Bounce Rate is a Lie. Here’s How to Force it to Tell the Truth (for Webmasters only) 2018
Have you ever had a shockingly high bounce rate? I mean high — like, 90% or more? I see this all the time, even on my own site. But it doesn’t always worry me. Some people really panic when they see bounce rates over 50%. Most people assume that if a bounce rate is over 50%, it’s their fault. Their content or landing pages just aren’t performing the way that they should. But guess what? It’s not true. Or, at least, it’s not always true. There may be absolutely nothing wrong with your content, and you might still have high bounce rates. Don’t freak out. While keeping a low bounce rate on average is an important part of optimization, most people don’t actually know what a bounce rate really means for their conversions.
Most don’t even realize that a 90% bounce rate isn’t out of the norm for certain parts of your website.
In other words, your high bounce rate isn’t telling you the whole story.
What your bounce rate isn’t telling you
To Google, high bounce rates mean low-quality web pages.
If users are frequently leaving your pages without engaging in some way, it must mean your content isn’t relevant enough, right?
Well, yes and no.
A bounce is any “session that triggers only a single request to the Analytics server.”
If someone performs an initial action, like a site visit, but doesn’t interact a second time (by viewing another page, clicking a CTA, etc.) within 30 minutes, it’s considered a bounce.
Basically, it’s a “one-and-done” event.
A bounce rate is simply the percentage of those single-page visits. The formula looks something like this:
But here’s the problem: Google can’t tell the difference between a good interaction and a bad interaction.
Google only watches an interaction for about 30 minutes.
If someone came to your site and was scrolling around reading your homepage for 15 minutes, stepped away to check the mail, and then came back to click on another page… it’s still counted as a bounce.
It can also be registered as a bounce if someone is using multiple tabs because Google doesn’t know the difference.
Even if they buy from you (in one tab) and browse in another, if one tab is open and unengaged for 30 minutes it’s still a bounce.
You get a conversion, but Google doesn’t even credit you for it because it happened in a separate browser tab.
The other issue is that not all of your traffic sources are accurate.
Google often lumps different types of traffic together under “Direct Traffic.”
This means that some of your referral or email traffic gets bundled together with your direct traffic.
For example, Google now forces websites to use SSL certificates as a part of their ranking process. Many sites are now redirecting non-SSL traffic (from HTTP pages) to HTTPS pages.
This should be considered Referral traffic, which wouldn’t count against your bounce rate because there’s no real interaction to be made.
But it’s not. It’s considered direct traffic and automatically penalizes you with a bounce.
It’s really not fair.
When Groupon performed an experiment by de-indexing themselves, they found that 60% of their direct traffic was actually supposed to be organic traffic.
This misclassification of traffic can really hurt your bounce rate because you’re looking at “non-engaging” traffic and it might be designed that way.
If someone literally can’t engage with your website (because they’re redirected), but you’re not aware of it, it just looks like you have a really bad website.
Timing and traffic sources can hurt you.
But Google doesn’t know how to tell the difference between a good source of referral traffic that eventually converts and a non-converting visitor who is just browsing your site.
So what can you do about it?
The good news is that you can get Google to tell you the truth about your bounce rate as long as you know what to do.
1. Track in-page events, not page views
Because Google can’t predict the intention of a site visitor, you have to tell it what to look for.
Google needs to track both page views and events to determine accurate bounce rates.
Events track how a visitor is specifically engaging on your site. If they watch a video, sign up for a newsletter or click a link, it knows.
Events are important because they let you count interactions that don’t involve other pages loading (Google Analytics typically counts an interaction only if it opens another page).
But events require custom code on your website to track them properly.
To set up an event, go to your Admin Dashboard for your Google Analytics account and choose the ‘Goals’ link under the ‘View’ column.
Next, create a new goal and choose the “Custom” option and then go to the next step.
Give a title to your goal — something that will be easy to identify — and then be sure to select “Event” as the type.
You’ll then be given more options to customize your event.
Once you’re finished, you can save your goal. This will create a unique event that can be tracked on your website.
You will then have to input that code on your website, which may require the help of a developer or a website admin.
Code varies depending on the platform you’re using, so be sure you know how to do that.
Yes, all of this is an extra step in the process.
But if you’re worried about your bounce rates, creating events will give you a much more accurate look at what people are doing on your website.
Adding events to your pages will notify Google that an interaction has taken place so that the visit won’t be treated as a bounce.
2. Use inbound traffic segmentation
As I said earlier, one of the biggest problems with Google Analytics is that it can’t always tell the difference between direct traffic and other traffic sources.
There is a workaround to this problem, but it does take a little elbow grease.
The solution is to create different landing pages for different traffic sources.
Unbounce calls this “inbound traffic segmentation.”
The idea behind inbound traffic segmentation is that different sources have either “warm” or “cold” leads coming to your site.
If you create unique paths for all major traffic channels, you will be able to see which channels are driving or generating what.
Take a look at this example of a homepage, landing page, and ad optimized for Facebook traffic:
Or this one:
People clicking on the ad go directly to a conversion point (a sign-up).
These types of landing pages are more likely to have lower bounce rates because there’s already a level of buy-in when the visitor gets there.
They already clicked the ad, so they were curious at the very least.
Because Google has a hard time telling the difference between direct traffic and social traffic, this segmentation provides a bit of clarity that is otherwise absent.
And you might want to take into account the type of device used by your visitors.
If someone is using a mobile device, you want to optimize your landing pages for mobile, because bounce rates are typically higher for mobile users.
This might require tracking traffic sources over several months to see where the bulk of your traffic is coming from, and what devices are most used.
If you notice that your bounce rates are higher for mobile users, you don’t have to panic right away, because bounce rates will always be higher for mobile.
But if you notice that they’re high for desktop users as well, you’ll have a better idea of how to optimize your site to help visitors convert.
And that will give you a better idea of how to improve your content if you do want to make tweaks.
3. Test your site speed for “phantom bounces”
So what are your options if you don’t want to spend hours creating new landing pages and events in Google Analytics?
There are a few other ways you can test your website to make sure your bounce rates aren’t being jacked up for other reasons.
If your bounce rate is over 50%, one of the first things you want to check is your site speed.
Use a tool like Google’s PageSpeed Insights.
The faster your site is, the lower your bounce rates should be.
In one study by Google, they found that 53% of mobile ad clicks never resulted in a pageview when the page took more than 3 seconds to load.
In another case study, one brand migrated their website to a better hosting service and saw their bounce rate go from 50-60% down to 2-5% almost overnight.
In addition to the significant bounce rate improvement, they also saw other engagement metrics improve:
- Pageviews By Mobile Users saw a 132% increase (3,761 to 8,755).
- Overall Pageviews increased by 100% (8,416 to 16,849).
- Average Pages Viewed Per Session had a 113% increase (4.17 to 8.91).
- Average Pages Viewed By Mobile Users Per Session increased by 145% (3.27 to 8.05).
That’s some truly incredible differences from such a simple fix.
Before the migration, they attributed their high bounce rate to what they called “Phantom Bounces,” or bounces that happened because the site didn’t load in time and a visitor clicked away.
Google doesn’t know the difference between a standard bounce (they came, they saw, they left) and a phantom bounce (they didn’t see because there was nothing to see).
So if you’re not sure why your bounce rate is so high, test your site speed.
It could just be that making a small adjustment to your page loading speed drops your bounce rate enough that you don’t have to fiddle with the more complicated solutions above.
4. Delay the timing of your pop-ups
Google is a little vindictive about certain elements of your site, and may actually punish you for having them.
Pop-ups, for example, should be a good thing for your bounce rates.
They’re designed to give users a way to interact with your site, which should keep your bounce rates low.
But if they’re not designed in the way Google likes, your bounce rates might be affected without you even realizing.
Here’s what I mean.
Google has been putting a lot of emphasis on mobile-friendliness, especially with their new algorithm. This also means that Google will penalize sites that don’t properly display on mobile.
Interstitials — things like pop-ups, overlays, and modals — are seen as intrusive, and Google doesn’t like to rank sites that have a lot of interstitials on their mobile sites.
Here are a few examples of interstitials that Google doesn’t like:
- Standalone interstitials that must be dismissed before users can access your content
- Pop-ups that cover content and that users are forced to close to continue reading
- Deceptive page layouts whose above-the-fold portion looks like an interstitial
You can still have interstitials without the penalty, however. But it’s all about timing.
Like site speed, bounce rates tend to be higher on sites that have “immediate” or intrusive pop-ups, or those that appear in the first second or so of a page loading.
Pop-ups that are delayed by 3-5 seconds, on the other hand, tend not to be penalized by Google, and can actually improve bounce rates.
An alternative is to use exit pop-ups, which load when a user signals that they are about to leave your site.
Exit pop-ups tend to be less invasive and can help create last-minute conversions.
And according to Google’s Webmaster Trends Analyst, John Mueller, Google’s interstitial penalty isn’t triggered by exit pop-ups.
You might have to create an event for Google to properly track your exit pop-ups (see #1), but if it lowers your bounce rate, it’s worth it.
If you’re set on using pop-ups when a page first loads, though, then delay the timing of when they appear by a few seconds and test whether or not your bounce rates remain the same.
Also, be sure to test how pop-ups are affecting your site speed.
If you’ve gone to all the effort to optimize your loading speeds, and then your pop-ups ruin your progress, you will be right back where you started.
5. Take exit rates into account
Google isn’t just lying about your bounce rates.
Another area you want to pay special attention to when looking at your bounce rates is your exit rates.
Exit rates are calculated based on the last page a visitor was on before they left.
A high bounce rate will (almost always) increase your exit rate.
Let’s say you have an e-commerce site, and a shopper is clicking back and forth between Product A’s landing page and Product B’s landing page.
If they finally leave after viewing Product B for the 10th time, Product B’s landing page exit rate will be high, along with the bounce rate.
And Product A’s page will have a high bounce rate.
That’s because bounce rates and exit rates are calculated differently, even though they affect each other.
So what does this mean for you?
It means that you have to put your bounce rates into context.
If you notice pages that have abnormally high bounce rates but low exit rates, then you have to look at your user behavior, not necessarily page views alone.
You might have to pull a custom report to truly understand what’s happening.
Google won’t be able to tell you why certain pages are doing well.
If you want to know the truth, you have to look at your analytics within the context of your audience, their behavior, and the type of page they’re visiting.
Certain landing pages will almost inevitably have higher bounce rates than others, but in context, that’s not necessarily a bad thing.
If you’ve gone through all the effort of the above steps, and still aren’t seeing results, make sure you know your exit rates.
Because not all bounces are bad for business.
Don’t take Google’s word for it.
Google is smart, but it’s not as smart as you.
It can tell you that someone left your site, but it can’t tell you why. That’s why you have to do some digging.
First, be sure to set up event tracking.
Yes, it’s an extra step. But it can tell you which elements of your site are causing people to leave.
Next, create separate landing pages for different traffic sources. That way you can tell where your traffic is actually coming from and why certain visitors convert while others don’t.
And test your site speed. Seriously. It could be the reason your bounce rates are at 95%.
Finally, be sure to take into account other things that influence conversions, like interstitials and exit rates. Remember, it’s all about context.
Google won’t tell you that, but I will.
What changes have you made to your site that noticeably reduced your bounce rates?