Optimizing your website normally starts with something you want to improve. This might be how it looks, how it feels or how to interact with it. If you want to improve your business you know, that optimizing those elements just on a hunch might be risky. Your most important impuls for changes should be your visitors and in this article I am going to show you how to learn more about them using Google Analytics.
From vanity metrics to people
Opening the first page, Google Analytics presents us basic numbers on the dashboard, like the visits, the pageviews or the average time spend on our website. Those are in fact very interesting numbers, but chances are that they are not the most important ones you might need to understand how your business is really performing. Let me give you an example from our own projects to understand the meaning of this.
vanity and actionalbe metrics
If you are running or starting your own business, chances are that you have read the „Lean Startup“ by Eric Ries. In this case you might recall that the less important numbers some website owners tend to fall for are called vanity metrics. The opposite of them are called actionable metrics. Those are the numbers you should really observe.
Like vanity metrics the actionable metrics are different for each business. So you need to figure out what the important numbers are for you. What I learned though is, that numbers that represent people are very often a more reliable and actionable metric. Eric Ries confirms this by writing „numbers are people, too“. In the second part I am going to give you some more examples on that.
An example for a web specific vanity metric are page views. Our own platforms perform with a couple of million page views per month. Mentioning this to my pears normally results in a respectful astonishment. In fact, this number is not telling anything about the (financial) success of the business. First of all because of the simple truth, that most websites are for free and nobody pays to see the content and use the tools. We don’t even sell anything directly on them. If I would run a shop with this number of page views and users, I must be very dumb to not easily earn a lot more.
Anyway, page views are still important to us, because we sell advertisements on our websites and some of them on a cost-per-mile basis. But if you think about ad blockers or ads on mobile vs. desktop the ad impressions are a far better metric here.
User segments in Google Analytics
If you used Google Analytics before you probably remember unique visitors. This represents the number of real people who visited your website at least once, but hopefully a couple of times. Even though there was a lot possible combining the unique visitors with other metrics in custom reports, it still wasn’t it.
A while ago now, Google Analytics introduced user segments. They allow you to filter a specific report by users or visits with specific attributes or events. You can find the segments on most reports when searching for the blue circle in the head of the report and clicking on the arrow left to it to open the list of available segments. The custom segment is All Visits. Other interesting segments are Mobile Traffic or New and Returning Users. Go and play with those segments in the various reports and you will already learn a lot more about your users than you might have known before.
Still, the next step up the Analytics mastery is to create custom user segments according to your business needs. To do so click on Create New Segment above the opened list of available segments. You can now select your segments based on demographics, technology, behaviour and more attribute groups.
For starters, I created a user segment for every content type we have on www.word-grabber.com: word finder, blog posts, word profiles and word games.
If you are running a blog along with other content on the same website you might like the readers user segment, so I included a screenshot from the settings. Be sure to use Users as a filter. Those are now users who visited at least one blog post at all even if this happened a lot of visits ago.
Readers – users who read at least one article
I created similar reports for every content type mentioned above and also created some for users who have never visited a specific section.
Find your user segments
Creating a readers user segment for any website with editorial and non-editorial content makes sense. Still, this is just the beginning. Below you’ll find two examples of even better user segments I discovered playing with the already created reports.
I was most surprised when I discovered that there is an enormous group we totally missed. 1/3 of the visitors never used the main tools, but was using the „only supporting“ content. This is content we just build to either attract users to the tools, like editorial content or to make the tools more useful, like additional information on words. One might now optimize both content types to be even more useful without the tools or to better convert those users to the main tools.
There is the other group I would call converted visitors. Those are visitors who found our site through a blog post (blog post = landing page) and stayed using the tools. This user segment is visiting more often, creates more page views and stays longer than the average user. This is what finally tells you that all your efforts to create great content to make people aware of the cool stuff you build pays off.
How user segments help to improve your website
Understanding your users is the first step to improve your site. The next is to optimize those metrics. The HOW to optimize is again a very unique question for every website. Still, I am giving you an example on a rather tricky optimization we are preparing and the ideas and challenges we face.
In the beginning of our word search tools, people often came to us asking about the meaning of a specific word. This showed us that we needed to create some kind of dictionary where people could look up the word definitions themselves. We started with a simple link to other ressources and added more and more information to keep people on our site.
Now we are asking ourselves if there is more content needed for the purpose of our users and how we can optimize the existing word information.
What are the right metrics?
If you want to make a fact based decision you need to test the right things. Without asking at least a couple of thousand visitors in a costly poll, what other metric could tell us that the word profiles are getting „better“?
A longer time spend on a page might be a misleading metric, because the better they find something, the sooner they might leave.
The same counts for the bounce rate. It is a nice goal to lower the bounce rate and help visitors find other information on your site two, but that still doesn’t tell you if the information shown is useful to them or they left disappointed.
Instead of the two metrics mentioned, we concentrate on the following three to measure if our optimization was successful.
Returning visitors. What does a happy visitor? He comes back. So we are going to measure the number of returning visitors before and after the optimization. This will only include visitors with word profiles as their landing page.
More page views from tool users. We also expect an impact on our loyal users. If they notice that the word profiles are worth clicking, they might create more page views.
Less clicks on external sources. The more useful a word profile is, the less people click on the external links to find more information.
Beside statistics, we think about getting feedback from users with a contact form and a poll. The problem with this is that those interactions are done by just very specific groups and it is hard to apply those facts on all of the visitors.
Another more and more popular method in testing are heat maps. They more or less measure where the users actually look or click most. This might be interesting later to get an idea of which part of the new content is used to reorder the elements.
The great thing about user segments is that Google Analytics has a gallery for them where everyone can share his segments. You should try some to get an idea of how they work, even if they don’t totally fit your own website specifics.
This was just an introduction where I shared how to create user segments and what you can do with them. Feel motivated by my insights to start your own experiments. And don’t hesitate to add something in the comments below.