
No matter what you do with your visitor data — no matter how good your analyst is, no matter how advanced your segments are, no matter how customized your reports end up being — it can't tell you what users want. Oh, the data can show you patterns that come close — funnel drop-offs for this type of intent, groups of pages accessed most for that other thing — but you still have to infer rather than analyze, and you're inferring against data with a lot of noise.
Long story short, traffic doesn't give you precise data about user intentions. Full stop.
There are really two things that most sites do to get good qualitative data. One is having an exit survey — i.e., asking users directly about what they want. Companies like iPerceptions, OpinionLab and ForeSee can get you direct comments from users, and this is excellent for intent. The other thing is something most sites already have, but not a ton of site owners mine for intent data: internal site search.
Analysis of internal site search can reveal a lot more than sheer number of people searching or items directly being searched for. It's a direct verbatim from your customers and prospects. Consider the following when analyzing your internal site search:
1. Groups of queries determine what's broken. If you take the time to study your top queries and group them into buckets, you'll see patterns emerge. The things people search for most are things that aren't obvious from your navigation structure. Make sure you fix that issue.
2. Pages that lead to a lot of searches may require usability tweaks. If there are pages where the number of searches is disproportional to the overall visibility of the page, then the content of the page needs to more closely align with the searches.
3. Site searches tell you what people need most. If some things are trending on your site or there are common themes that surface, make sure you feature those things more prominently. Your prospects are telling you to.
4. How successful you are with, well, internal site search. Make sure the tool that's giving you the treasure-trove of data isn't itself broken. Make sure there's enough space to accommodate user queries (users will shorten their search terms if there's little space, and that's more likely to lead to failure). As importantly, review the search depth for your most popular terms if visitors are refining their queries all the time for something that's used often. If this happens a lot, devise a plan to either change your site search engine or customize it for the terms that fail the most.
Now, just knowing what visitors want doesn't get you all that much. There are dozens to hundreds of issues you need to fix, and almost no company has resources to fix everything. Still, if you pay attention to what visitors want in their own words, prioritizing should be just a tad easier.

Tim Ash is the author of the bestselling book Landing Page Optimization, and CEO of SiteTuners. A computer scientist and cognitive scientist by education (his PhD studies were in Neural Networks and Artificial Intelligence), Tim has developed an expertise in user-centered design, persuasion and understanding online behavior, and landing page testing. In the mid-1990s he became one of the early pioneers in the discipline of website conversion rate optimization. Over the past 15 years, Tim has helped a number of major US and international brands to develop successful web-based initiatives. Companies like Google, Expedia, Kodak, eHarmony, Facebook, American Express, Canon, Nestle, Symantec, Intuit, AutoDesk and many others have benefitted from Tim's deep understanding and innovative perspective.
Connect with Tim on Google+