My Take: Google Analytics Audit with Verified-Data.com
A forensic approach to Google Analytics data quality and compliance.
I first wrote about Verified-Data six months ago. Verified-Data is a sophisticated Google Analytics auditing service from my good friend Brian Clifton, author, speaker and overall data quality and privacy expert. I recently dug back into it and thought I should do another review.
What’s the price of good data quality? Worse, what’s the cost of making good decisions informed by bad data?
The notion of “good decision informed by bad data” is particularly pernicious. As a decision maker, you might think you can rely on good data when it is not actually the case. You just don’t know it. You confidently make a decision but you’ll realize much later it was an error costing time, money and even reputation.
The Value Proposition
My bag of experience includes the creation of the Web Analytics Solution Profiler (aka WASP) over a decade ago — at the time, it was the first tool specifically created to make data validation and auditing accessible to non-technical marketers and analysts. Over the past 20 years, I have done countless implementations and more often than not had to pick up the mess of a botched implementation or something that was set once and forgotten thereafter. Data quality is a constant battle and nowadays, the requirements and expectations are even higher.
Although there are many other tools available, almost all of them are spitting out technical and hard to decipher HTTP requests, or they crawl and merely confirm if a given tag is present on a page. Some are attempting to look under the hood — the configuration of Google Analytics itself — but fail to provide actionable results. That’s why Verified-Data stands apart.
Verified Data is unique in the way it uses a forensic method to audit your Google Analytics data. Combining off-page testing with on-page testing brings visibility of what data is being collected (API checks) and what data is missing (visit simulator). This provides a unique and complete picture of your Google Analytics data quality.
What’s Cool, New & Improved
Since I reviewed the service six months ago, a lot has changed. Let’s look at some cool features.
The number and types of validations are quite impressive — over 200 different tests. Here are some of the validations:
- Inaccurate data — pollution from robots, spammers, test users, staff, etc.
- Misconfiguration — wrong filters, poor segmentation, incorrect Adwords linking, missing ecomm variables.
- The dreaded collection of personal information, including checks for dimensions containing credit card numbers, email addresses, phone numbers, and many more.
- Duplicate counting — duplicate tracking pixels.
- Rogue counting — ghost accounts you are not aware of.
- Tracking code coverage report — the percentage of pages with the correct tracking code firing.
- Missing data — are file downloads and outbound links tracked?
- Error page differentiation — are they tracked correctly?
- Cookie report — list the cookies set, their properties and validate their consent i.e. for GDPR.
- Validate cross-domain tracking — so easy to get wrong.
- Smart checks of campaign parameters and spam referral.
- Audits scan can be on-demand or scheduled and either scan new pages or re-use the previous same pages.
- You can select the region from which you want to do the audit and the browser for the simulation.
- A nicely laid-out email is sent once a scan is completed.
- Validation thresholds are determined through some pre-checks and the user is free to use those “auto-adjusted” values or enter some strict values to pick up “warts and all.”
- Audit results link directly to the relevant Google Analytics reports — support handy!
- Reports are clearer and easier to read, with improved layout and graphs.
- Speaking of reports, there’s a cool new API feature coming up which will allow you to bring the audit results directly in your tool of choice (I’m already thinking of a nice Tableau dashboard, although
- In order to make auditing easier, there’s a “consent override” feature so you can review what happens when consent is given — or not.
- The historical audit scorecard becomes a useful reference if you ever need to demonstrate what you’ve done to ensure compliance.
- And of course, the risk of personal data collection is thoroughly checked.
There’s nothing better than actually looking at a real site and a real audit.
The first run was with the default thresholds and the result pointed out to some improvement opportunities.
It is very rare to see an initial audit reach a quality score over 70. They must have had a real expert in to help them. — Brian Clifton
I got to the task and reviewed the following points:
- Under Account Structure & Access, it warned about people with generic emails (ex. GMail addresses) with access — it’s always a good practice to enforce the use of corporate emails only.
- Under Data Validity, some events were being fired before a pageview, resulting in missing landing pages. A couple of events had long labels which were truncated. And there was still one referral considered to be spam.
- The way E-Commerce Tracking is setup up for this site makes the Add to Cart and Checkout steps irrelevant, so as expected, those were flagged as issues.
Obviously every site has unique data thresholds that the site owner can adjust to match the business. The initial audit typically pinpoints two areas of potential improvement: fixing Google Analytics misconfigurations and correcting tracking issue.
Despite the depth of my experience, Verified-Data was able to pin-point and help me fix data issues which would likely have remained hidden and unknown.
After some adjustments, the overall score went from 68.8 to an amazing 84.4% score. It is then the role of the implementation expert, along with the analyst and the business user, to find the balance between hypothetical perfect data quality and the effort, cost and real impact of acceptable data.
It’s nice to see how Verified-Data has improved over the past six months and the roadmap ahead is quite promising, including an API to retrieve the audit results and attacking the audit of Google Analytics 4.
This really sets Verified Data apart as the solution for Google Analytics implementation audit. The AI and forensic aspect built into the solution is the key to its evolution and the detection of new patterns in the data. I would love to see this service extended to include the scanning & collection of other known tags, such as Facebook Pixels, Adobe Analytics and many more.
Why not give it a try ? No strings attached, you can get an audit for four the four key sections under “Core Principles”: Account Structure & Access, Tracking Code Coverage, Data Validity and Error Page Tracking, which amounts to over 50 different tests. The pricing model is quite flexible with a pay as you go option ideal for occasional spot checks or a subscription model certainly well suited for agencies or larger organizations.
Stéphane Hamel is a seasoned independent digital marketing and analytics consultant, innovator, teacher and speaker with a strong interest for user privacy and the ethical use of data.
If you enjoyed this article, you should follow me on LinkedIn and while you’re at it, why not click the nice little clapping hands on the left and follow me on Medium!