Google Analytics is without a doubt the single most popular digital analytics tool of all time. Full stop.
While it’s impossible for anyone to give the exact number, according to many different sources it’s used by over 50% of websites on the internet. And because about 30% of the remaining half is not using any analytics at all, it makes Google Analytics the dominant market leader.
There is no other digital analytics product which gives its user so many powerful features and is also free. It becomes even more powerful when synced with its cousins from Google Marketing Platform, Tag Manager, or Data Studio. Not to mention AdWords (Google Ads)…
So it shouldn’t surprise anyone that for many companies today Google Analytics is a no-brainer when it comes to building the foundations of a data-driven marketing environment.
Like every engine, Google Analytics too has to be taken good care of. It can, and should be serviced and optimized. It’s like an insight motor that runs on data fuel. So if you want to make sure that your engine produces maximum energy and minimum fumes, this analytics tuning is something that should be done quite often. Especially if the quantity of data being processed by Google Analytics increases in the course of time.
Remember, analytics is just a tool. It is your company’s or team’s individual setup that should help you do things, make decisions, let you understand, measure and above all utilize all the insights you may obtain from data. You’re the driver and it’s you who should have confidence behind the steering wheel of the analytics machinery.
If you’re reading this article, you probably suspect that your current Google Analytics setup could be better or (what’s also possible) you’re not using it as often as you may think you should. And you’re probably right! But you know what? You’re not the only one!
I’m a UX consultant. My team works with growing companies helping them to build a data-driven culture for sustainable and organic growth. We usually start with a Google Analytics audit.
90% of the accounts we work on could be optimized. To say the least.
And there’s nothing wrong with that! As I mentioned above, if GA is not being serviced and tuned it may stop producing valuable insights or, the worst case scenario, stop being used at all. And it happens quite often.
If this is something that has happened to you, then this article is here to help! You’ve just made the first step in fixing what’s broken, so you’re about to become a part of the 10% of those whose analytics runs like the most efficient, ecological, insight-producing engine.
You may want to ask, if Google Analytics is such a great product, why does it need to be adjusted? Well, that’s a tricky one. As I mentioned, untuned GA is like an instrument or an engine, and in the default state it would and it will work, but different uses require different customizations or adjustments.
You can play a piano which wasn’t tuned, sure, but if your aim is to become a professional musician, you need someone to tune and customize it according to your needs. The more accurate you need your data to be, the more certainty you need.
Every business has its own characteristics and thus the Google Analytics setup should reflect that. So what might work for an ecommerce store will not necessarily be needed by a SaaS.
The following checklist will help you make the first 5 steps to improve Google Analytics performance regardless of your type of business — or help you ensure you get set up correctly in the first place.
So let’s go!
1. Set up the 3 views (Raw, Master, Test)
Google Analytics is not retroactive. There’s no magic undo button that when hit would fix the data collected, say, three months ago. Every change you make in Google Analytics can have an irreversible impact on how and what data is being gathered. This is very often the main cause of why people are afraid to change anything in GA setup at all. Sad but true. We wouldn’t want that happen to you, would we?
Luckily, there’s a very simple and easy-to-implement solution to that problem which will give you not only confidence in using Google Analytics comfortably, but also the ability to test, experiment and even undo without the fear of losing or affecting data. This solution is setting up the proper structure of Views for each Property you use.
It works like this: if you use only one View and if, for example, you set GA to filter (in this View) traffic from specific IP, this data will not be collected. At all. There is no way to undo this command to filter out some portion of data. It’s gone, whoosh, GA didn’t see it and held no record of it.
By the way, a Property is like a type or a class of analyzed object (e.g. your website) and Views are different collections of data sets related to this object. In general, you should always use 1 Property for each analyzed product as multiplying Properties for objects of the same type will only unnecessarily duplicate the same data collected and as a result slow down the GA or/and make people using it more confused. Unlike Views, you can and you should create 3 views for each Property.
This step is recommended for these Google Analytics users who have only one View. Most likely its name is something similar to ‘All website data’.
The best (and also recommended by Google) set are these 3 Views: Raw, Master, Test.
Let’s discuss each of them:
Raw
This is a view with no settings implemented, no filters, no parameter exclusion, no nothing. Everything (including the trash data). It is an archive that is used to collect all the data from the beginning of the tracking code implementation. It will be our safety net where we will be able to retrieve all the past data and gain some reference.
Raw, once set should be this way forever. Without a very good reason you shouldn’t change anything to its setup. Raw is raw. Point of reference. You can stick a Do not touch label on it. Trust me, Raws have saved many data analysts’ lives. I mean insights.
Master
This is a main reporting view with the most updated filters and settings – best to analyse traffic and gain insights. Any member of the team that wants to check up on the latest data, and make decisions, this is the go-to view. You ought to use Master as much as you can. However, you should be careful when changing anything to its setup. Remember? No undo button in GA. The label to stick on this view: Fragile.
Test
This is your Google Analytics sandbox. This view should be used for testing new setup features (like filters). If you need a new filter (e.g. excluding the traffic from an home IP address of your new employee), you can just implement it on the test view first. After the filter is assessed as correct, it could be implemented on the main (Master) view without risking its data. The sign: Playground.
OK, so where to create the new views? It’s simple, all you need is to follow these steps.
How to:
- Go to Admin panel
- In the 3rd column select View Settings (as mentioned, in 99% of the cases it will be named All website data view as it’s the one that Google Analytics creates by default for each new Property)
- Change its name to ‘Raw’
- Click Save button
- Copy that view
- Call the new view ‘Test’
- Confirm by clicking Copy view
- Now follow the next steps of this article and get back here when you’re ready
- Ready? Points 2, 3, 4 and 5 checked? Good so let’s move on: Go to the Test view settings and copy it
- Call this view ‘Master‘
2. Increase data accuracy with the lowercase filter
Have you ever typed a website URL with Caps-lock turned on? I bet you have. We all have, hundreds of times. But do you know that for Google Analytics CRAZYEGG.COM/SNAPSHOTS and crazyegg.com/snapshots are two completely different sites (resources) even though your server will most likely send the exact same resource to the browser? You’re right, that’s strange. But we’re not here to discuss why GA does it, but to fix the results of it.
All we need to do here is to set up a filter that will treat all addresses to be treated by GA as if they were typed in lowercase, but that’s not just it. The main reason why we’re doing it is to have all the data (e.g. pageviews, avg. session duration, bounce rate) related to specific resource (eventually it will be only one e.g. crazyegg.com/snapshots) in the same place.
Additionally, this could also help you with proper UTM tracking of your campaigns.
How to:
- Go to Admin panel
- In the 3rd column select Filters
- Click Add filter
- Name the filter ‘Force URL to lowercase’
- Filter Type: Custom
- Choose Lowercase option
- Filter Field: Request URI
- Confirm by clicking Save
3. Set up a Self-IP removal filter to increase data accuracy
I bet you and the members of your team are visiting your website after every major update to make sure all the recent changes have been applied and that everything works as it should. New blog post? ‘Hey guys, we have a new post on the blog! Make sure you’ve read it and shared it on Twitter’. Sounds familiar? If so, know that so called internal traffic may have a negative impact on the accuracy of your data. Such traffic will not give you any useful data to produce insights. It affects the so called clearness of data.
To remove the noise from the data regarding your users’ behaviour, visits, sessions, etc. coming from your internal traffic, we have to exclude it by setting the self-IP removal filter. It’s easy to do, but trust me, you’ll get rid of a lot of waste from your data by doing it. It’s totally worth doing it. Think of all your offices, computers, phones, tablets, employees, contractors, agencies you work with, working-from-home-days, just-check-this-one-thing-even-though-I-am-on-a-vacation… All this is internal traffic. In order to fix this, we’ll need separate filters for each team, office, team-member, etc. By the way, if you have a dynamic IP number, you’ll need a slightly more robust solution. [If you need help—feel free to contact us]
How to:
- Go to Admin panel
- In the 3rd column select Filters
- Click Add filter
- Name the filter: ‘Exclude name_of_the_team_member/branch/office [IP address]’ (for example: ‘Exclude HQ traffic [52.3.98.157]’)
- Filter Type: Predefined
- Select filter type: Exclude
- Select source or destination: traffic from the IP addresses
- Select expression: that are equal to
- IP address: [paste the subject IP address]
- Confirm by clicking Save
By the way, checking your current IP address is as easy as typing ‘what’s my IP’ in Google.
4. Removing your own domain from self-referral (increases accuracy)
Now this one is an absolute must-have. Although it’s so easy to implement, it’s still not in Google Analytics by default yet, so we’ll have to fix it ourselves.
If you want to produce valuable insights from your data, you have to make sure your users’ visits are being tracked correctly. It’s all about the session (= entire visit) tracking here and making sure they are not being broken by the actions which aren’t breaking them. Let’s consider the following example:
Your visitor googles something. She lands on one of your blog posts. She reads the article, clicks to another one, reads it, and then clicks on the top navigation logo and goes to your home page. If your domain address is not on the Referral Exclusion List then this one session (logically it is a one session) will be split into 2 sessions:
1st: Traffic source: Organic Search; Pages per session: 2
2nd: Traffic source: Referral from (yourdomain.com); Pages per session: 1
Again, luckily the fix is pretty easy to implement.
How to:
- Go to Admin panel
- In the 2nd column select Tracking Info
- Choose Referral Exclusion List from the list
- Click on Add Referral Exclusion
- Enter your domain address (e.g. yourdomain.com)
- Confirm by clicking Create
5. Get rid of the most common bot traffic
Making decisions based on data generated by bots is definitely not a good idea for obvious reasons. That said, I don’t suggest filtering bot traffic on the Raw view where we collect all the data with all rights and wrongs and so it is with bot traffic. But getting rid of bots from the traffic we analyze is paramount if we want to say the data here is clear and accurate. Google Analytics lets us easily filter out the traffic generated by the most popular bots (if a bot is popular, does this mean it was a quarterback or a cheer-leader in high-school?) making the data human-only. Well, until another, not necessarily popular, bot shows up in town.
We very often observe that a high bounce rate can be caused by the traffic generated by bots. I bet you see how devastating this can be to your decision making. Imagine you may get rid of one of the most annoying marketing headaches by fixing just this one thing.
How to:
- Go to Admin panel
- In the 3rd column select View Settings
- Check ☑︎ the Bot Filtering
- Confirm by clicking Save
Don’t ask me why this is not checked by default.
This is the first, most important step in fighting against bots. But this war is never over. Every week new ones show up attacking the clearness of our data. So in order to keep your analytics safe and sound, you should come back to this step regularly, setting up the new bots traffic with custom filters.
(Now, you can get back to the step 1. to finish setting up the 3 main views)
You Ready to Start Using Google Analytics?
Phew! That’s it. Having these 5 points covered, you’re joining the exclusive club of tuned Google Analytics owners. You’ve just covered the list of patches to the most popular analytics problems.
I hope that the most important takeaway from this article for you will be to start using Google Analytics in general. Don’t be afraid to use data! It’s there to help you. I believe one of the best things teams can do in order to become more successful is to build data-driven culture around them. If you need any help with that, or found any problem in your GA setup that doesn’t fit the descriptions above, feel free to get in touch.
Pssssst… Not yet using Crazy Egg but want to get a taste of the value of heatmaps when it comes to analyzing your visitor data? We have a free Google Analytics Heatmap ready for you to try!
Free Webinar: How to get Maximum Value from Google Analytics and Crazy Egg to Increase Conversions
Google Analytics is near ubiquitous on websites around the world. Yet, almost 90% of implementations aren’t right.
Join us on September 26th to learn a few simple things you can use to fix your Google Analytics installation and, make sure you’re getting maximum value from your data by using the depth of Google Analytics with the powerful interface, visualizations and user recordings from Crazy Egg. Sign Up Now
