You set up your Google Analytics tracker, goals, and filters. Now you need to learn something from all that data you’re collecting. When you look at the reporting for your web property, the first thing you’ll see is an audience overview that shows a variety of seemingly useful metrics, like average session duration, number of users, and bounce rate. That’s fine, but is it useful?

Knowing that the bounce rate for all of your sessions is 35% doesn’t really tell you much. You don’t have any context for why those sessions bounced or if some areas of your help had higher bounce rates than others. If you want your data to answer questions, you need to be more specific. Stop looking at all sessions and start using segments.

Segments allow you to filter your data. You could create a segment for users that enter your help on the default page and another for users who don’t. You can apply these segments and start getting more interesting data. Now you can compare users who entered the help on your default page to those who didn’t and see what’s different. Is every page in your help actually page one? Now you have the segments and data to find out.

Here are some interesting segments:

• Sessions that searched or used the table of contents
• Sessions that searched but didn’t exit on a search result
• Sessions that entered through the main landing page
• Sessions that came from one of your other web properties
• Sessions visiting from a server machine
• Landing page bounces

If you’re having trouble coming up with segments to create, take a step back. Figure out what questions you want your data to answer, for example:

• Are users able to navigate my help without using the table of contents or search?
• Are people using my context-sensitive help?
• Are user searches effective?

Then, create segments to help answer those questions.