October 26, 2017 | Sarah Danks

How would you define bounce rate?

Ask that question of ten people who deal with any aspect of online marketing and you’ll get ten variations. I’d wager a guess each and every person you ask has a slightly different connotation (yes, I meant that) of what they perceive bounce rate to be. Not to say they’re wrong…

…but are they right?

Now, before your feathers get ruffled, consider that until I started researching this post, I had my OWN definition of bounce rate — and it was only *mostly* correct.

If you live in the world of websites, you know clients pretty much always want to know, “What’s my bounce rate?”

As marketers it’s our job to smooth the muddied waters and tell them

• what bounce rate actually is,
• that it’s always contextual, and
• there are more important KPI on which to focus.

Yes, I just said that.

So What IS Bounce Rate, Really?

Let’s back up for a second. Generally speaking, a bounce is defined thus: a user lands on a web page and “bounces” immediately off.

I mean, at its core, it’s as simple as that.

Okay, so let’s get a bit more technical: bounce rate is when a user lands on a web page and does one of three things:

1. hits the back button,
2. closes the window, or
3. types a search term or URL into the address bar without taking any other action (i.e., visiting any other pages on the site, filling out a form, etc.).

*yawn*

But, that’s just MY definition. And who cares what I think, right?

Since I spouted off a second ago about people in the industry having varying ideas of what bounce rate is, why don’t we clear that up first.

I asked ten colleagues in the online marketing world to define bounce rate in their own words.

Some are search marketers, some are designers/developers, some are strategic digital strategists, some are conversion rate optimizers and I even threw in a token web content writer, just to cover my bases!

10 Marketers’ Definitions of Bounce Rate

Here it is: a compilation of 10 marketers’ definitions of bounce rate. Unabridged, (mostly) unedited, and in their own words:

1. Jeff Sauer, Digital (Nomad) Marketer & President of Jeffalytics:
   “Bounce Rate is a way of measuring the number of people who left your website after viewing only one page. While the goal behind this metric is to give you an indicator of which pages on your site fail to draw in visitors, I find that bounce rate is a terrible indicator for success of a company. This is because of two reasons: if someone leaves your site after visiting a page, they may have found what they were looking for and left. Maybe that is a good thing. The other reason is that there might have just been a problem with your data collection. For this reason, I say don’t fret bounce rate and focus on generating revenue for your company instead. Focus too heavily on bounce rate and you might soon find yourself out of business.”
2. Paul Kragthorpe, Director of Technology at WebRanking:
   “Does anyone define it better than Avinash? I can’t! But, to me it’s going to a website, not finding what I want immediately, and leaving to find what I was actually looking for.”
3. Oli Gardner, Creative Director & Co-founder of Unbounce:
   “The percentage of people who leave your website without visiting more than one page or performing a conversion action. For a standalone landing page, bounce rate is inversely proportional to conversion rate.”
4. Angie Schottmuller, Growth Marketing Advisor:
   “Google Analytics’ meaningless, malarky bounce rate = enter and exit the same page. Legit bounce rate = enter and rapidly exit (<5 seconds) the same page.”
5. Paul Jahn, Local Search Marketer:
   “The percent of people who land on your website via whichever avenue, it doesn’t resonate with them & they then hit the back button or type in a different URL.”
6. Christy Arneson, Senior Marketing Manager at Amazon.com:
   “Bounce rate reports the number of customers leaving a landing page without completing a conversion action. Tracking where customers/visitors go after exiting can help reveal opportunities for optimizing UX.”
7. Julie Kosbab, Independent Consultant & Troublemaker-in-Chief at Betweenstations Digital:
   “Bounce rate is a measurement of your web site’s one-hit wonders.”
8. Kayla Hollatz, Copywriter, Brand Strategist and Owner of KaylaHollatz.com:
   “I’ve always defined bounce rate as the percentage of visitors who click out of your website after only viewing one page.”
9. Charissa Moore, CEO & Marketing Copywriter at House of Bliss:
   “Bouncing? That’s when people don’t have an emotional connection with your brand or business, and click on to your competition. Gross.”
10. Dave Dechant, Creative Director here at ThinkSEM:
    “When a visitor decides near immediately that your web page is sh*t and realizes how many better things there are to do in this stupid world.”

See, what’d I tell you? Ten different answers to describe the same term. Now, are any of these marketers’ definitions wrong? Heck no. Colorful, perhaps, but not incorrect.

But did you notice how emotion tends to creep into their “definitions” (i.e., they postulate WHY the visitor bounced)?

Taking emotion, opinion and conjecture out of the equation, let’s see what the technical definition is. According to The Google:

“A bounce is a single-page session on your site…specifically, a session that triggers only a single request to the Analytics server, such as when a user opens a single page on your site and then exits without triggering any other requests to the server during that session.”

Isn’t that what we all just said? Sort of, but:
a) Google’s definition is statistical only, and what makes this most interesting is
b) what they consider “request” to mean.

And why do we care what GOOGLE thinks bounce rate is, by the way?

(Insert zombie-like voice: Must do what Google says…)

Google-Defined Term: Website Interaction

Since Google is the most popular search engine, it stands to reason we should care what They think bounce rate is. Plus, with almost half the websites on the web using Google Analytics as their tracking platform…

…well, you get the point.

So, to get back to bounce rate: Google Analytics considers a single-page session with no interaction to be a bounce.

Big deal, right? Who cares if that’s how Google denotes a bounce?

Well, what about web content that takes a long time to read?

PS: Google Can’t Measure Scrolling…

Here’s the deal, folks: Google Analytics can’t “see” — and therefore track — scrolling. “So? Who cares?” you scoff. Well, I care, for one. As I’m sure all bloggers do.

Any copywriters of long content on the web should care — see, when you write a long blog post, article or any in-depth web page, it takes people a long(er) time to read it. They’re bound to be scrolling for a while. (You know, kind of like you’re doing right now…)

Stop, stop — if people are reading the content, spending time on the page, scrolling, that’s gotta count for something! Scrolling is engagement, isn’t it? Not according to Google Analytics.

So, if it takes a visitor longer than thirty minutes to read your post? Bounce. And/or if they don’t interact (as Google defines it) in any way with the content (even though they read it all)? Bounce.

Hm.

No reason to panic, right? Since said visitor spent thirty minutes on the page (or however long it took them to consume the copy), Google Analytics will have them logged as spending a long time on the page, even though they’re a bounce. Whew! Problem solved.

Right?

WRONG.

…Nor Can Google Track Time Spent on Bounce Pages

Since Google cannot tell without actual actions (e.g., clicking a link, playing a video) what the visitor is doing on the web page, they can’t accurately tell you that the visitor is THERE AT ALL.

So, a visitor that lands on your page but clicks no links or moves to any other pages before backing out is not only considered a bounce; they’re considered to have spent zero seconds on the page.

Even if they actually spent the better part of an hour consuming your content.

BOOM. Did your mind just explode a bit? Mine did when I first learned it. I assumed — how silly of me — that The Google knew how to track how long a visitor stayed on a page, regardless of activity taken (or not taken, in this case).

You can read all about how Google Analytics tracks visitor time spent on websites (explained very well by Google’s own Justin Cutroni) if you’d like to see the science (math, whatever) behind it all.

At this point you’re still not convinced bounce rate is something to get upset about. I mean, seriously, we don’t want people on our site who’re just going to bounce back out! The ones who, in the words of Avinash Kaushik, “came, puked and left.”

Even if they’re reading your content, they’re not DOING anything else, so what’s the use in beating a dead horse?

Keep in Mind: Bounce Rate is Contextual!

But, hold on. What about that long blog post? You know, the one Google tells us visitors aren’t “interacting” with? (Because it’s taking them longer than, say, 30 minutes to read it.)

Well, let’s say these visitors are finding your post via your Tweet about it. Then, they click on the link to it (Google tracks that there is a new visitor on the page). Since you can’t measure scrolling as visitor engagement (even though in 2017 we know it IS engagement), Google doesn’t know what your visitor is doing…

…even though s/he is gobbling up your 2,500-word blog post and scrolling like crazy, gazing at the imagery you have embedded throughout the content, laughing at your witty anecdotes, etc.

Let’s say said visitor ends up spending twenty minutes on your blog post, reading every word — but then, yes, s/he hits the back button and leaves. BOUNCE. Zero time on page. (According to Google)

Here’s why you need to keep bounce rate in context:

Why are these visitors bouncing if the content’s good enough to consume in full? Well, they came, they saw, they left — and then they went back to Twitter and retweeted your link. So, action was taken — just not on that actual blog post.

OR, let’s say visitors are reading all of the content — including all the comments — and then they leave a comment (yay! INTERACTION!), but this all takes longer than 30 minutes. Since Google’s “timer” for measuring activity ends at 30 minutes, guess what? BOUNCE.

What we’ll see in Google Analytics is a page that has a high bounce rate. BUT, while Google tells us this page is a failure (because no one’s spending time on it), what’s really happening is: visitors land on our page, consume our content, and then interact (whether it’s heading back to Twitter to RT our post to all of their followers or leaving a comment).

Ideally, yes, we’d like interaction with other pages on our site/blog, too — but hey, this isn’t a failure — it’s “free marketing”! Enter: context.

Yehoshua of Analytics Ninja explains bounce rate well here in this oldie-but-goodie:

“When bounce rate is looked at in a monolithic fashion without exploring the nature of the site or page types, it is quite possible to draw incorrect conclusions about user behavior on one’s site. Always keep things in context.”

(Which, by the way, is an extremely helpful bit of advice for life in general.)

So, when you see a page on your website with high bounce rate, don’t assume that’s a bad thing! It can be annoying, however, to have to delve into the analytics data for each and every page to separate the actual bounces from the non-bounces.

Want to “Fix” Your Bounce Rate Data?

It’d be nice if Google Analytics would separate out the visitors who come to web pages and bounce IMMEDIATELY off (as Angie Schottmuller pointed out above) from the ones who’re consuming your content.

Don’t forget, for Google a bounce = no time spent on site (even if visitors are spending massive amounts of time reading blog posts, longer content, etc) so you can’t track this yourself with the way Google Analytics code is set up.

If Google changes their search algorithm more than once a day (on average), why doesn’t their Analytics code move with the times, as well? Especially in today’s online world of parallax/infinite scroll sites, blog posts and endless social media feeds (think how far you can scroll on Facebook. Twitter. Instagram. Etc. etc. etc.).

You’d think they’d re-define how they measure bounce rate, since online behavior has changed and continues to evolve, instead of letting us try to separate the wheat from the chaff ourselves.

If you use Google Analytics and you’re interested in being able to track more delicate types of visitor interaction than what Google’s capable of, you do have options. Check out more of Cutroni’s brilliance as he explains how to set up advanced content tracking.

It’s an article chock-full of great information, including how to add code to track events such as scrolling, visitors reaching the bottom of the page content, etc. IF you do add code to your site that measures, for example, visitors scrolling on a page, Justin notes:

“As soon as someone starts scrolling I consider them engaged and not a bounce. So this event will drop your bounce rate. Also note that these events WILL change your time on site calculations. You should see time on site increase.”

With this code addition, you’ll be measuring interaction very differently than the current Google Analytics code does, so your data will drastically change — most likely in a huge cliff-like type of way.

Image credit: Raaw by Markus

Wait, I Need Special Code to Accurately Track Visitor Engagement?

Yeah, sorry. The workarounds for being able to track certain visitor actions involve more time, energy and — of course — special coding, but are worth it if you want/need to measure stuff that Google can’t.

Because, see, marketers are getting wise to the fact that just because Google’s code doesn’t measure it, doesn’t mean it shouldn’t be measured.

I’m not sure why at the end of 2017 we have to add “workaround code” to be able to track something Google should just be tracking anyway, but the point is, while Google doesn’t track “non-action” events, there’s code out there to allow you to do just that.

Bounce Rate, Schmounce Rate!

…why don’t we call the whole thing off? All of this controversy: How do I define bounce rate? What does it mean for my website? What’s the context? How can I work around Google Analytics code to actually measure visitor engagement?

I mean, you have to ask yourself one final question: Does bounce rate even MATTER???

Sure, it’s been a “KPI” you’ve been worried about for years. But honestly, does the sun rise and set on this metric? As Jeff Sauer pointed out earlier, “focus too heavily on bounce rate and you could find yourself out of business.”

Here’s a real-life example: take Justin’s Advanced Content Tracking post I mentioned earlier. I clicked on THAT post from a different article, read through it, grabbed the quote (and the URL of the page), but then hit my back button to get back to the original post (since there was no target=”_blank” code :)).

So, unless Justin has his own special tracking code added (which he most likely does — sorry, Justin, I didn’t have the time or patience to wade through your source code to find it!), Google Analytics will have counted my visit as zero time spent on the page. Yet I not only consumed the content (and learned from it), I’m sharing it here.

Pretty sure that’s a win. (Since we all know how important links are for SEO. Ha.)

But, to Google all of that is a “bounce”.

It’s so confusing! As Nick Eubanks wrote for Search Engine Watch:

“A high bounce rate can be indicative of a number of things but usually falls into one of two categories:

1) You’re acquiring the wrong kind of traffic to your page(s), or
2) You’re acquiring exactly the right kind of traffic to your page(s).”

Gah! It’s like bounce rate is the new Schrödinger’s cat!

If bounce rate can vary so widely in what the heck it means, is it irrelevant? Or, at the very least, should we even be paying much attention to it at all?

Now, before Oli bundles his undies about all this — I’m too late, aren’t I??? — we have to discuss when bounce rate (as Google defines it) matters. Because up to this point, it’s been all about blog posts and websites with lengthy content.

Why Doesn’t Google Change The(ir) Definition of Bounce Rate?

Because of their agenda, of course.

Through all of this jibber-jabber you MUST be thinking: all this is bounce rate according to Google. But, if Google’s all about staying on the cusp of web user behavior, why haven’t they changed the way they define/perceive/track bounce rate in today’s scroll-heavy world?

And, technically they have; a couple years ago the definition was “…the percentage of single-page sessions (i.e., sessions in which the person left your site from the entrance page without interacting with the page.).”

Well, it’s simple: because their bread-and-butter is paid search.

Aha. See, on a standalone landing page built for the sole purpose of converting paid traffic (as Oli mentioned in his definition), a bounce is a much, much bigger deal. Surely when you’re paying for traffic you don’t want visitors bouncing at all.

In this context, Avinash’s definition (“I came, I puked, I left“) holds true. Or does it?

Since PPC landing pages are built (rather, should be built) with a single goal — for a visitor to take some type of conversion action — the fact that any traffic bounces is bad. Right?

Wrong. Again, context.

Wait, what context? How can it be a good thing if a landing page that’s built to bring in paid traffic and convert it on the spot has a high bounce rate?

Consider this: what if the action item on the landing page (the goal) isn’t an “engagement” event? I.e., it’s not filling out a form or clicking to download a white paper. What if it’s a phone call instead? The visitor lands on the page, gets the information s/he needs, and calls up the company. Since that was the point of the page, there’s no reason to stay on it, so s/he backs out/closes the window/etc.

Is that a bounce? According to Google: YES.

*Sigh* Here we go again.

Okay. Not all landing pages fit into that bucket — *MOST* of them require visitors to fill out a form (or take some other type of “clickable” action) in order to be counted as a conversion. So yes, bounce rate of any kind is a bigger deal with paid marketing landing pages…

…but depending on what your actual goal is, the data can still be confusing.

Bounce Rate: It’s a Love-Hate Relationship

The take-away here is that bounce rate — no matter how you define it or which tracking platform you use — needs to be taken with the proverbial grain of marketing salt.

For those of you using Google Analytics to track blogs and websites with long content, be prepared to see a “high bounce rate” on your long-form content. Because if the only “action” a visitor takes on a page is scrolling, Google Analytics code doesn’t (can’t? won’t?) track that.

Right now, the only way to track that type of event is to add special code to your site. Which is a pain.

And, keep in mind if you do change the code to track scrolling and other events, your existing Time on Site will skyrocket, creating a massive historical event in your data. Why’s that matter? If it’s a client site we’re talking about, you can be certain they’ll sit up and take notice at any drastic changes in their Analytics data.

For those of you tracking standalone landing page performance (with Google Analytics or, really, any tracking code), bounce rate probably means more — but again, always look through the data to be sure you know what you’re actually looking at.

At the end of the web page — I mean the day — you don’t need to care about how Google defines it or how YOU define it…

…just make sure you know what bounce rate means for YOUR website.

And, above all else: KEEP IT IN CONTEXT.

For those of you who’ve stuck with me thus far, thank you (and for those of you who haven’t, well, you tuned out a long time ago and probably bounced anyway)! But, it’s not as if this is the ONLY thing “wrong” with Google Analytics — check out Andy Crestodina’s laundry list of 21 Inaccurate Traffic Sources.

And also: how much stock do YOU put in bounce rate???