Bounce Rate – The Complete Guide

In Blog by Vladimir0 Comments

Bounce rate is one of the simplest, yet also one of most misunderstood metrics in web analytic today. If I had a penny for each time I got asked about bounce rate and/or heard something that was entirely wrong, well… I’d have a lot of pennies. Like, a jar full of pennies.

Hence I decided to assemble this article about the most common questions about bounce rate, why is it important and why you shouldn’t care about it, how to properly measure the bounce rate for your website and decide if it requires improving or not.

What is Bounce Rate?

The simplest definition: a Bounce rate is the percentage of the visitors who come to your website and leave without visiting a second page.

This sounds very straight forward. Avinash Kaushik, one of the best known web analytics of our time, “redefined” the bounce rate as following: “I came, I puked, I left.” There is a lot of truth to that, and if we were to take this one, single statement out of the context, we would be right in assumption that bounce rate is an inherently bad metric, and the higher it is, the worse the situation. But here is the deal – this interpretation is relative since it depends on a huge number of other circumstances.

There is no such thing as a “good” or “bad” bounce rate on its own; it must always be interpreted in relation to other available data!

The (highly simplified) philosophy behind the bounce rate is that a visitor came to your website, saw that it was unsuitable for him and decided to immediately leave. The “yuck” part, indeed. What are the major reasons your website could have a high bounce rate?

  1. Low quality. The searcher could not find a good answer for the query on your page.
  2. Bad user experience. The website was so badly designed that the user screamed in horor while frenetically hitting the “X” symbol.
  3. Technical problems. An extremely high bounce rate more often indicates in technical problems of the website or tracking issues.

But what actions actually count as a “bounce”?

  1. Hitting the “Back” button (without visiting a second page on the website one has landed)
  2. Turning off a browser or tab 
  3. Entering a new URL or hitting a bookmarked link
  4. Clicking on a link on your page that leads to an external website.
  5. Leaving a browser window or a tab open when going AFK (away from keyboard), causing the session timer to expire after 30 minutes (default duration)

Out of these, the “back” button seems to be the most common thing the users do, even though that specifically does not apply to myself. SIDENOTE: I always have dozens of browser tabs open at all time over several monitors I work on. The moral of the story? Do not make conclusion how other behave based on your own!

Mythbusters – why bounce rate doesn’t matter!

Let me jump straight into the core of the issue and explain to you why a bounce rate is not a metric that is decisive or even important on its own!

Let us observe two searches, one with a very low bounce rate and the other with a fairly high bounce rate.

The first search I conducted was something simple and extremely popular: “SEO guide for the beginners”. It is no surprise that in results the first entry was a Moz Beginner’s guide to SEO, followed by a Kissmetrics SEO guide and eventually Wordstream SEO basic guide. (I can recommend all three as highly valuable and well organized guides!). As soon as you land on the Moz guide, you will see the entry page, the printable PDF version of the guides as well as the link to the first chapter. The landing page is simple, easily understandable, and the short intro does everything to convince the visitors that they are in the right place:

New to SEO? Need to polish up your knowledge? The Beginner’s Guide to SEO has been read over 3 million times and provides comprehensive information you need to get on the road to professional quality Search Engine Optimization, or SEO.

It is simply impossible to assume that this is a page that will have a bounce rate of 90%! (Remember, even downloading a PDF file, registering for newsletter or any other interaction with a website is considered an action that will make that visit not count as a bounce.) I can’t peak in their respective analytics, but it is a safe bet that these websites have a “good”, very low bounce rate, since they provide the exact information to the audience searching for those specific keywords.

Let us try the second example, and this time, we will make a very straightforward question: What is the capital of Canada? (I won’t blame you if you don’t know.)

Wikipedia will tell us right away that Ottawa is a capital of Canada,  and you probably won’t even have to click the link, but let us assume that you actually did and visited the page. Once on the Wiki, you might fly over several sentences, check out an image or two and then leave that page.

Effectively bouncing.

Now imagine how many queries a Wiki has where people come to see and read a single entry and leave afterwards?

Recall on how many websites you found a solution for your problem (how to do something in Excel, or fix your phone) and left without checking out any other articles? Dictionaries, weather forecasts, the best gear for your favorite class in World of Warcraft?

All these websites have a very high bounce rate… and absolutely no reason to be concerned. Their content is good, highly visible in search engines and has good overall performance – with the “exception” of that one little thing, the high bounce rate. Which leads us to the next important question, one which strikes fear into the hearts of many website owners:

Does Bounce Rate effect my website’s ranking?

This is the core of all fears related to the bounce rate; the idea that Google might be checking our bounce rates and demoting our website’s ranking when they find that a website has a high bounce rate (whatever that meant).

Let us quickly revisit the point we made in the previous part, when we compared websites with a high bounce rate to the websites with a low bounce rate. If the conventional “wisdom” that high bounce rate is bad would hold, then all the dictionaries, wikis and similar websites would fall down in rankings (imagine the chaos that would ensue on websites such as Yahoo Answers or Quora!).

So what the heck does the bounce rate mean, if it means anything!? How can I use that metric to gain meaningful insights?

We return back to the relationship between data. No data on its own has a meaning; only the data that are in correlation to each other can be interpreted correctly and provide us with actionable insight.

Before everything else, we should be aware of the fact that bounce rate depends so the type of websites and the industry the belong to.

One very simplistic way of determining how good or bad the bounce rate is would be the following:

  1. Up to 40% bounce rate is considered an excellent bounce rate
  2. From 40% to 55% we have a good bounce rate, which some consider an average
  3. From 55% to 70% is a slightly higher than average (could be considered bad)
  4. From 70% and above is being considered a very high (very bad ) bounce rate

Frankly, this means little to nothing. Yes, I would be worried if your bounce rate is 95%, and would be content if a website has only 15% of people leaving without clicking on anything else, but I have seen news portals with a bounce rate of 85% and above performing incredibly well (both in terms of ranking as well as revenue terms) and I have seen websites with a 10% bounce rate that had very low rankings and no profitability to speak off.

So, what IS a good Bounce rate!?

Average bounce rate for websites per industry would be as follows (thanks to Kissmetrics!):

But I digress (ever so slightly).

It has become apparent that it would make no sense to compare websites from different industries to each other, so we are certain that Google isn’t doing that (once again, think of all the websites that have a crazy bounce rate and yet rank incredibly well for their specific keywords). But is Google using a bounce rate as a relevant metric to determine how well your website will rank in SERPs?

In short, the answer is a resounding “NO!”.

Google has repeatedly said that they consider the bounce metric to be “spammable”, “noisy” and even “that some site are expected to always have a high bounce rate”. It is also very prone to manipulation, as I will demonstrate shortly. But Google doesn’t need our data to rely on, which we can manipulate, since it has a sufficient other sources to draw knowledge from (Chrome, Android, Google toolbar data, which they now openly admit to use to help calculate rankings).

Personally, I think that Google might be using a bounce rate to some minor extent in certain cases, but it is by no means as important as the majority of webmasters assume and fear. I personally never made a huge fuss over bounce rate nor do I worry about it significantly so that I actually attempt to reduce it on any of my web projects.

Let us return to another question that comes hand in hand with this one.

Should I worry about my bounce rate?

No – but you should use the bounce rate as a significant signal to improve your website’s functionality and performance.

If your website has a very low bounce rate, somewhere up to 20%, the chance is good that your tracking system is not properly configured (read: broken) or that your website in general has technical issues.

If your landing page has a huge bounce rate, you need to check into your channels that are leading traffic to it. You might find a specific channel that is the culprit and brings an excessive amount of traffic that doesn’t convert for you (social media or badly adjusted AdWords campaigns).

Radical drops in the bounce rate over relatively short time periods are sign that your website is experiencing some technical issues. There is also a scenario known to happen where the tracking would work badly, and visitors would be dropped between their landing page and the next page, which would also result in an increased bounce rate.

But personally, I would use the bounce rate as one of the best metrics during A/B testing, where checking the different versions of a landing page to discover the behavioral patterns of the audience. It will tell you if your adaptations and changes are moving in the right directions, but only if you combine them well with other permanent metrics, such as time spent on website, pages per visit and conversion rate.

What is the difference between bounce rate and exit rate?

Simply said, a bounce rate is a percentage of people leaving your landing page without any further interactions with your initial landing page; exit rate is the percentage where the visitors are leaving the entire website regardless of where they have entered it. These two are being confused often enough that I thought it would be important to state this detail as well. Once a person enters a website, it should be led to the final customer funnel (for a registration, purchase or something else). These funnels will have their weak spots, and some of them (registration page, checkout page) will have big exit rates. It is important to follow these metrics and see if you can derive an understanding why this is happening as well as how you could prevent it.

Is Bounce Rate important metrics for my AdWords campaigns?

Personally, I have always used a bounce rate in my AdWords overview and tried to derive as many actionable insights from those reports as merely possible. The first telltale sign that something is wrong with your paid advertising campaigns is often enough the quick rise in the bounce rate – and I have a perfect example for it!

During my time where I was running an AdWords account for a retail online store, one of the ads we had running advertised a specific Samsung Tablet in our offers. The campaign was set, and released early in the morning. When I checked upon it several hours later, I was amazed to find out that the bounce rate was increased to a staggering 98%! (Needless to say that there weren’t any conversions at all)

I immediately rechecked everything in the campaign; ad texts, keywords, placements, audiences, anything that might give me a hint why this entire story performed so badly. But everything was OK, the price, the keywords, negative keywords were in place, the product searched for and presented was the same, no significant price differences between various providers… I was clueless. I turned off the campaign and shrugged it off.

It was only per chance that I saw a TV commercial of a local telecommunication provider with a special offer that included that very same Samsung Tablet (along with the Smartphone) for the new subscribers. Naturally, the interest peaked and people went looking for the specific product; by entering the same keywords as defined in my AdWords setting, they arrived to the device in question, but were in fact looking for the offer from the telecommunication company. Situations like these happen quite often, so one needs to be on top of the game if he or she want to drive high quality leads for their services and products.

How can I change my bounce rate?

Next to the old practices of optimizing the user experience and removing annoying pop-ups, redirects, take-over ads, streamlining the experience and custom-tailoring the messages for targeted audiences, there is a way to make the bounce rate work for you.

This is how it works; essentially, the last session will always be counted. If a user comes to a landing page, spends on it a minute, 5 minutes or stays there over a 30 minutes (effectively letting the timer run out), the analytics will count these as a visit with the duration 0. GA can’t tell how long a user was actually reading a page; it can only tell how much time it spent on a previous page after he/she switches on a new page. When in doubt, it will count a 0.

What you as a Google Analytics power user can do is to set an event for the specific landing page. Once event fires, the visit will not be counted as a bounce. Effectively this means that you can set, for example, a time limit of 60 seconds to trigger an event, or scrolling until a specific part of the page. Imagine this, if a user comes to your website, spends there mere 5 seconds and decides to leave, well, that is a signal that the website quality is truly low. However, if a user spends a minute or five minutes reading a specific content, we can say that, from our point of view, his visit shouldn’t be counted as a bounce even if he leaves. By the way, this is not cheating, it is a perfectly legal technique, since the bounce rate needs to provide the analytic feedback to the webmaster, and he/she set the rules of what is useful for his/her content.

Conclusion

Hopefully you will find the topic of a bounce rate slightly less demystified after this article. Should you find any omissions, have a different opinion or would like to ask any questions, please feel free to do so in the comment section!

Leave a Comment