Sponsored Search and Online Marketing

I have been really involved in investigating sponsored search and related forms of online marketing.

Other than sponsored search, there are few technological innovatives in information retrieval that has had such impact. The previous innovation with substantial impact is Page Rank, and that was in the mid-90s (see http://infolab.stanford.edu/pub/papers/google.pdf).

Bill Gross (www.idealab.com) is generally credited with creating the sponsored search paradigm in 1998, launching Goto.com that later became Overture and then Yahoo! Search Marketing. The two largest sponsored search entities, Yahoo! Search Marketing and Google, account for the majority of the sponsored search traffic and with content servicing platforms.

It is difficult to find an area of research with more immediate impact and also long range implications in a variety of areas including commercial and social.

Most of the academic research has been in the online auction area (see for example http://research.yahoo.com/workshops/ssa2005/sched.html).

Some of the search engine marketing firms like Enquiro (http://www.enquiroresearch.com/) have conducted some research in this area.

Dan Fain and Jan Pederson have written a nice history of sponsored search - http://www.asis.org/Bulletin/Dec-05/pedersen.html

I have done some research on sponsored search, including:

The effectiveness of Web search engines at retrieving relevant content, including sponsored links - http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/jansen_ecommerce_ipm2006.pdf

The searcher bias against sponsored links – http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/sponsored_links_jansen.pdf

What causes a searcher to click on a sponsored link – http://ist.psu.edu/faculty_pages/jjansen/academic/jansen_factors_sponsored_link_clicks.pdf

Whether or not sponsored links are as relevant as organic links – http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/jansen_tweb_sponsored_links.pdf

What is the click through rate (using real search engine data) on sponsored links – http://ist.psu.edu/faculty_pages/jjansen/academic/jansen_sponsored_links_one_listing.pdf

Overview and history of sponsored search - http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/jansen_overview_sponosored_search.pdf

Click Fraud - http://ist.psu.edu/faculty_pages/jjansen/academic/jansen_click_fraud.pdf

Handbook of Research on Web Log Analysis

I have co-edited a book on Web Log Analysis with Amanda Spink, Queensland University of Technology, Australia; Isak Taksa , Baruch College, USA entitled Handbook of Research on Web Log Analysis. As lead editor, I spent a great deal of effort ensuring that the book covered the spectrum of Web Analytics (as did Isak and Amanda).

The book’s expected publication date is October 2008, and it’s Webpage is here - http://www.igi-global.com/reference/details.asp?ID=8282.

The book’s table of contents is:

Introduction
“Research and Methodological Foundations of Transaction Log Analysis”
Bernard J. Jansen (The Pennsylvania State University), Isak Taksa (Baruch College), and Amanda Spink (Queensland University of Technology)


Section 1 Web Log Analysis: perspectives, issues and directions
“Historic Perspective of Log Analysis”
W. David Penniman (Nylink)

“Surveys as a Complementary Method to Web Log Analysis”
Lee Rainie (Pew Internet & American Life Project) and Bernard J. Jansen (The Pennsylvania State University)

“Watching the Web: An Ontological and Epistemological Critique of Web-Traffic Measurement”
Sam Ladner (York University)

“Privacy Issues Associated with Web Logging Data “
Kirstie Hawkey (University of British Columbia)


Section 2 Methodology and Metrics
“The Methodology of Search Log Analysis”
Bernard J. Jansen (The Pennsylvania State University)

“Uses, Limitations, and Trends in Web Analytics”
Anthony Ferrini (Acquiremarketing.com) and Jakki J. Mohr (University of Montana)

“A Review of Methodologies for Analyzing Websites”
Danielle Booth (The Pennsylvania State University) and Bernard J. Jansen (The Pennsylvania State University)

“The Unit of Analysis and the Validity of Web Log Data”
Gi Woong Yun (Bowling Green State University)
“Recommendations for Reporting Web Usage Studies”
Kirstie Hawkey (University of British Columbia) and Melanie Kellar (Google)


Section 3 Behavior Analysis
“From Analysis to Estimation of User Behavior”
Seda Ozmutlu (Uludag University), Huseyin C. Ozmutlu (Uludag University), and Amanda Spink (Queensland University of Technology)


“An Integrated Approach to Interaction Design and Log Analysis”
Gheorghe Muresan (Microsoft Corporation)

“Tips for Tracking Web Information Seeking Behavior”
Brian Detlor (McMaster University), Maureen Hupfer (McMaster University) and Umar Ruhi (University of Ottawa)


“Identifying Users Stereotypes for Dynamic Web Pages Customization”
Sandro José Rigo (Universidade Federal do Rio Grande do Sul), José Palazzo M. de Oliveira (Universidade Federal do Rio Grande do Sul), and Leandro Krug Wives (Universidade Federal do Rio Grande do Sul)

“Finding Meaning in Online, Very-scale Conversation”
Brian K. Smith (The Pennsylvania State University), Priya Sharma (The Pennsylvania State University), Kyu Yon Lim (The Pennsylvania State University), Goknur Kaplan Akilli (The Pennsylvania State University), KyoungNa Kim (The Pennsylvania State University), Toru Fujimoto (The Pennsylvania State University), and Paula Hooper (TREC)


Section 4 Query Log Analysis
“Machine Learning Approach to Search Query Classification”
Isak Taksa (Baruch Colleg), Sarah Zelikovitz (The College of Staten Island ity University of New York), and Amanda Spink (Queensland University of Technology)

“Topic Analysis and Identification of Queries”
Seda Ozmutlu (Uludag University), Huseyin C. Ozmutlu (Uludag University), and Amanda Spink (Queensland University of Technology)

“Query Log Analysis in Biomedicine”
Elmer V. Bernstam (University of Texas –Houston), Jorge R. Herskovic (University of Texas –Houston), and William R. Hersh (Oregon Health & Science University )

“Processing and Analysis of Search Query Logs in Chinese”
Michael Chau (The University of Hong Kong), Yan Lu (The University of Hong Kong) Xiao Fang (The University of Toledo), and Christopher C. Yang (The Chinese University of Hong Kong)

“Query Log Analysis for Adaptive Dialogue-Driven Search”
Udo Kruschwitz (University of Essex), Nick Webb (Institute for Informatics, Logics and Security Studies), and Richard Sutcliffe (University of Limerick)


Section 5 Contextual and Specialized Analysis
“Using Action-Object Pairs as a Conceptual Framework for Transaction Log Analysis”
Mimi Zhang (The Pennsylvania State University) and Bernard J. Jansen (The Pennsylvania State University)

“Analysis and Evaluation of the Connector Website”
Paul DiPerna (The Blau Exchange Project)

“Information Extraction from Blogs”
Marie-Francine Moens (Katholieke Universiteit Leuven)

“Nethnography: a Naturalistic Approach towards Online Interaction”
Andrade Braga (Universidade Católica do Rio de Janeiro)

Conclusion
“A Review of Methods in Presented in the Handbook of Weblog Analysis”
Isak Taksa (Baruch College), Amanda Spink (Queensland University of Technology), and Bernard J. Jansen (The Pennsylvania State University)

The Interplay between Branding and Search

I’ve been doing research on and investigating a search engine’s brand effect, namely how this brand affects how searchers perceive that engine’s performance and the usefulness of the results.

Findings? .... Brand has (can have) a major effect!

I ran a study where participants thought they were using 4 different search engines (Google, Yahoo!, MSN, and No Name) for four different searches. However, no matter what search engine the participants thought they were using, the results always came from Google (both in content and appearance).

There was a 25% different in participant evaluation of the usefulness of the results depending on the search engine, even though the results were identical in both content and appearance. Overall, Yahoo! scored the highest, closely followed by Google, then MSN, and then No Name.

Although, I call this branding, this really falls into the larger category of affective (i.e., emotional) aspects of people interacting with technology.

See here for a short paper on the branding study - http://ist.psu.edu/faculty_pages/jjansen/academic/pres/chi2007/jansen_branding_of_search_engines.pdf

See here for a press release - http://live.psu.edu/story/24878

Sponsored Search Special Issue

I was recently the guest editor for a special issue on sponsored search for the International Journal of Electronic Business (http://www.inderscience.com/browse/index.php?journalID=38&year=2008&vol=6&issue=2). As a form of online advertising, sponsored search is certainly one the of the most important innovations in search in years and very important aspect of Web marketing. Sponsored search provides the funding for the major search engines, allowing them to offer the 'free' search that has become such a part of our daily lives.

Given that, one would think that there would be a lot of academic research in the area of sponsored search. Currently, there isn’t. Other than some marketing articles and some papers on online auctions, there isn’t much academic activity in the area.

I expended a lot of effort just trying to get enough good articles for the special issue! The number of folks doing research outside of the search engine and internet marketing firms is limited at this time. Although, I am betting it is going to increase.

If you are interested in an overview, history, and technology of sponsored search, here is article that I co-authored on the topic (http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/jansen_overview_sponosored_search.pdf)

User Intent in Web Searching

One of my research areas is trying to determine user intent (i.e., the general classification of content indicated by a searcher’s query).

The general three classifications, from Broder (http://www.sigir.org/forum/F2002/broder.pdf#search=), are informational, navigational, and transaction. Informational searching involves looking for a specific fact or topic, navigational searching seeks to locate a specific Web site and transactional searching looks for information related to buying a particular product or service.

My research has explored the sub-classifications of these three broad classifications.

More importantly, I’ve been trying to classify automatically queries based on these classifications or their sub-classifications. So far, we have been making good success; about 75 percent accuracy using some simple methods and about 90 percent using more complex approaches.

The first full research manuscript is here (http://ist.psu.edu/faculty_pages/jjansen/academic/pubs/jansen_user_intent.pdf), with two more in the works using other methods.

Read the press release on this research here (http://live.psu.edu/story/29879).

Visit to Pepperjam

I recently visited Pepperjam (http://www.pepperjam.com/) located in the … perhaps … unlikely location of Wilkes-Barre, Pennsylvania given that Pepperjam is an internet marketing firm. The whole visit was really great!, with the only hitch being me getting lost in Wilkes-Barre. :-)

Kris Jones is the founder, president, and CEO, and Mike Jones is the COO. Interestingly, Kris and Mike were high school buddies, which makes me think how often pairs of folks have started companies (e.g., Apple, HP, Microsoft, Google, Yahoo!, Berkshire Hathaway, etc.).

Once I got there, I ‘immediately’ gave a talk to the search division and most of the senior management team (http://www.pepperjam.com/our-team.php) on the topic of search engine branding. We had a great follow-on discussion that touched on branding, search engine quality, the search engine market, academic studies (or the lack thereof) in the search marketing area, and search engine marketing.

We followed this discussion with a tour of the Pepperjam operations. It is a high energy place, with a diverse work force. From a review of the Pepperjam Website, one can tell that the company has been very successful.

The company has a really ‘non-PA’ working environment that is more reminiscent of what one would find in a Silicon Valley company. I asked about the ability to hire qualified employees. With the numerous colleges in the area, Kris said this was not a problem. Additionally, being located in Wilkes-Barre really reduces operating costs for the company. Concerning the opportunity for discussions and interchanging of ideas with others in the field, Pepperjam leverages the numerous conferences in the search engine marketing domain.

I was really impressed with Kris and Mike’s commitment to their hometown and the positive impact that Pepperjam is having on the local area. Really wonderful ‘industry – local community linkage’ that can serve as a role model for others.

Kris has authored a book from Wiley on search engine marketing from Wiley Publishing (http://books.google.com/books?id=jtYsGgAACAAJ&dq=Jones+Internet+Marketing)

IDEAS @ IST – an undergraduate research program

I have always been really enthusiastic about undergraduate research, having mentored several undergraduates in thesis and other projects (for a list of some of these students, see http://ist.psu.edu/faculty_pages/jjansen/academic/acad.html#students).

With support from the College of IST, I lead the kick off an undergraduate research program -- Inter - Disciplinary Educational Application Scholarships at IST! (IDEAS @ IST!) with a focus on recruiting potential freshmen / sophomore undergrads as research assistants (RA) for faculty projects.

Among the strategic goals of the program is to prepare a pool of undergrads to be competitive for graduate school. The specific program objectives are to identify promising freshmen / sophomore students, get these students involved with faculty research programs early in their academic career, and leverage the talent of these students to enhance the research efforts for faculty.

The specific strategic goals of the effort are: (1) to enhance the IST research program by leveraging undergrad students, (2) use the IST research program as another item in the IST undergrad recruiting efforts, (3) prepare a pool of our undergrads to be competitive for graduate school and in their professional careers

IDEAS @IST! started it first cohort in Spring 2008. With freshman / sophomore RAs, it is many times "an investment in the future" at first, but there is the potential for good payoff down the road. Several universities have leveraged their undergrads in their research programs with great success (e.g., MIT, Sanford, and Berkley). Several professors (including me) have had good experiences with undergrad RAs, including co-authored publications.

The first semester was really successful with approximately 30 students participating. We kicked-off the semester with a “getting started” dinner with students and faculty. We ended the semester with a research poster session, dinner and guest speaker, Dean Christian Brady (http://www.scholars.psu.edu/about/deansmessage/) of the Schreyers Honor College at Penn State. See Dean Brady’s podcast of his presentation here (http://engage.shc.psu.edu/?p=390)

I really believe that IDEAS @ IST! is an extremely worthwhile program that can only help undergraduate students, improve the marketability of an undergraduate degree, and enhance research productivity. The core concepts of can be applied to similar programs in other universities and settings.

Google Online Marketing Challenge

This spring, I taught a course centered on the Google Online Marketing Challenge (http://www.google.com/onlinechallenge/). Basically, teams of students had to recruit a SME and run an Adwords campaign for this SME for 3 weeks.

The Google Online Marketing Challenge was certainly one of, if not the, biggest academic endeavor ever undertaken. Final numbers for the Challenge were:

Total Universities - 468
Total Classes - 629
Total Student Groups - 4,317
Students @ 5 per group - 21,585
61 countries (44% from EMEA, 40% from the Americas and 16% from APAC)

The course was a great experience for me, and from all indications, the students and local businesses as well. We ended the course with a celebration dinner at a local restaurant (see story here - http://www.istbuilding.com/2008/04/30/google-online-marketing-chall...).

We ran a local competition giving awards to the top three of five teams (Amazon gift certificates for each member). We also purchased t-shirts for everyone in the course (designed by a student committee), and took everyone in the course out for a celebratory dinner at a local Indian Restaurant. …. I didn’t want any of the teams to leave without a prize, so I purchased small prizes for members of the other two teams. :-). …

It was a really fun course for me and a great learning experience. The course has a positive impact, causing at least two students to consider marketing minors, provided great insight into technology for the business major students, and has been a point of interest for the seniors in the course as they go on job interviews.

The Challenge has also generated some good press. See http://live.psu.edu/story/29647.

Overall, the Challenge was just great. A big thanks to Google and especially Lee Hunter (Google) and Jamie Murphy (University of Western Australia) who did the heavy lifting to get the Challenge off the ground and executed.

Interesting Technology, Information, and People Problems

Hey!

At times one runs into real life TIP (technology-information-people) problems.

There is a school that I am doing some volunteer work with that has a large ESL population.

The school is trying to ‘reach’ the parents of this demographic group better, including getting the parents information on school events of importance to their kids, etc. (the problem).

The school has an excellent electronic Bulletin distributed via privately run servers that provides news on school happenings (technology)

Based on discussions with county’s migrant workers and ESOL staff, keeping calendars is just not part of the lifestyle for these parents (culture)

And even if it was, most don’t have computers to get the Bulletin (digital divide).

They do have cells phones. But, the cell phone numbers are continually changing (economic)

Which made me think of the US government’s efforts in the 1900 to provide basic low cost land line service (policy)

School solution? --- Monthly potluck dinners (people)