Lab Report: Bounce Rate

Our lab site, FreshPlans, has always had a high bounce rate. We haven’t been concerned about that, for several reasons:

• A site with daily posts may get people arriving each day for the new post and leaving again.
• An informative site may have people coming directly to the info they want by search, getting that information, and leaving happy. With our particular population, we know, they are likely to print the page for future reference rather than reading it online.
• A site with affiliate links may have people clicking through to make a purchase.
• FreshPlans is our lab site, not a business, so who cares?

However, we do always have to say, “If this were a client of ours, we’d work on the bounce rate.” So we’ve decided to work on our bounce rate the way we work on a client’s site — that is, using data instead of guesses and assumptions.

We found out some interesting things.

First, we went to Google Analytics and sorted our pages by bounce rate. You can do this in the Content section, clicking on Pages and then on Bounce Rate in the results. This will sort your pages so that the pages with the highest (or lowest, as you prefer) bounce rate are at the top.

We saw right away that we have thousands of pages with 0% bounce rate — searches and translations, mostly. We also have thousands with 100% bounce rate, largely searches for things we don’t yet have on the site or translations which presumably showed the visitor that the page wasn’t what they expected. Filtering those results out, we found multiple pages with bounce rates ranging from 10% to 90%.

Our next step was to check the keywords that had brought people to the pages. After all, if the bounce rate is high, that should mean that people didn’t find what they wanted. Our Monkey Lesson Plans page has an overall bounce rate of 85% — but people stay there an average of 4:22 minutes. They arrive with keywords like “monkey lesson plans,” “monkey classroom theme,” “monkey bulletin board,” and “monkey themed classroom.” Four and a quarter minutes is long enough to get monkey classroom ideas, print them out, and be happy.

By comparison, our page of Christmas lesson plans has a bounce rate of 12%. People stay there an average of 35 seconds. This page  has a bunch of links to the various Christmas-related lesson plans at the site, so it looks as though visitors skim the page and click through to the lesson plan about Christmas Around the World or the Grinch or what have you. They’re also probably happy, but the bounce rate is much lower.

So far, this is about what we expected to see.. The interesting part came when we discovered that a lot of our high bounce rate pages were actually just images. We have a page, for example, that shows a Little Red Hen mask. It’s not really a post or page, just photo storage. People reach this rather useless page through image searches for “little red hen mask” and “little red hen printable,” and it does them no good at all.

Clearly, we should fix this problem and make sure our photos are stored differently, or at least that they clearly lead to the lesson plans they belong to. Should we also change our monkey lesson plans? Perhaps not. Over 5,000 people have visited this page, all reached it by searching for reasonable keywords, most stay there long enough to read it, and it seems to us that it contains useful ideas. Compare that with our engineering lesson plans, which enjoys a much lower bounce rate but has been visited by only a few hundred people. It also shows reasonable keywords and a reasonable length of visit, and we’re not convinced that it’s actually a better page.

However, if FreshPlans were a business site, we would want people to visit more pages, to explore, to shop. That’s what you want at your website. So here’s our advice:

• Use Google Analytics to identify the pages with higher and lower bounce rates.
• Check the keywords bringing people to the pages with higher bounce rates.
• Examine the patterns you see closely, and identify the problems with the pages that have higher bounce rates.
• Fix the problems you find.
• In the absence of identifiable problems, add links to other pages on your site, invitations to other parts of the site, and calls to action that will lure visitors to pages where they will shop, subscribe, or otherwise convert.

As for FreshPlans, we’ve lowered the overall bounce rate by three points so far by increasing the links to other pages on our site and reducing the links going out. We’re not convinced that this is an improvement in the quality of our site. Fixing the photo storage issue will, however, be an improvement, and we wouldn’t have thought to fix it if we hadn’t examined the bounce rate.