Sign up for weekly recaps of the ever-changing search marketing landscape.
Interview With Louis Rosenfeld, Author Of Search Analytics
As search engine optimization (SEO) professionals, we focus on keywords, aboutness, and information scent. What keyword phrases best describe a webpage, a graphic image, and/or a video? How can we ensure that information scent gets stronger from search engine results page (SERP) to a web page on our websites?
One of the untapped places to conduct keyword research effectively is the analytics data from your site search engine. In his book, Search Analytics for Your Site: Conversations with Your Customers (*use Discount Code “SEL” to receive 15% off the book now*), information architect Louis Rosenfeld describes how website owners can implement various metrics on their site search engine to:
- Better understand searcher needs
- Improve your website content, site navigation, and search performance
- Understand keyword phrases in context
- Get useful qualitative and quantitative metrics
Lou has graciously agreed to an interview for Search Engine Land readers. For the next two columns, I wanted to sum up what came out of our phone conversations, email, and other correspondences.
Q. What are site search analytics (SSA)?
Rosenfeld: Your users are telling you what they want from your site—in their own words—when they use your site’s search engine. Are you giving them what they want?
SSA is simply a set of tools and methods to help you harvest users’ queries, test and measure how well they’re performing, and actually see how well you’re serving users. That’s it in a nutshell.
Q. That is what your search analytics book is all about. How are web searchers different from site searchers? How are they similar?
Rosenfeld: Searchers often have a better idea what we want from a specific site than they do from the Web in general. So, as you’d expect, site search queries tend to be narrower and more relevant to the site itself.
For example, if you’re examining the University of Michigan’s SSA data, you won’t see nearly as many name queries (e.g., “UM,” “univ michigan”) as you will with Web search keywords. Instead, you’ll see queries that are unique to you (e.g., “football tix,” “get my transcript”).
Interestingly, knowing these more specific queries might help you invest in search engine advertising more wisely—you may be able to avoid broad, expensive keywords (e.g., “University of Michigan”) and instead bid on less expensive, more precise ones (e.g., “Michigan football tickets”) that may still be under your competitors’ radars.
Q. What types of patterns do you look at when you review site search analytics?
Rosenfeld: It really depends, as there are so many ways to categorize queries; here are just a few:
- Audience type: who seems to be doing the searching?
- Topic: what are the queries about? Products versus tutorials versus forums versus something else?
- Content type: what kind of content do they seem to want to retrieve?
- Tone: technical jargon versus marketing jargon versus plain language terminology
- Seasonality: does the nature of users’ information needs change over time?
Because different people see different patterns, it’s ideal to have many people do the analysis—at least enough to start seeing if consistent patterns emerge. You might even have the individuals on your team do a card sorting exercise with a set of your site’s queries and then compare results.
The good news is that even if one or two people look for patterns in your site’s query data—for an hour or two every quarter—you will learn something useful. I guarantee it.
Q. I truly understand that. On a number of our client sites, we discovered that users/searchers keep going to the same blog articles over and over again via site querying. From that data, we were able to pinpoint keyword phrases that never occurred to us, even with all of the Web-based keyword research tools available to us. On that note, what is a search exit? How can website owners utilize search exit data to provide a better searcher experience?
Rosenfeld: I like Google Analytics’ definition just fine: “The percentage of searches that resulted in an immediate exit from your site.” The percentage of search exits per query is a common failure metric, along with measuring queries that retrieve zero results (which I happen to like even more).
It’s useful to flag and study queries with high search exit rates—in fact, it’s one of those things to do on a regular basis. Look for patterns—are there types of queries (e.g., product names) that have surprisingly high search exit rates? Better take a deeper look at how product name queries are performing and why they’re failing. That’s where qualitative research—specifically user studies involving task analysis—work really well hand-in-hand with exploring your analytics data.
Q. What do search queries (keyword queries) that get zero results typically mean? Why is that a problem for website owners?
Rosenfeld: Zero search result queries are generally due to one of two varieties of failure:
- You don’t have the content. This should lead to a follow-up question: why don’t you? Your customers are searching your extreme mountain biking products site for an insurance policy that’ll cover them if they fracture their skulls on the way down Pike’s Peak. Maybe there’s a business opportunity here that you’re missing by not offering such a policy?
- You have the content, but users can’t find it. Your content may not be written in plain language. Or your content authors don’t or won’t follow guidelines for titling and tagging their documents. Forget the constant cajoling: now you can reverse engineer the process and show content owners what happens when they don’t follow those guidelines.
Q. I have often stated that “How + Why = ROI” which means that website owners should understand how people search (which we can get from interpreting log file data), and why people search.
Rosenfeld: Absolutely. But keep in mind that SSA, like other types of analytics, can’t tell you how or why; it can only tell you what. Site search analytics data gives us a fantastic perspective on what happens when users search our sites. But it’s just one tool in the user researcher’s toolkit.
Stay tuned for Part 2 of Lou’s interview, where we will go into more depth about search sessions, precision scores, relevancy scores, how to use site search analytics to improve navigation, and some insightful tests you can use on your own site.
*Editors Note: We would like to thank Rosenfeld Media for the 15% discount for our readers. Neither Search Engine Land nor Shari Thurow has any financial interest in sales of the book, although Ms. Thurow has endorsed the book as a top read for SEOs.
Image from Rosenfeld Media, used under Creative Commons license.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.