Leveraging On-Site Search Data

BY Jason Burby
March 14, 2006

Search is a very hot topic. There‘s a great deal of discussion about its importance, appropriate spending on search engine ad words, the perfect balance between organic and paid search, and recent changes to search engine algorithms.

Though those topics are important, it‘s also important to think about the right answers for your business and site. You quite possibly have more to gain through the other "search" -- the little-talked about but very powerful on-site search. What I‘m referring to is the search box on Web sites that‘s frequently used when visitors can‘t find what they‘re looking for, be it a product, store location, article, or anything else.

This lower-profile search may very well touch more visitors than the other search mechanisms that drive people to your site.

About 18 months ago, I wrote a column that details how to measure the effectiveness of your on-site search and identify ways to improve it.

Today, I want to focus more on understanding the terms people are searching for and how you can use this information to improve your site overall. Examining data from your on-site search can provide a tremendous amount of insight about visitor‘s intent, as well as what‘s easy to find and what‘s harder to locate on the site.

Aggregate Search Terms

It‘s often easiest to start by looking at the most common terms people search for. Create a list for a period long enough to give you a good sampling. The most recent month is a good place to start. Then, review the different search terms and group them in common categories. What types of things are people searching for? Look for common patterns or large percentages of people searching for the same thing or the same types of things.

Search Terms by Site Section or Page

Now that you‘ve viewed search terms for the overall site and found some key trends or identified common terms people search on, you want to understand the terms by site area. Look at the most common search terms based on where people search from based on site section or even individual pages. This will allow you to understand specific problems or concerns people are having on those specific pages compared to other sections or pages throughout the site.

Segmentation Based on Visitor Type

Another way of understanding what visitors are looking for and thinking about is to analyze search terms based on different visitor segments. An easy task is to compare the search terms used by first-time visitors against returning visitors and previous buyers. By understanding the different terms they‘re searching for, you can better understand what questions they need answered and what their key interests are.

Insight Through Search Analysis

Recently, when analyzing on-site search data for a company that sells apparel and related products online, we identified a few interesting findings.

First, most visitors search within a specific line of products by color, not by style. This helped drive a change to the way products were displayed and could be sorted. It also helped adjust what colors would come up based on different terms. If a specific product color is officially "pumpkin," for example, we want to make sure when someone searches "orange" the pumpkin-colored product returns. (I admit I don‘t know the difference between "orange" and "pumpkin").

We also found a lot of searches regarding order tracking. Even though it was listed as a key option in the top nav bar, people just weren‘t seeing it. Through a simple treatment to the order-tracking button, we were able to reduce confusion and need for people to search for it.

I‘m not saying you‘ll find the same issues on other apparel sites, but you should look at your own search terms and see what you do find.

If you‘re serious about improving your on-site search strategy, check out that earlier column and pay special attention to the "Search Return Exits" metric covered to identify the low-hanging fruit.

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

print this article | e-mail a colleague | send feedback | Read Feedback

Jason Burby is the director of Web analytics for ZAAZ, a Web business consultancy implementing data-driven business initiatives for long-term clients across the U.S. Using performance scorecards, A/B testing, tool reconfiguration and other techniques, Jason helps companies better use Web analytics data to improve site business results.

He‘s worked with Washington Mutual, Wachovia, T-Mobile, Converse, Alaska Airlines, Microsoft, Sprint, Levi Strauss, Qwest and A&E Television Networks. Jason speaks frequently at conferences and seminars helping spread the word on the effective use of Web analytics. In addition he is the co-chair of the Metrics/KPI committee of the Web Analytics Association.