Whenever we start an analytics project for a new client, the first thing we do is a complexity scan of the Google Analytics Account of this client. We go through basic and advanced settings, check the data quality and consistency – we try to find weak spots and make suggestions on how they can be improved. Most of the times we encounter a number of recurring problems which appear in almost every new account we work on. In this article I wrote down the five most frequent problems. And I also provide tips on how to solve them.
1. Wrong goals settings
One of the most useful features in Google Analytics is defining and tracking your own goals. Even though goals can provide really valuable insights on how your site is performing, a lot of accounts have set them wrong, or don’t have set them at all. When it comes to goals, there are three most frequently occurring situations:
- Goals are not defined at all – this is, obviously, the worst scenario. Which users are the most engaged with your content? In which step to goal completion is the biggest drop-off of users? What is your goal conversion rate? Without defining goals, we are not able to answer these questions. Defining goals might be a bit challenging at first, but it’s really essential if we want to measure how your business is performing.
- Goals are defined, but not correctly – in this case, goals are defined, but they aren’t tracked correctly – we can simply identify this by checking how many times a certain goal was completed in the last 7 to 30 days. If we see a low (or suspiciously high) number of conversions (or not conversions at all) it’s a strong indicator that there might be a mistake in goal definition. There can be many different reasons for this – typo in destination URL, mistake in RegEx definition, etc.
- Goals represent only the macro-conversions – it’s important to set the goals in such a way that they reflect the user’s behavior on the site. There are many interesting steps or actions which happen before conversion (macro-conversion). Tracking these will provide you with a better picture of the users who are getting closer to the main conversion. It might also identify problematic steps in a funnel with a high drop-off rate which could be optimised. Micro-conversions differ from client to client: it could be newsletter subscription, product page view, contact form submission, whitepaper download, etc.
2. Filters are not applied in GA
Filters in Google Analytics are a really useful tool to make your data more precise and consistent. By using filters you can:
- filter out internal traffic (filter out internal IPs, IPs of your digital agency or the development team)
- strip out redundant parameters from URL
- segmentate traffic based on source, device, location…
- and a lot more!
However, it’s important to understand that even though filters can help clean up data, filters can also modify your data to a completely useless form. Also it’s important to remember, that once you apply a filter to the view, data which were affected can’t be changed to its original state. That’s why it’s always a good idea to keep one unfiltered backup view. So even if something will go wrong, you will still have an access to the raw data. Next to that you should create one testing view, in which you can test the effects of newly created filters – better be safe than sorry!
3. Only one view
The problem with having only one view is that either we have a view with filters – so we don’t have access to raw data, or we have a view without any filters, which might cause inconsistencies in the data. We always suggest to create at least three views:
- Master View – view with all filters and goals – used for reporting and analysis.
- Testing View – view with only internal traffic, where you can test new settings, filters or goals
- Raw Data View – view without any filters – a backup view in case of data corruption in the main view
By applying filters, we can create more specific views – you can create a view for traffic sources or devices, or a view which will display data only from specific subdomains.
4. Custom dimensions and metrics are not defined
Custom dimensions and metrics are essential for the collection of business specific data, which aren’t collected by default in Google Analytics.These dimension will help you with filtering data or creating segments. Using custom dimensions and metrics is necessary for advanced analysis, which can’t be done with just default tracking.
Sending custom dimensions and metrics requires an additional setup in Google Tag Manager (or editing tracking code), but the benefits are huge and we would recommend everyone who is not using them now, to think about dimensions and metrics. It adds great value to Google Analytics reports.
5. Events are not used
Events are a great way to track interactions on your site that are not page views. Using events, we can track form submissions, scrolling, video playing and much more. Basically, anything that isn’t page view can be tracked as an event – these events can help us with defining goals, which are not possible to define as URL destination (for example newsletter subscription or sending a contact form).
BONUS: Annotations are not used
Annotations are really simple and a very useful feature of Google Analytics. It lets you add notes to a certain date. Are you starting a new campaign? Will a new version of GTM container be published soon? Is the design of your site going to change? It’s really easy to add an annotation concerning these kind of events, but most clients don’t use it. Why is it so useful? It might give you insights when you see a huge spike in some metrics or other data which you have been collecting. For example, if you made an annotation about a new version of a GTM container which went live last Monday and you see a huge increase or decrease in page views on that date, you will have a better idea what the problem might be. You can make annotations for yourself or you can share them with other accounts which have access to the GA account.
I hope this article was helpful to you and it will help you fix or prevent some of the problems I mentioned.