One of the services our Analytics team provides, is a comprehensive Google Analytics health check. In fact, we have conducted one for almost a third of UK universities, gaining invaluable insight into common and complex issues and providing solutions to overcome them. To give you the chance to reflect on your Google Analytics set-up, Joe, our Digital Manager, has compiled a list of the five most common issues found in our audits…
What We’ve Learnt From Our Google Analytics Health Check
Starting work with any new client involves understanding what normal looks like to them. So ahead of a full Google Analytics audit, we first run a “discovery stage”. It’s as part of this stage that we do a Google Analytics health check. During the process, I examine several elements within the client’s Google Analytics (GA). The simple aim is to understand what data is present and to assess the quality of that data for making informed and robust business decisions.
Here are the top five issues we’ve found with the Google Analytics set-ups that we’ve come across:
1. Google Analytics View Configuration Issues
For those new to Analytics, a View is the lens by which you ‘view’ your analytics’ data. A view, or views in most cases, sit under the Property, forming the most dynamic part of the whole Google Analytics (GA) account.
Each view is individually configurable to include, exclude and understand data in different ways depending on what it needs to show.
We find that most issues arise not within the features that distinguish views from one another, such as different filters. These are often configured correctly. The problems often occur due to inconsistencies in the constant settings that should remain the same between views.
As an example, let’s take a UK University who create a separate view for each one of their markets: UK, EU, USA. The filters successfully segment traffic from each market. However, the UK is the only view that has Goals set up, bot filtering turned on and is linked with the company’s Google Ads account. Both the USA and EU views have e-commerce reporting turned on with the currency set to US dollars; they also maintain the UK GMT zone.
Now, some of these choices may be intentional. But in most cases, a view configuration like this would hinder reporting as the data between views would not be entirely relative making cross-market comparisons difficult.
2. Goal Tracking Issues
We regularly find either little to no conversion tracking or a large amount of tracking setup by different suppliers for their activities (some of which are no longer active). We encourage our clients to move away from this additive approach and develop an integrated model allowing for tactic specific tracking inserted into an existing configuration; this keeps things neat, fast and scalable.
In Analytics, events and goals are kept separately, one under ‘Behaviour’ and the other under the ‘Conversion’ section.
Events are configured externally either in the site’s code or through a tag management system such as Google Tag Manager (GTM); events are then pushed to the Analytics account and appear in the views.
Goals are different and are configured inside the view settings in the GA admin section. You can create goals around a variety of different types of user behaviour including creating one from an event as it enters the account.
I regularly see event tracking of ‘non-interaction’; e.g. a promotional banner impression that often causes site-wide issues with bounce rate.
Regardless of configuration, Events and Goals are essential in understanding user behaviours and onsite experiences. They also form the conditions by which to measure outcomes against organisation/marketing objectives and to segment audiences for targeting purposes which is a fundamental part of a comprehensive remarketing strategy.
3. Direct Traffic
Direct Traffic plagues all web analytics accounts. Traffic is labelled as ‘Direct’ when it enters the GA account with no referral information e.g. GA doesn’t know where it’s come from. We’ve seen cases where it can make up to 65% of all traffic to the site, which makes it hard to attribute behaviours to specific sources.
There are various reasons for Direct Traffic, including but not exclusive to:
- Untagged Links in Emails – Email channels rarely pass over referral information for security reasons so ‘in email’ links need to be manually tagged
- Traffic from Mobile Apps – Again, apps rarely carry over referral information, but links can be tagged in most cases
- Your Site is Not Encrypted – If your website domain starts with http:// rather than https:// it means that it’s not encrypted. For security reasons, a user’s browser will refrain from passing over referral information to a ‘non-secure’ site.
4. Personally Identifiable Information (PII)
“We’ve found PII.” This is never something we like telling a client! But it’s important to do so as soon as possible as the ramifications can be huge.
PII mainly appears in the site’s URL, which then gets pushed to GA and is viewable in the content breakdown. In most cases, this occurs during user interactions with forms that query the site during logging in, changing personal details, searching, etc.
The most commonly found instances of PII:
- Email Addresses
- First Names and/or Last Names
- Public Usernames
- Partial or Full Addresses
The simplest way to check for PII is to scan behaviours and the URLs listed against them within a 30-day lookback window. If you see one instance of PII, chances are there will be many more.
For advice on how to find out if there is PII in your Google Analytics data, check out this post.
If you do find PII, you need to act quickly. Normally this would be a case for developers. However, there are things you can do to prevent any further instances of PII from reaching your GA account. Get in touch to find out more.
5. No Link to Google Ads
PPC advertising is a big part of many of our client’s digital strategy. Google offers a simple linking feature between its Google Analytics and Google Ads platforms.
Without this synchronisation between the two products, marketers and advertisers miss out on crucial data such as campaign attribution in Analytics and onsite Conversion insights in Google Ads.
As universities have a long recruitment period, there are multiple touchpoints over many months, therefore understanding this and not using last click attribution is crucial as Google Search will almost always win!
Do any of these issues raise a flag for you? If you’re wondering how healthy your Google Analytics set-up is in comparison, get in touch to request a free Google Analytics health check with Joe.