Analyzing Search Behavior for SEO

Many people believe the primary goal of SEO (define) is to achieve top positions. If a site has top search engine positions, the target audience will always click on the site link, then make a purchase or type of inquiry. And even if site visitors don’t make a purchase, they’re left with a positive branding experience because a company must be outstanding and trustworthy to be listed at the top of Google, Yahoo, MSN Search, and

Man, I wish SEO were that simple. And I wish search behavior were a straightforward, linear process. But it’s not. Search behavior involves more than the querying process, the behavior most search engine optimizers focus on.

Types of Search Behavior

For SEO and sales conversions, many information retrieval and usability experts have identified a number of different search behaviors, including but not limited to:

  • Berrypicking
  • Querying
  • Refining
  • Expanding
  • Browsing/surfing
  • Pogo-sticking
  • Foraging
  • Scanning (eye-tracking)
  • Reading

The one landmark search behavior that was identified long ago in a 1989 paper by Marcia J. Bates is berrypicking. Many of the other types of search behavior are actually subsets of berrypicking. What’s so phenomenal about Bates’ research is search behavior isn’t a linear process.

(Note: This is an academic article. If you want to read it, which I highly recommend, don’t expect any quick tips or tricks for optimizing a Web site for commercial Web search engines.)

SEOs and Search Behavior

Though Bates doesn’t specifically mention the Web in her research, her paper truly changed the way I approach search behavior. As I mentioned in a previous column, many SEO professionals consider the word “search” to mean only querying behavior. Unfortunately, this attitude is precisely what makes many SEO professionals extremely narrow-minded.

For some odd reasons, most SEO professionals seem to believe the true measure of a successful SEO campaign is top positions, and they use Web positioning software to gauge their success.

This narrow view of search behavior is firmly based on the belief “search” means only querying behavior. In other words, if you get your client to come up in the number-one position in Google for “press releases” in any way possible (such as an invisible frameset), you’re a successful SEO professional.

From that single query evolve many other types of search behavior. After a person types in a series of keywords into a search box and clicks, he will either read or scan search results to see which listings best match the query. If the search results are unsatisfactory, the searcher may refine the keyword phrases. Then, he will either read or scan more search results.

If the searcher clicks a link on a commercial Web search engine, he will read or scan the landing page to determine if the content matches the query. Browsing a site for further information is also common. Searchers often pogo-stick between SERPs (define) and commercial Web sites to find the best answer to their questions.

Look at all of the search behaviors I just mentioned: querying, scanning, reading, refining, and pogo-sticking. Do SEO professionals even consider optimizing Web sites for all types of search behavior?

As long as SEO professionals continue to obsess over positioning, they’ll continue to produce substandard content and poorly developed and architected Web sites.

Matching Search Behavior, Design, and Information Architecture

Understanding how your primary and secondary target audiences query should be an important part of persona development. It still amazes me that I don’t see Web site usability professionals incorporating search behavior into personas.

For example, I remember trying to come up with a navigation scheme for a pet supplies Web site. With a card-sorting usability test, usability professionals ask participants to group topics into categories, then to present possible category names. On this particular site, one navigation button label was “pet supplies.” I assumed “pet supplies” was a logical navigation button label. It seemed logical and self-explanatory.

Out of curiosity, I did a little keyword research. I noticed people tended to search for “pet products” more often than “pet supplies” on Google and Yahoo. Intrigued, I decided to conduct an A/B test for the navigation button label. I was genuinely surprised to see participants clicked on “pet products” more often than they clicked on “pet supplies.”

Had I not begun my Web design/development career as an SEO professional, I’d have never thought to utilize keyword research as part of the usability process. I knew to incorporate the keyword phrase “pet products” in title tags, meta tags, cross-links, and navigation. Important keyword phrases became an integral part of the site’s information architecture.

The result? All types of search behaviors are addressed. Query-based searches (from both commercial Web search engines and site search engines) are more accurate. Site visitors are provided with a clear sense of place when browsing, scanning, and reading Web pages. Through effective cross-linking, pogo-sticking behavior was minimized. The site’s ROI (define) increased over 37 percent within three months by effectively addressing all types of search behavior.


I knew about many types of search behavior long before I read Bates’ paper. Yet it took reading the article, writing a paper about it (I’m back in grad school), and truly absorbing its meaning to get the big picture. When I realized Bates’ research also applies to Web sites, I swear my brain went “CLICK!” I felt I reached a higher level of search understanding. It was a very Zen moment.

Search behavior evolves. Search behavior consists of many different processes. When Web developers, SEO/SEM (define) professionals, usability professionals, and information retrieval scientists can see the bigger picture, they produce more successful Web sites.

It took an academic article to inspire an understanding of the big picture for me. I’ve read a number of other academic articles that gave me a greater appreciation of information retrieval algorithms and heuristics, but I’ll leave that column for another time.

Want more search information? ClickZ SEM Archives contain all our search columns, organized by topic.

Related reading

Brand Top Level Domains