What you’re about to read today on the Analytics & Site Intelligence blog here at MoreVisibility is something derived from a often-heard question in the web analytics / measurement world:
“How come the %Exit metric doesn’t say 100%? Everyone has to leave a website at some point, right?”
Measuring how visitors interact with your website – and how they consume your website – is the critical way in which you can refine and improve your website to provide an increased level of visitor satisfaction, and a higher profit margin.
Entrances, bounces, and exits are three foundational, bedrock metrics that Google Analytics uses in many different ways to help you measure your website’s effectiveness. As it turns out, there seems to be slight to moderate confusion about what these metrics represent, and what metrics like bounce rate and %exit really mean.
Let’s start by defining the aforementioned three foundational metrics:
Entrances – This is the number of entries by visitors into the pages of your website.
Bounces – This is the number of single-page visits by visitors of your website.
Exits – This is the number of exits from your website.
Pretty clear so far, right? Good. Now, it starts to get slightly trickier. How about we define another metric:
Bounce Rate – This is the percentage of single-page visits to your website. Bounce Rate is calculated by dividing bounces into entrances.
So, when you have a 60% bounce rate, 60% of your entrances left your website on the same page they entered from. In other words, they did not view another page. Also, when you have a 60% bounce rate, Avinash Kaushik’s head explodes. 🙂
Still with me? Great! Now let’s define the last of our metrics:
%Exit – This is the percentage of site exits from your website. %Exit is calculated by dividing exits into page views.
A-ha! You probably already knew or had an inclination toward what the first part of the definition would say, but then reading the second part, you start to see why the %Exit metric isn’t ever equal to 100%.
Now let’s take the final step.
Does everyone have to enter your website to view it? Yes. Does everyone have to leave your site at some point? Yes, they do. So, while the %Exit metric isn’t going to be 100%, the number of total exits should equal the number of total entrances.
A few images will help you understand everything you’ve been reading so far. I created a custom report in the new Google Analytics platform showing entrances, exits, page views, %exit, entrances / page views, bounces and bounce rate metrics. I also have the page as the dimension for my custom report. This first image shows the scorecard, which tallies up the totals for all metrics for the dimension that I have chosen:
As I expected, the number of total exits equals the number of total entrances (everyone has to come in to some page and everyone has to leave at some point). But the %Exit metric reads 58.92%, not 100% as you would initially think. So, I threw in the page views metric to give you some clarity as to how Google Analytics is calculating %Exit. 13,467 exits divided into 22,857 page views, all multiplied by 100% will give you 58.918%, which is rounded up to 58.92%.
However, viewing the very first line item of my custom report shows that the number exits does not equal the number of entrances for any given page, because not everyone leaves your website on the same page that they entered it from:
Here, the %Exit metric makes a little bit more sense, as it is not tallying up page views for all viewed pages on the site – only the /index.php page. 43.91% of all page views on the /index.php page resulted in an exit from the website from this particular page.
I hope that you were able to obtain some clarification and some deeper understanding of how some of the common metrics that you see in Google Analytics are tabulated and used by Google. Remember to always know what data you’re looking at – that is, get to be familiar with the way metrics are computed in Google Analytics (or, your web analytics platform of choice), as it will help you glean those insights that we all strive for!
Bounce Rate – the most popular two words in Web Analytics today. It’s become a cliche, a catch-phrase if you will. Everyone is talking about Bounce Rate and how good, how bad, how low or how high it is, and quite a number of folks have started to use Bounce Rate as an evaluation metric for success. I can safely speak for everyone involved with Google Analytics when I extend a huge “Thank You!” to all of you who have embraced it!
Interestingly, Bounce Rate is one of the only metrics in Web Analytics that we want less of. We want lower bounce rates, not higher, and fewer bounces, not more. A question I get asked at least three times a week by clients and co-workers alike is “How do we lower our Bounce Rate?” There are a lot of things that you can do, but there are only so many options that have proven to be effective over time. Today, let me share with you five different things that you can do – today – to start decreasing your bounce rate, by keeping your website’s visitors engaged with your website.
1. A “Higher” Call-To-Action
Have you ever heard the expression “Out of Sight, Out of Mind“? A persuasive and engaging call-to-action that is very low on a page, say, below the fold of a page, can cause visitors to lose focus and get distracted by your content / video / latest web 2.0 toy, which may cause the visitor to hit the back button or close their browser before visiting the next page on your site. No matter how nice of a call-to-action you have and no matter how attractive the offer or pitch may be, it needs to be highly visible to your website’s audience so that they can react (positively) to it and click on it, thereby lower the number of folks who bounce off of the page.
2. A Sync with your Ads and your Landing Pages
No, I’m not talking about N’Sync – I’m talking about a strong connection between the ads and the messaging you are using with the page that you are directing all of your future visitors to go to. One of the biggest factors that could be driving your Bounce Rates higher and higher is a mixed message that you are sending to your potential visitors. For example, if your ad copy says “15% Off!”, you need to make sure that “15% Off!” is the very first thing that a visitor sees when they hit your website. If you have “multiple sizes and colors available”, direct the visitor to a page where they can choose their favorite color and the right size. Using a promo code in your ad? Create a unique landing page and have the promo code appear right away on the page, so that visitors will feel the connection between your marketing message and what’s really happening on the website.
3. Improper Tagging on your Website Pages
A silent but very deadly killer, untagged pages of your website can only do your website harm. When some pages are missing the Google Analytics Tracking Code, visitors reaching those pages will have their referral cookie updated, thereby resetting information like “google / organic”, the campaign, and the keyword they used to reach you. At all times, when uploading a new page or section to your site, stop and make sure that the Google Analytics Tracking Code is present on your new page(s) first before uploading. This will save you a lot of head-scratching, unnecessary report ugliness, and will decrease your Bounce Rate, all at the same time!
4. Writing for your audience
Khrysti / SEO Team – I haven’t forgotten about you, because I am still writing “Content Is King!” That statement definitely translates to the Analytics side of things, and helps reduce your Bounce Rate. Use a combination of Google Insights for Search, Google Ad Planner and Google Trends for Websites to get an idea of the type of traffic that your website can receive, as well as valuable demographic information which could represent your future audience. Once you are comfortable with the type of audience and volume you expect to receive, write your website’s content appropriately and specifically targeted, so that visitors will feel a connection with what you’re saying. To use an exaggerated example, you wouldn’t want to talk about the fashion stylings of the cast of “The Hills” if your website sells motorcycle insurance (This, unfortunately, happens a lot on the web and it leads to a high number of bounces).
5. Testing, Testing, 1…2…3!
Finally, it’s essential that you incorporate some program of testing and experimentation on your website on a weekly or monthly basis. Each and every week (or few weeks), you should think about some element of your website or some element of an advertisement that you’ll want to experiment with, to see which version is the more profitable and successful one. Google Website Optimizer is a fantastic product where you can easily create as many experiments as you’d like, and see clear results in no time. You can also create a Website Optimizer experiment from start to finish in well under 10 minutes, which means you won’t have to be bogged down with hours of set-up and design time. Testing and experimentation with Google Website Optimizer is one of the best ways to decrease your Bounce Rate over the long-run, while sky-rocketing your conversion rates at the same time!
So there you have it – 5 great things that you can do today to start lowering your Bounce Rate, keeping your website’s visitors engaged, focused, and happy with you!
Starting today, chances are that your Bounce Rate is going to go up, and your Time on Site metrics will start to become more realistic.
Should you panic and freak out? Should you hide under the bed and lock the door to your room? Should you pause all of your campaigns? Obviously, you shouldn’t do any of those things, but I should explain what’s going on before you reach a state of dementia.
Google Analytics has changed the classification of the setVar function – Custom Segments that appear in the Visitors >> User-Defined report. Previously, whenever a user reached a page that was making use of this setVar function within the Google Analytics Tracking Code, Google Analytics would consider that what they refer to as an “interaction hit”. Interaction hits, like Pageviews, Events, Transactions, and Experiments with Google Website Optimizer are what Google Analytics uses to calculate Bounce Rate and Average Time on Site.
So, for example, let’s say a visitor landed on a page of your website, but left without visiting any other pages. That is, as we all know and love, considered a Bounce. But, let’s say that that same page was using the setVar function. Before today, that visitor would NOT have been counted as a bounce, because Google Analytics would have fired off two “interaction hits” – one for the pageview on that landing page, and one for the custom segment caused by the usage of setVar. However, from here on out, a user that only views 1 page of your site and leaves will be counted as a Bounce, setVar function or no setVar function.
This also has an affect on your Average Time on Site metric. This is calculated by Google Analytics by taking the time stamp of when the first pageview on a website occurs, and subtracting that from the time stamp of your second pageview on a website. Now, if you don’t visit a second page and you bounce, Google Analytics cannot do the math, because it has nothing to subtract from, so it reports a 0:00:00 average time on site.
Previously, because of our setVar friend, Google Analytics would be able to do math, because it would have the time stamp of that first pageview, AND, the time stamp of the setVar function firing off. Since these happened so close together, you could easily see extremely low average time on site numbers, like a second or two. Clearly this was confusing and didn’t make sense, which is another reason why the good folks at Google Analytics have decided to make this change.
When should I use Custom Segments / setVar?
There are a few good places to use this function. One place is on the receipt / “Thank You” page that a user sees after they buy something from your store. This way you can identify anyone who reaches this page as a “shopper” or “customer” or as “awesome”, or anything that you want to call people who reach this page. Then, in your Visitors >> User-Defined report, you’ll be able to do some analysis on this segment of people.
You can also use Custom Segments / setVar based upon an option they select on a “Contact Us” or inquiry form. Let’s say your form has a question that asks users to select between: “Executive, Director, or Marketer”. You can use setVar here to identify people based upon their selection, and analyze the behavior of each custom group of people.
So don’t call your pay-per-click manager or SEO engineer and proclaim that the sky is falling – chances are very good that you will not notice too much of a difference, if at all. If you don’t use Custom Segments / setVar, then you have nothing to worry about. If you use setVar on multiple landing pages and in different places on your site, brace for impact because your Bounce Rate is going up. But hey, think of it like this: it’s for the better – this is now a much more accurate calculation – a truer representation of your real Bounce Rate and Time on Site metrics.