3 Steps to Decode Your Top Google Analytics Questions

What are the most common google analytics questions we get? Usually, why is my “bounce rate” [fill in the blank]? And a few other “bounce rate” associated questions. Many analytics experts say Bounce Rate is the most misunderstood of all Google Analytics metrics.  And I hear more discussions about Bounce Rate than other more relevant data points, such as conversions!

Other Google Analytics concepts are misunderstood as well – such as average time on page. There is a lot of confusion around the meaning of different data points and how to interpret them in relation to one-another. For example, here’s a question we got recently:

Blogs with high avg. page times also have high bounce rates? Those data points seem to conflict. Any ideas?

So let’s clear all this up. But for that, we’ll need to start from the beginning. And here’s why:

Bounce rate and other Google Analytics measurements are just formulae. To be able to understand and interpret a measurement such as bounce rate, and other Google Analytics metrics we need to know three things:

1. The technology
2. The formula
3. The contexts

1. The Technology: How Does Google Analytics Work?

The reason why we need to start with “the beginning,” is that 90% of the time when we troubleshoot a tracking issue, the cause is improper google analytics code installation. So our first step is always to check the code.

The collection of data is dependent on properly implementing the technology and having the right environment for the tracking mechanism to activate.

So, to interpret the data, you need to understand the tracking mechanism, and what can affect its ability to do its job.

How does Google Analytics track your website visitors’ activity?

First, you need a Google Analytics account. Once you set that up, you’ll get a snippet of Java code to add to the code of your website. If the code is installed properly, the mechanism through which Google collects data is activated.

There are three sources of data:

• The HTTP request of the user
• Browser/system information
• First-party “cookies”

The first two provide browser and computer data, such as what operating system your visitor uses, and what region they are in.

The first-party cookies are used to obtain user session and ad campaign data. For example, the data used to calculate bounce rate comes from these cookies.

These cookies are what makes or breaks most of the marketing-relevant tracking. Without them, GA may know you had a visitor but not where this visitor came from, or what actions they took. In a minute here, we’ll talk about how the “health” of these cookies can impact tracking and data accuracy.

What if my code is not installed properly?

Google Analytics has a validation tool through which it will tell you if your code is detected. But there are instances in which the code appears to have been installed correctly, yet you’ll still notice reporting issues.

There are many many possible reasons for this. Other code on the page may break the GA code, or you have duplicate analytics code, or you’re trying to do a cross-domain installation or other advanced tracking which may require additional code than just what Google provides out of the box. A few installation issue red flags are:

• Zero data overall or on just certain pages
• Your own website showing as a referral to your website
• 100% bounce rates, may indicate some pages are missing the code

What else can affect Google Analytics’ ability to collect data?

Have I mentioned, there are many reasons why Google may not be able to collect data or collect incorrect data?

• Javascript errors on page – if the javascript code doesn’t execute (fire up)
• Using a browser with javascript turned off by default, which doesn’t allow the code to activate
• The visitor’s browser doesn’t accept cookies, which disables user session tracking
• The visitor uses a firewall or other method to block and delete cookies
• Your visitor deletes cookies manually (research shows 30% of visitors manually delete cookies monthly)
• Your visitor uses an opt out / do not track browser plugin
• Browsing using the “incognito” feature of the browser – will track session data but will expire the cookie upon exit, so when that visitor comes back it will look like they are a new visitor
• In-app browsing (when you click on a link in Facebook, by default it will take you to that website within the app itself, not by using your phone’s internet browser)- there are commonly reported cookie issues with mobile apps’ browsers

In conclusion, Google Analytics data is imperfect. But understanding the limitations can help you evaluate your results fairly. Some have dubbed google analytics reports “surveys,” because they don’t capture 100% of the data but they give solid directional guidance.

So now that you know how GA tracks data, let’s take a look at some of the most discussed performance indicators.

2. What’s the formula?

What is Bounce Rate?

According to Google, “Bounce Rate is 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).” It’s important to note that bounce rate is a measurement associated with a specific page, not the website. You can calculate a site-wide average but that won’t mean much unless you have a 3-page website.

So let’s say I see a post on Facebook and click on the link. I will land on Page A of the website.

I read, and then I close the browser window or tab. In other words, I leave Page A. That’s a bounce.

If I stay on that page for 20 minutes and then I leave, it’s still a bounce.

Let’s say I go to Page B first, then click over to Page A. When I get to Page A I decide I hate it, and leave the site in three seconds. Well, that’s NOT a bounce for Page A.

A bounce is when someone leaves the site from the same page they entered, regardless of how long they stay on that page.

Exits vs Bounces:

So, how can a high bounce-rate page have a high average time on page? Well, this question compares apples to oranges, because the key to the answer is in a different measurement: Exits.

Average page time is calculated as the total time on that page divided by Pageviews of that page minus Exits from that page. Bounces are not anywhere in that formula.

Average page time =    total time on that page  / (Pageviews of the page – Exits from that page)

So how are exits different from bounces and thus how is bounce rate different from exit rate?

A page’s number of exits is the number of times some leaves the site from that page. A Bounce is an Exit, but an Exit is not necessarily a bounce.

Example: Let’s look at bounces and exits for our fictional Page B.

• Joe comes to Page B from a Facebook ad. He reads Page B and leaves. That’s a bounce and an exit.
• Jane comes to Page B from Page A of the website. She reads Page B  and leaves. That’s an exit.
• Mary comes to Page B from Page A. She reads Page B and clicks on Page C. That’s neither a bounce or an exit for Page B.

So our page had 1 bounce and 2 exits out of 3 visits. The bounce rate of our page is 100% (one instance where the session started on page B and ended with page B), and the exit rate is 66% (3 visits to page B, of which 2 exit the site on page B).

Average Time on Page:

Again, average time on page does not use Bounces in its formula, but rather Exits. You CAN have a page with high bounce rate and low exit rate.

But let’s start with an interesting and very important fact – Google will record the average time on page for a bounce as 0. Even if the visit is 20 min, Google Analytics will think that visit was 0 min. That’s because GA only records the time after you left a page to visit another page of the website. Even if that’s not a bounce, but rather the last page visited, if you exit the site from that page GA won’t count the time you spent on that pre-exit page. So Average Time on Page is HIGHLY inaccurate.

So theoretically,

Average page time =    total time on that page  / (Pageviews of the page – Exits from that page)

But given that GA doesn’t count the length of the exit visits, Avg page time is more of a measurement of the average time spent on page by those who enter the page from elsewhere on the site and don’t leave the site from that page.

So if you have more of these long visits than bounces, your average time on page can seem high, especially when compared to Bounce rates. If you notice your page having a high bounce rate but a low exit rate, you should expect the avg page

time to seem high.

Or not at all, because if it’s an intermediary page in a process, maybe the avg time will be 10 seconds. Which leads us to the third consideration.

3. What’s the Context?

What the data means, or how we can interpret it, depends on your marketing context. Marketing context is a short name for a big bag of variables – your goals, your user experience, your desired path to conversion and more.

What should I worry about when it comes to Bounce Rates?

A high bounce rate can be a no-issue if you want people to “grab and go.” Is this a lead-magnet page, where the visitor comes and downloads materials and then leaves? Well then don’t worry about bounce rate and focus on how well that page converts a visitor to a download (and capture that email address, emails still has the best ROI). Or a “where to find us” page. The purpose of such a page is to take the customer off the website, so don’t worry about bounce rate.

For blog post pages in general, a high bounce rate is not necessarily a bad thing. It’s possible your visitors are reading, getting their fill and moving on. You’ll also notice that bounce rates tend to be lower for repeat visits, so focus on how you can get them to come back.

There will be pages that have 0% bounce rate where that means absolutely nothing. The “Thank You” page will be one of those, because there shouldn’t be any people coming directly to that page and then leaving. The “Thank You” page always has a precursor, which disqualifies it from the bounce measurement.

When should you worry about bounce rate?

When the goal of the page is to convert the visitor to an action that would take place on another page of the website. So if your page is Step 1 of a conversion process, yet the Bounce rate is high, that page may not be doing its job to move the customer to the next step. Your home page may be that Step 1 page. Or maybe a products or services page is that starting page.

How about exit rate?

The funny thing about exits is that EVERYONE will eventually exit the website. So again, the context matters. Are 90% of your visitors exiting on the “shopping cart” page? Well, then you may have a problem. But if they exit from a blog post, again, it may be a no-issue.

And Avg. Time on Page?

I’m not a fan of this measurement, because of the limitations I described earlier. The only time it’s worth investing energy into worrying about this measurement is when you see wild variations after a period where the numbers seemed steady.

If you’re trying to estimate how “sticky” your site is, use Average Session Duration instead.

OK, so now, when faced with Google Analytics questions, go through the three steps:

1. Could technology be impacting these numbers and how?
2. How is the data calculated? How will understanding the variables in the formula impact my interpretation of that data?
3. How do the numbers play in the context of my goals for this page, my users’ site experience and more?

Or just shoot me a note and I’ll help out, because I’m a data head.