One thing that many GA users learn the hard way is that once the data has been sent in, you can't change it. For this reason, maintaining a well-functioning Google Analytics setup is crucial for obtaining accurate data and insights from your website or app.
If you're not regularly checking your data, the risk is that problems with the data can go undetected. This ultimately leads to a long list of gaps and caveats in your data, which can undermine trust in the data within your organisation. Without that trust, you're not going to be able to confidently take action to optimise your website or marketing activity.
In this article, I'll outline a comprehensive GA4 maintenance checklist, categorising checks into five sections:
You can access the template and make a copy here. Feel free to remove things that aren't relevant to you, or change the frequency of the checks to suit your organisation.
Data collection
1. New events or modifications
If there are lots of people who can add or edit events, either in GA or GTM, then what tends to happen is things get added that don't fit best practices or your naming convention.
Ultimately, your setup can end up a mess, so regularly checking to see what's been added is one way to control things.
2. Data collection limits
There are certain limits around data collection, which could mean missing data if exceeded.
The most common ones to look out for are:
Length of event parameter value - 100 characters
Length of user property value - 36 characters
Length of user property name - 24 characters
Event parameters per event - 25 event parameters
A full list can be found in Google's documentation. Is there any evidence that any data collection limits listed in the Google documentation are close to or being reached?
3. GTM documentation
Does your documentation have the latest changes? If you don't have GTM documentation, you can find a template here.
Configuration
4. Data Import
This feature can be used to upload data from external sources and join it with GA data.
If data import is in use, check if the data appears correctly in reports.
5. Configuration limits
Like data collection limits, there are also limits on the configuration side. For free GA users, the most common ones to look out for are:
User-scoped custom dimensions - 25
Event-scoped custom dimensions - 50
Conversion events - 30
For a full list of limits, see Google's documentation.
6. Filters
Are live filters working as expected? Are there any filters in testing that are ready to go live?
7. Test property
It's good practice to have a test property for testing changes that could have a major impact on your data before you move them to the live property.
In order for the test property to be useful it's good to keep it set up in the same way as your live property. Check that it's in sync.
8. Account change history
Check the change history to see if there have been any configuration changes that may cause issues. This is another check that shouldn't be needed if your organisation is restricting access to the right people. The reality is that this isn't always the case.
Data Protection
9. Personally identifiable information (PII)
It's against Google's terms of service to send PII, and sending it to GA could be a data protection breach too. Typically if you find it it's passed within query parameters in the URL but it could be passed in other places too. Look for things like email addresses, physical addresses, names, phone numbers, credit card details and passwords.
10. Consent Management Platform (CMP)
If you are based in the EU you likely have a consent banner on your site. It may have been working fine when it was initially set up, but it's important to ensure that it continues to.
Are there any new cookies that need to be categorised by the CMP? Are there any GTM tags that are not integrated with the CMP?
11. User management
Who has access to the data? Is there anyone that should be removed? Keep editor access to only those who really need it. It's good practice to periodically review this.
Data Quality & Reporting
12. Unwanted referrals
Check your referral sources. There should be minimal or no traffic from your own site(s). Ensure that any payment gateways are also not appearing here. If they are, they should be added to the referral exclusion list.
13. Sampling
How badly does the property suffer from sampling when using explorations? Over what time period does sampling typically kick in?
14. Explorations limits
Is the property close to the limit of 500 shared explorations?
15. Tracked hostnames
Are there any unwanted hostnames appearing in the data?
16. Cardinality
Cardinality is caused by custom dimensions with a high number of unique values (>500 per day). This high granularity of data causes GA to consolidate data into a row named '(other)'.
Check if reports are being affected by cardinality, or if there are any dimensions that have a high number of unique values.
17. Thresholding
Thresholding is caused by Google Signals. This feature combines data that Google has on users (who have opted in) with GA data to identify users across different devices. This gives a more accurate user count, and for those who are included gives a more accurate picture of their user journey.
The downside of Google Signals is that GA withholds rows in data tables in GA if they have small amounts of data. This is called thresholding.
Determine if your reports are affected by thresholding. If this data is important to you, you may want to switch reporting identity to 'device based', which uses only the GA cookie.
18. Not Set Landing Pages
Sessions that contain no page_view event will have a "not set" landing page. A certain level of this is unavoidable because it's caused by normal user behaviour.
Analyse the traffic to "not set" landing pages and track how it trends over time. Sudden changes may indicate issues with your tracking setup.
19. Channel Reports
Regularly review channel reports for potential issues and trends. Identify changes in traffic sources and assess how much traffic falls into the 'unassigned' category. Issues here may indicate issues with your campaign tagging processes.
20. Content Grouping
Verify that content grouping is working as expected. Ensure that all pages are appropriately categorised, and there are no uncategorised pages.
21. Conversions
Analyse the trend of conversions over time. Identify major changes or any conversions that have dropped to zero, which may require investigation.
22. Topline Metrics
Keep an eye on overall traffic, engagement rate, and event counts for irregular activity. Sudden spikes or drops may signal issues with tracking or website performance.
23. Audiences
Check that any defined audiences are collecting data as expected. Ensure that your audience segments are properly set up and functioning.
24. Reporting Interface
Review the reporting user interface (UI) for any customisations or changes. Ensure that any customisations do not introduce errors or issues in reporting.
25. Error Tracking
Monitor error-tracking data for trends that require action. Identify recurring errors or issues that need to be addressed.
26. Site Search
Verify that internal site search tracking is working as expected. Ensure that search queries and results are accurately recorded.
E-commerce
27. Verify Against Backend
Compare Google Analytics data to backend data to ensure consistency. Significant discrepancies may indicate tracking issues or data integrity problems.
28. Checkout Performance
If you track the checkout funnel, analyse the data to ensure it appears sensible. How does it trend over time? Any anomalies may suggest problems in the checkout process.
29. Cart-to-Detail Rate
Does the cart-to-detail rate seem reasonable? Have there been any sudden changes?
30. Item-Level Data
Ensure that item-level parameters are consistently applied across all e-commerce events. Inconsistent data can lead to inaccurate product-related insights.
31. E-commerce Journey
Evaluate buy-to-detail and cart-to-detail rates to gauge the effectiveness of your e-commerce journey. Anomalies may indicate usability or conversion problems.
Conclusion
Regularly performing these Google Analytics maintenance checks is vital for maintaining data accuracy, ensuring compliance with privacy regulations, and optimising your analytics setup. By systematically reviewing and addressing issues, you can trust the data you collect and make more informed decisions. Remember that a well-maintained analytics setup is the foundation for effective digital marketing and business growth.
Comments