In understanding search user behavior, no one is a more respected voice than Gord Hotchkiss. He and the research team at Enquiro have built a solid reputation as the leading experts in understanding not only what happens on a search portal, but also why. Gord’s voracious personal curiosity extends into areas as diverse as neurology, psychology, genetics, sociology and anthropology, always with the goal of understanding why we do the things we do and what that means for marketing strategy.
In addition to being the CEO and co-founder of Enquiro, Gord is a director and the past Chairman of SEMPO, a columnist for MediaPost and Search Engine Land, a regular presenter at all the industry shows and a popular keynote speaker.
SMS: Enquiro has taken a fairly unique approach to search engine marketing research. What prompted you to choose eye-tracking over other methods of data collection?
Gord: Eye tracking is only one of the tools we use. We’ve also done survey work, standard usability work, and have even done our first neuro-scanning study with a local university. That said, we do like eye tracking as a tool to gain insight into user behavior because it gives you a deep data set, especially when you combine it with post-session questions. It allows you to combine and compare what people physically see with what they remember seeing. Also, some of the most interesting findings in our eye tracking research come when we slice interactions down on a second-by-second basis. It provides tremendous insight into how people digest and assimilate the content on a page.
SMS: You often touch on neuroscience and its relationship to marketing in your writing. Can you identify a couple of current or recent research areas that you think will affect how marketers approach online optimization and advertising?
Gord: Dr. Gary Small at the Semel Institute at UCLA is doing some really interesting work on digital “natives” versus digital “immigrants” and how our brains are being rewired through exposure to online technologies. As far as I’m aware, he’s done the first functional Magnetic Resonance Imaging (fMRI) studies showing which parts of the brain are activated when we use search engines. This is a particular area of interest for me, as I have my own theories about the path taken through different areas of the brain while one views a SERP. Also, Peter Pirolli’s work (Information Foraging Theory: Adaptive Interaction with Information) should be required reading for anyone in search.
Finally, the work done on evolutionary psychology by Leda Cosmides and John Tooby, as well as on behavioral economics by Daniel Kahneman, the late Amos Tversky, and — more recently — Dan Ariely, is fascinating reading for anyone in marketing.
SMS: Over the past couple of years, we’ve seen the so-called Golden Triangle on the search results pages largely replaced with a more scattered, blended focus. How does personalization, and now Google’s SearchWiki, change the equation? What potential new behaviors can we expect in user interaction with the new SERPs?
Gord: The Golden Triangle was a product of the SERP layout — a text-based list of ordered results. Really, the only variant was the degree of relevancy at the top. So, two factors will impact that in the future — the visual saliency of relevancy as engines get more adventurous in their results layout, and the actual relevancy of the results to the user. For example, we interact with images in a substantially different manner than we do with text, so this impacts scanning behaviors. Also, the SearchWiki idea has the potential to let us define our own relevancy.
One of the things we’ve found is that our own search interactions are largely conditioned and subconscious. We search by habit. As the environment changes — i.e., the layout of the results page — we can’t rely on habit to guide us through. And that makes it tougher for us to search. Engines need to give us a reliably consistent search experience so we can create conditioned habits (e.g., starting scanning in the upper left is a conditioned habit). These conditioned behaviors will be in flux as engines experiment, but ultimately we will go to the engine that offers a familiar and efficient experience. We’ll go to the one that makes it easiest to find what we’re looking for.
SMS: When discussing your findings, you talk a lot about perceptions as they pertain to the quality of search results. In what ways are perceptions important in searching?
Gord: Perceived relevancy is huge. It dramatically impacts our satisfaction with the search experience. And these perceptions of relevancy happen very quickly, often subconsciously. Pirolli’s work on information scent is very helpful here in understanding how this happens. Right now, that scent is largely carried in the text of the listings, particularly in the title of each listing, and our criteria for relevancy is not particularly sophisticated. The query we use, repeated in bold type in the listing, is a key reinforcement of relevancy. The faster you can give the perception of relevancy in a listing, the faster you’ll capture a click. Of course, once you deliver perceived relevancy, you had better deliver real relevancy after the click, otherwise you’ll quickly lose users. But when it comes to interactions on the search page, perceived relevancy is the whole ball game. Real relevancy only matters after the click.
SMS: Could you explain exactly what you mean by the “brand lift of search”? Is branding as important as everyone thinks in terms of getting that initial click?
Gord: Originally, we were asked to find out if the appearance of brands on the SERP lead to higher brand recall; if there is a latent branding impact even if the result isn’t clicked on. This is a very complicated question that’s difficult to cover quickly, and it comes down to the metrics you choose to use in determining branding impact. Brand is a tremendously broad and somewhat ambiguous concept. So, some aspects of branding are very effectively delivered in a search experience — for example, the likelihood of making it into a prospect’s consideration set while they’re researching a purchase. And in that regard, yes, brand is just as important as everyone thinks in getting that initial click. In fact, it’s more important than everyone thinks.
Others aspects aren’t — for example, if you’re looking at search to create an emotional brand connection like a TV ad might. Personally, I think the whole idea of branding is being completely turned on its head right now anyway, so we’re embarking on a fool’s errand if we’re trying to nail the concept down in over-simplified metrics.
Previously, we had this concept of “pushing” controlled brand messages out. We have to realize that brand control is an illusion — your brand is whatever someone perceives it to be. Part of that is based on your brand messaging, but it also depends on experience, word of mouth, news reports, and many other factors, most of which are totally beyond the control of the supposed owner of the brand. The customer is the real owner of the brand — because their concept of the brand exists uniquely and totally in their mind — and it differs with each individual. All the advertiser can do is contribute to that, hoping that their efforts strike a positive chord.
SMS: Let’s shift gears and talk about post-click user behavior. Based on your research, what are some of the signals users are looking for when deciding on the quality and relevance of a website? What could marketers do to amplify those signals?
Gord: Pictures are huge, as are clear and relevant headings and subheadings, with lots of white space around them. Also, intuitive calls to action (prominent links) that reinforce information scent and give clear paths to follow. Some of my personal favorites are prominent bullet lists or feature comparison matrices.
Another best practice is what I call “intent clustering.” For any given landing page, think about what the top two or three user intents would be going forward from that page. Then cluster content and navigation options around those intents, reinforcing in multiple ways — a picture, a subheading, some links, a text description and a few bullet points. Things reinforcing scent around a potential path can be very compelling and intuitive to the user.
SMS: To sum up, what are some of the key user behavior trends that all search marketers should be aware of?
Gord: Here are some:
- We take the path of least resistance to the information we’re looking for.
- The more competitive your market and the more alternatives there are for the user, the less patience they’ll have with a poor experience.
- Keeping the above in mind, remember that you can’t drive an online prospect anywhere they don’t want to go.
- Branding online is about experience, not exposure (courtesy of Jakob Nielsen).
- We are creatures of habit online as well as offline. Don’t assume people have their brains fully engaged when they’re searching for you or visiting your website.
- Looking is not the same as seeing. We look at many things, but only a few make it through our attention filters and are actually seen and understood. And it’s not always the most obvious things that are seen.