Product managers care deeply about the engagement of their users. After all, engagement is an indicator that their product is providing value to users.

Of course, engagement can be measured in many different ways. One of the most important qualities of a successful product is that users find what they’re looking for. To measure this quality, let’s introduce a new metric to our tool kit: the bounce rate.

Definition of Bounce Rate

The bounce rate is defined as “total number of visits where only one page was viewed”, divided by “total entries to the page”.

Let’s make this more concrete with examples. Say that we’re looking at the following 5 visits:

• Visit #1: 2 pages viewed, user ID 123
• Visit #2: 3 pages viewed, user ID 456
• Visit #3: 1 page viewed, user ID 123
• Visit #4: 6 pages viewed, user ID 789
• Visit #5: 1 page viewed, user ID 468

To calculate the bounce rate, we can ignore the different users, and focus just on the number of pages viewed on each visit. Remember, not all data is relevant – in this case, user ID doesn’t contribute to the definition of bounce rate.

2 of the visits can be classified as bounces. Remember that we define a bounce as “a visit where only one page was viewed”. Therefore, we can calculate the bounce rate as:

2 bounces / 5 visits = 40% bounce rate

The key callout here is that every one of your web pages has a different bounce rate. When defining bounce rate, be sure to clearly define which page you’re using. Generally speaking, landing pages are most relevant for calculating and monitoring bounce rate.

This informative infographic from KISSmetrics clearly walks through bounce rate: definition, equation, ways that people can bounce, benchmarks, and tactics for improvement.

Insights on Bounce Rate

One of the key misunderstandings around bounce rate is that “higher bounce rates are worse.” That’s simply not true!

A high bounce rate isn’t necessarily bad – it just might be the way that users typically interact with those kinds of products.

Bounce rate is a mixed metric – that is, people can bounce in a negative way, but they can also bounce in a positive way.

For example, say that your feature is a signup form to a newsletter.

A visitor would bounce in a negative way if they came to your website from an ad, found out that your newsletter didn’t satisfy what they were looking for, and left your website without interacting with anything on the page.

A visitor would bounce in a positive way if they came to your website from a Google search, read about your newsletter, found it valuable, signed up, and then left the site to go to their inbox to confirm their email.

Therefore, overall bounce rate is not a sufficient metric to monitor. It needs to be considered alongside other signs of engagement.

The Opposite of Bounces

Just to round out our discussion, let’s talk about what the opposite of a bounce is.

If single page visits are called “bounces”, then let’s call multi-page visits as “sticks”. We can now talk about positive stickiness and negative stickiness.

A visitor would stick in a positive way if they came to your website, accomplished their key task, and found so much value in the rest of your website that they decided to click through to other pages on your site.

This positive stickiness generally stands out the most in e-commerce.

Just think about Amazon: when you come to the homepage, you see lots of different deals and product offerings. You decide to click through to another page on Amazon. Your visit is now a positive stick.

Its inverse is a negative bounce – you came to Amazon, couldn’t find anything on the homepage that interested you, and you left.

A visitor would stick in a negative way if they came to your website, got confused, and tried to click around to find more information. This generally happens in cases where a visitor has a particular expectation about what they’ll find, yet fails to find what they’re looking for.

Here’s a personal example.

I once tried to book a cruise for my friends. I came to a cruise website where I expected to be able to book directly on the homepage. After I read the homepage, I couldn’t find any sort of online form that would let me book.

Confused, I clicked into the “trips” page, where I still couldn’t find any way to book. I then clicked on the “contact” page, and still couldn’t find out how to book a trip. In frustration, I left the website. In this case, I had negative stickiness.

The inverse is positive bounce.

On a different cruise website, I saw the booking form on the homepage, booked it within the same page, and left the site immediately afterwards. I found the value I was looking for, accomplished my task, and paid for the product.

Making Bounce Rates Work for You

So now we have four different kinds of behavior: positive bounces, negative bounces, positive sticks, and negative sticks.

How might we make the bounce rate more insightful, so that we can actually determine whether users find your product valuable?

First, remember that you can always pair quantitative metrics with qualitative feedback and observation.

Set up user testing sessions where you ask a user to run through a real-world scenario, and see how they interact with your product or with your landing page.

What thoughts are going through their heads as they interact with it? What’s confusing? Did they achieve their objective?

Second, consider how you would define positive bounces, and how you would define negative bounces. Which actions do users take to get to a positive bounce? Which actions do users take to get to a negative bounce?

Here’s one example of a signal of positive bounce. An engaged visitor might read your page from top to bottom and find the value they’re looking for, then leave your website – so you can define a good bounce as “spent at least X minutes on the page” and “scrolled all the way down to the bottom”.

Once you know which actions define positive bounce and which actions define negative bounce, you can create new action-based metrics. Split your aggregate bounce rate into positive bounce rate and negative bounce rate, based on what kinds of actions they took.

Third, segment both the positive bounce rate and the negative bounce rate by user segments. Consider the various kinds of users who will be coming to your website, and the different personas. Who are you targeting, and why?

For example, Google Analytics enables you to break out by age, gender, and location. You can also segment by new visitors vs. returning users, browser, device, and acquisition channel. Furthermore, you can segment them by different affinities. As an example, you can see how bounce rate breaks down by Technophiles, Sports Fans, and Cooking Enthusiasts.

Once you’ve completed this exercise, you have a baseline for bounce: it’s split up into positive bounce and negative bounce, and segmented by your different kinds of users. The next logical step is to minimize negative bounce.

Note that I didn’t say that we should maximize positive bounce. The reason here is that you might want to optimize for positive sticks instead of positive bounces. But in either case, you definitely want to reduce negative bounce.

With your segmented bounce rates, determine which segments have the most opportunity for added value. From there, determine what kinds of user research to conduct or what sort of A/B tests to run.

Summary

Bounce rate is a mixed metric, which can be misleading. High bounce rates aren’t necessarily bad, and low bounce rates aren’t necessarily good.

That being said, bounce rate is used in many different industries, and many product managers and product marketers track at an aggregate bounce level. Bounce rate is a helpful tool in our tool kit, but it shouldn’t be used in isolation.

To drive more user value, view the world from the perspective of your user. What would be a positive bounce? What would be a negative bounce?

Then, track positive bounce and negative bounce separately and break them out by user segment. Look for the segments that have the most opportunity and the most value, and conduct A/B testing and user research to further optimize.

We’ve now covered four important key performance indicators (KPIs) that should be part of any product manager’s toolkit – bounce rate, net promoter score, conversion rate, and retention rate.

As data becomes easier to gather, product managers need to know how to make data-driven decisions to best serve their customers and their companies.

Let us know in our Slack community what metrics you’d like us to cover next!

 

---

Have thoughts that you’d like to contribute around bounce rate? Chat with other product leaders around the world in our PMHQ Community!