Interview with Penn State Professor and 'Search Behaviorist' Jim Jansen
Jansen is an assistant professor at the College of Information Science and Technology at Penn State, and an expert on Web searching, sponsored search and personalization for information searching. He's authored more than 150 publications in the area of information technology and systems.
The interview with Jim Jansen covers topics like his recent research on click behavior, his ideas on how Google and other engines can improve user experience, as well as his current Twitter research.
Read on for more.
In your study on click behavior with integrated paid and organic search results, you cite “negative bias” and “distrust” as the reasons searchers click on sponsored ads (like Google sponsored links) less than 16% of the time. Are these distrust issues solvable? If so, whose job is it to solve these issues? The engines or the advertisers?
I take the approach that most things are solvable. The hard issue is deciding what is really a problem and what is just ‘the way things are.’ In Web searching, there are queries that are probably better addressed via organic results, so one shouldn’t look to advertisements to address 100% of Web searches.
That being said, there are certainly queries where ads are appropriate, and these should be the focus.
The solution is both engineering and advertising. The search engines must continually improve the ad platforms to serve relevant and targeted results. The advertisers must write compelling ads.
There is also an education component that search engines must address, either directly or via word of mouth. This education component is that these ads can effectively speak to user needs.
One of the conclusions of your research is that combining paid and organic listings in a sort of "integrated SERP" may overcome the bias against pay-per-click advertising links and improve the management of "screen real estate." Can you discuss why you think this idea will work?
Well, I am not sure this is really the final solution. However, what I was getting at is that by separating the sponsored links (i.e., PPC ads) from the organic links (i.e., non-ads), it calls attention to them. Some users respond to this attention in a negative way (e.g., these must not be good). The result is that many users, notably Firefox users, just use ad strippers and don’t even see the ads at all.
The only element that distinguishes sponsored links from non-sponsored is money. However, there are companies that spent a lot of money on search engine optimizing their Website. The search engines do not identify these to the consumer. Why sponsored links?
I understand that it is not as simple as this in real life, but the separating of sponsored links is adding to the negative bias.
An integrated listing may be something to try.
I’ve read that you’re now conducting research on automated searching assistance systems to help design more advanced search engines. Can you talk about what this means in practice? And do you think “query reformulation” is the future of search?
As of this interview, there has not been a major advancement in Web searching in more than a decade; the last were page ranking and sponsored search that hit the streets at about the same time.
So, we are really leveraging the intellectual capital of the last decade.
However, we are really seeing signs that we might be on the cusp of another technological advancement, with several algorithmic publications and prototype systems emerging that could really change the nature of Web search. Most have a similar slant, which the technology taking more of the computational load of searching strategies and tactics.
We’ll have to see what happens.
Given your extensive knowledge of the click behavior of searchers, what in your opinion can search marketers do to improve their own search marketing performance efforts and overall click-through rate?
At the core are the foundational elements of marketing and advertising, have a good product that consumers want (or convince consumers that they want it), rely mainly on word of mouth advertising, focus on customer relations, and build a solid brand image.
For a more direct answer to your question, ideally you want each customer to have a personalized experience. One can do this with keywords advertising in a way that has not been possible in the past.
However, it is some what of a balancing act. You can’t get too personalized or the customer gets freaked out (i.e., ‘they’ know too much about me).
You seem to be a fairly active Twitter user (@jimjansen), and you obviously have an intimate knowledge of search. What are your thoughts on “real time search?” Is it a Google killer? Is Google in the best position to “do it right?” Is it overblown?
Twitter may turn out to be a really effective information source by providing real time answers to questions from trusted sources. Working with collaborators, we just finished a major analysis of tweets. About 20% of them were information seeking and providing, so the potential is there.
However, do I believe it is a Google killer? No, just as listserv and discussion boards were not. These were around long before search engines, and search engines emerged. So, obviously there is an information need that this type of service doesn’t provide that search engines do.
As for Google doing it right, yes. They have loads of talented folks and billions in cash reserves. Yahoo! and Microsoft have the potential but they are distracted with internal problems that slows them down.
You noted in a blog post that there probably wasn’t anything game-changing about Kumo/Bing (paraphrasing). Do you have any thoughts on what a game-changing search engine looks like? Will it have to do something fundamentally different from Google, improve upon a weakness, or is stealing share virtually impossible at this point?
Nothing is impossible. Bing is fine. As is Yahoo! Ditto for Ask.
Technology aside, the Google brand is flying high. It is so ingrained in our information seeking consciousness that is a more than a habit. It is like the 110 v electrical grid in the US – one would need a really good reason to change. A better analogy might be the metric system. There are some good reasons for the US to change to it, but our current system works better good and we’re use to it, so why bother.
I see two game changing scenarios:
(a) Some technological advancement occurs that Google overlooks or discounts, providing an opportunity for another company to make substantial inroads into the market.
(b) Something happens that causes people to change their habits, like losing trust in Google or switching their branding affiliation to some other company.
In either scenario, Google has to make a mistake. I use the analogy of the top twenty five college football rankings. For a team to move up, a team higher up the rankings has to screw up. I believe it is the same in the search engine market. Google will have to screw up for someone else to move up.
A real hot-button issue in search marketing, particularly PPC marketing, is attribution. The consensus is that “last click” isn’t a viable long-term answer. Do you see a solution to the answer readily available in any of the alternate models currently out there? If not, have you seen any indication as to where the solution will come from?
Several organizations are working this and there is some good research from SEM firms and search engines that show the temporal lag from “ad in front of face” to “potential purchase”. iProspect, Microsoft, Atlas (now a Microsoft company), and others are all working this.
We need two things, technology (i.e., a means to implement it) and acceptance (i.e., the search engines, the advertisings, and publishers see the value of it).
You teach a course on “Leveraging Search Engines for Successful E*commerce,” which honestly looks like a fascinating course, one I’d like to take. Can you tell me more about what specifically students learn regarding "designing and implementing e*commerce Web sites to leverage the informational technology of search engines and the Internet?" Also, with the ever-changing nature of online marketing, what sorts of things do you do to keep a course like this current?
Yes, I love this course. The students recruit real clients, typically small to medium size enterprises that have never had an AdWords account. They implement and manage an account for this SME and compete in the Google Online Marketing Challenge, which is a great program.
So, the students get to use a real world advertising platform with real money and working for a real client. Really what classroom education should be.
Additionally, it provides a great connect between a university and the community. I’m really driven to make my research applicable to my local community, which is more difficult than it might seem. We can easily tie research to long term, pie in the sky goals, but linking this research and teaching to the local community with immediate evaluation is challenging. I’ve been fortunate to leverage search engine marketing to aid some local businesses.