Warsaw School of Economics (Poland).

Closer to home, there were three Penn State teams in the Final 15, see Semi-Finalists in the 'Final 15' judging round.

The Google Online Marketing Challenge is a really excellent keyword advertising challenge and great as a course project. If interested, here is an article that can help one prepare for the Challenge.

The 2010 Google Online Marketing Challenge intent is open.

For Penn State Students: Keyword Advertising Couse Fall 09

2009-07-13T14:07:00.004-04:00

Hey, looking for a worthwhile course in Fall 09? Want to take a course where you work in a multi-disciplinary team managing a $30,000 advertising online campaign for a non-profit organization? Plus, do some Website design and search engine optimization? Then, enroll in IST440W.

Although an IST course, the main project will be an online advertising campaign using Google AdWords, a keyword advertising platform. Google is providing $30,000 per each non-profit for the semester. So, this will be some real world experience with real money for a real client.

The times for the course: M W 06:30P - 07:45P. Location is the IST building. The course will be capped at 40 students, primarily IST and advertising students.

Ignore the course description in the category and the prerequisites. Will waive them for the right students.

New Course Description: Indispensable for locating information, Web search engines also facilitate e-commerce and other transactions, with many organizations using search engines for online marketing, branding, and advertising. This course examines the online marketing environment via a real world client using the Google AdWords platform. Students will participate in the Google Grants for Non-profits Program, which provides $30,000 for a three month online advertising campaign for a non-profit organization. Students will gain practical, real world online marketing experience while getting all the excitement of working in the unique marketing challenge of the non-profit domain.

Students (seniors or juniors only) interested in registering need to send me (Jim Jansen, jjansen@ist.psu.edu) their name, resume, and Penn State id number (need id to get them registered, controls are on the course).

Trademark Infringement in Sponsored Search

2009-07-08T19:32:00.009-04:00

One of the blogs that I read frequently is by Eric Goldman. Rather than a Google-hater or Google-lover, he always has a very reasoned opinion based on the facts. Must be his lawyer nature! (note: don’t expect any sound bites; you’re going to do some reading!)

Anyway, this post on lawsuits against Google AdWords caught my attention. There are at least 7 lawsuits pending against AdWords for trademark infringement. The deal is that some companies bid on the brand name of other companies in sponsored search. Depending on your viewpoint this is call effective advertising, just advertisings, or trademark infringement (a.k.a., piggybacking, conquest buys, and adversarial advertising).

Okay, I am not totally up on the law concerning trademarks, although the use of a trademark term for shopping comparison seems reasonable to me. However, I *am* up on the user searching aspect of keyword advertising and Web searching.

The foundation of these suits seems to be that when a company bids on brand name of a competitor (or another company), when the ad appears it “causes confusing and is misleading” to the searcher. I find this a stretch, and generally I don’t see this position as valid as all.

This is why. First, many search engine now offer query suggestions. Type in a company name and often, you will get a competing company as a suggested query. Why? Because this is the way people search? They compare.

Second, the number of clicks on sponsored links is much smaller than the clicks on organic links (generally in the range of 15% - 85%). Given the branding relationship between organic and sponsored links, the number of ‘confused’ searchers is, by my guess, very small. However, I know of no research that has investigated this ‘confusion’.

Finally, there is the issue of terms within the entire query. For example, consider hotels. I usually, but not always, stay at a Marriott. What if I type in the query “Marriott Fort Lauderdale”? Would it be unreasonable for Websites concerning hotels in Fort Lauderdale to appear in the organic results? Would be reasonable for Websites concerning hotels in Fort Lauderdale to appear in the sponsored results? I believe so. The underlying need is a hotel in Fort Lauderdale.

I'm just not buying the consumer confusion argument for these cases.

Interview with Penn State Professor and 'Search Behaviorist' Jim Jansen

2009-07-06T10:12:00.004-04:00

This was a fun interview with WordStream.

Got to talk about online advertising challenges, sponsored search, search engine brand effect, Twitter research, automated searching assistance, and the Google Online Marketing Challenge.

WordStream is a company offering services and solutions for optimizing and expanding pay-per-click and search engine optimization efforts involving large numbers of keywords.

Organizing a course for the Google Online Marketing Challenge

2009-06-30T14:58:00.006-04:00

Am co-author for an article on planning and implementing a course to compete in the Google Online Marketing Challenge, Insights on the Google Online Marketing Challenge and Its Successful Classroom Implementation.

Really recommend it for professors who are considering (or want to enhance their efforts in) the Challenge.

The article provides some actionable items for implementation, a schedule, and learning resources. One of the most detailed articles on the Challenge that I’ve seen.

My co-authors are Theresa B. Flaherty (who spearheaded the article), Charles Hofacker, and Jamie Murphy, who, with Lee Hunter, got the Challenge going.

The citation for article is: Flaherty, T. B., Jansen, B. J., Hofacker, C., Murphy, J. (2009). Insights on the Google Online Marketing Challenge and Its Successful Classroom Implementation. Journal of Online Learning and Teaching, 5(2), 446-457.

Query Log Analysis: From Research to Best Practice

2009-06-24T09:34:00.003-04:00

I attended the Query Log Analysis: From Research to Best Practice Workshop hosted by Paul Clough from the University of Sheffield. The two day workshop was aimed at stimulating "discussion on the analysis of log files" and providing "a forum in which to share experiences and best practice”.

I’m a big fan of the potential of log analysis, so I found the workshop really enjoyable, interesting, and the start of something beneficial.

Paul did a great job in making the workshop both fun and worthwhile.

I gave two presentations at the workshop, one on log analysis research, Moving from Description to Prediction, that I’ve been doing. The focus was including more predictive aspects in and findings from log analysis.

The second presentation was a primer on Web log analysis, What is Web Log Analysis?, where I really focused on the foundational elements, both theoretical and methodological.

Overall, the workshop was really good.

One of the workshop aims was to bring together researchers and practitioners. In this regard, the workshop was heavy on the ‘research’ and light on the ‘practice’. So, this would be an area to concentrate on in future workshops.

Keyword Advertising Economic and Workforce Development Project

2009-06-22T07:52:00.004-04:00

I’ve been looking to make my research and teaching more immediately relevant to the local community, and this project concerning economic and workforce development in Pennsylvania is one of way of doing this.

Hopefully, we can make a direct impact on SMEs in our local community by leveraging knowledge students gain in the classroom.

We have some great advisors for the project, including IMPAQT, Pepperjam and BlitzLocal, three SEM firms.

A big thanks to the Pennsylvania Technical Assistance Program (PennTAP) for providing the funding.

If at first you don't succeed, let the search engine try

2009-06-07T07:33:00.002-04:00

I’ve been trying to move my research focus in Web searching from descriptive to predictive models. Nothing wrong with descriptive models (we need to understand the world around us) but at some point there is enough descriptive understanding where, in the technology fields, we can develop models that predict or influence.

If at first you don't succeed, let the search engine try is a press release concerning some co-research that I did to develop a predictive model of query reformulation using n-grams. One of my students, Danielle Booth, was a major contributor to this work.

We automatically classified queries within individual sessions in one of 8 states, and then used n-grams to develop Markov chains representing these query reformulation patterns. This search engine log, which was from Dogpile, also recorded when users accessed the query reformulation assistance that most major search engines now offer.

Some finding are:
- the patterns are short (3 – 4 states)
- there is a norm to move to narrow the scope of the query
- there appears to be a cognitive space at the start of the session and when switch vertical where the user is open to system assistance

The research is reported in: Jansen, B.J., Booth, D.L., Spink, A. (2009) Patterns of query reformulation during Web searching. Journal of the American Society for Information Science and Technology. 60(7), 1358-1371.

Thanks to Infospace for access to the logs!