Don’t Oversimplify Web Metrics

When viewing web metrics, many consultants, trainers, and professional coaches want to boil things down to good and bad. High counts of visitors are good, low are bad. Low bounce rates are good, high is bad. While some of these principles can serve as guidelines, page or user context is critical in analyzing how to improve a page.

A statement that illustrates this was recently made to us in reviewing a site’s metrics. During the review the client said, “My homepage is doing really well. It has the most hits by far on the site and the bounce rate is comparable to other pages. The time on page is really good; it’s over a minute and a half.”

Following the general principles this statement was true, high hits, reasonable bounces, and plenty of time on the page. But there had been several complaints that content was hard to find on the homepage. So where was the disconnect?

The disconnect was that the numbers were telling a more complex story that the general principles couldn’t illustrate. A homepage being the most popular page is common. This particular homepage had little content, serving primarily as an index to other more robust sections of the site. For that reason, the bounce rate should have been well below other pages because the primary purpose was to direct people to other parts of the site. Furthermore there wasn’t much to view or read on the home page so the time on page should have been quite short, likely under ten seconds.

The numbers were showing that people were struggling to find what they wanted and many were giving up after a minute and a half. Don’t rely on general guidelines when reviewing individual page stats. Rather analyze the purpose of the page to determine if the norms truly apply or if the numbers mean something unique.