Google Analytics Audit
A handy checklist to audit your Google Analytics implementation
Google Analytics is a dream for any digital marketer or an analyst and a goldmine for businesses to gain valuable insights on their web properties. The caveat here is that it should be set up and used in the right way to be helpful. The reality though is that the overwhelming majority of the Google Analytics implementations are flawed and under-optimized.
This post will look at the steps involved in conducting a detailed audit of the Google Analytics implementation. By the way, this is part of the review series of the Conversion Rate Optimization course at CXL. So let’s jump in.
To start with, you would need 2 essential tools (Chrome Extensions) to do a proper audit of Google Analytics namely,
- Google Tag Assistant
- Adswerve DataLayer Inspector+
Apart from those, the following 2 Chrome Extensions are very handy as debug tools as well.
- GA Debugger
- GTM/GA Debug
The Audit Process
We will cover the following areas in the GA audit and it will be very easy to do the audit on a simple Google doc or an Excel sheet.
- Account & Property Overview
- GA Property & View
- Verify Pageviews Accuracy
- Filters
- Default Channel Group
- Site Crawl
- Content Grouping & Query Parameters
- Events
- Goals
- PII (Personally Identifiable Information)
- EEC (Enhanced Ecommerce)
Let’s get into each of these in detail.
Account & Property Overview: Create a list of accounts, the number of users, the number of views, and if any filters that are being used. In addition, it is useful to note what is sending data to Google Analytics and there are 3 ways you can do this — through Universal Analytics, Gtag, and GTM.
GA Property & View: The important settings to verify here are the default URL (and if it redirects to any other URL), referral exclusions (to filter out traffic from your dev environment, and also cross-domain traffic).
Also, Verify view settings and the existence of specific views. A proper Google Analytics setup includes at least two different views and specific settings within each one. The raw view, if it exists, should not have any blocking filters applied. The raw view is your backup and safety net, so you don’t want to restrict the traffic going to it.
Custom dimensions give additional information than what is available by default and some of them are key to your analysis. A really cool reference list of custom dimensions to include is in this post by Simo Ahava.
Google Search Console gives a lot of useful information on how your site is performing on Search Engines and it should be enabled.
The remaining settings to verify here are the Site search and this should exclude the query parameters so that they don’t mess with the page traffic data.
Annotations (easy references for site changes and their impact on the site performance) and alerts (low traffic or conversions) are important for easy reference as well.
Sending Pageviews Correctly
Pageviews is a fundamental Google Analytics metric and you need to make sure they’re right. It is common to see sites underreporting or over-reporting pageviews and this is down to faulty Google Analytics implementation.
Underreporting happens when page views are not sent from every page, whereas over-reporting is caused by having both GA tracking and GTM enabled.
dataLayer Inspector+ is a useful tool to verify accuracy in reporting page views. In most cases, there will be low-hanging fruits that have the potential to bring a significant effect to your site.
Hostname Filter
This is meant to discover which hostnames are valid and the hostnames you’d want to exclude are your own domains and internal referrals. Also, you need to check if there are referrals that look suspect (high traffic and high bounce rate) and these could be bots.
IP Filters
The purpose is to exclude traffic from employees within your own company. You can filter a single IP address or if you are having a range of IP addresses, those can be blocked as a whole as well using regular expressions. This can be used to handle anonymous IPs as well.
The recording feature of the Tag Assistant is a cool tool to verify that the IP filters are working properly.
Default Channel Group
The key to look for here is to ensure the source/medium variables being used are the same as Google’s Default Channel Group or else they’ll get grouped under the dreaded “Other”. And if you want to use your own variables to group traffic, they need to be set under channel settings.
Emails traffic can often be grouped under “Other” and that’s because they are not tagged properly. Make sure to always use Google URL builder for all the links in your emails.
Direct is another place where the untagged traffic can be grouped. This will also include redirects from http to https and so on.
Site Crawl
With two different tools, you’ll learn to crawl your audit property to look for analytics tags that should and shouldn’t be there.
Here look for pages that do not contain the GA tag or the GTM tag depending on how you’ve implemented it. Also, look for utm_ parameters within the site and it is better to remove them since they can throw off your session accuracy and break traffic sources.
Good tools to use here are,
- Screaming Frog
- Website Auditor (part of SEO PowerSuite)
Content Grouping & Query Parameters
The idea is to discover if specific pages are broken into multiple lines due to the query parameters. Once you find them, exclude them using filters to have clean data of the pages. The other things which could be excluded are trailing slash and lower case site search filter. This filter could be applied at the account level, which is a more scalable option, or at the view level.
And now let’s get into the most important items that need to be tracked and tracked well — the Events, Goals, PII and Ecommerce.
Events
This is the most important part of the GA implementation and if you could get just this one right, it will make a big difference in tracking your website performance. So the things to check here,
Are there Events being tracked at all? If not, that’s a real danger sign. Things like outbound link clicks and downloads should be fully tracked.
Are there too many or too few Events? A good upper limit would be 25 Event categories as a lot more than would make it tough to analyze and draw insights from.
A few more things to look out for are, if the Event labels are duplicated, does the labels make sense, and are they specific to the things you want to track.
For eCommerce tracking, at a minimum, it should track product clicks and cart actions, whether add or remove.
Finally, check if there are actions that you need to track being missed. Useful items include social media links, blog links, media engagement, etc.
Goals
Like Events, you need to make sure that Goals exist, if they make sense, and the number of Goals optimal.
On the number, it is a good practice to keep the number of goals to a bare minimum and are used just for the important interactions on the site. There is more reason to be stingy with Goals since GA free allows just 20 goals and GA 360 limits it to 200.
It is always good to keep checking on the Goal numbers to verify everything is in place and not broken.
PII
PII or Personally Identifiable Information is a sensitive area that needs to be taken care of and Google is quite strict in its policy towards capturing PII. This includes information such as email, physical address, and phone number which can be mapped to an individual.
PII can be found under page content, site search, events, and custom dimensions and these are the places to look for PII while doing the audit. There are some cool resources available online like the ones below.
EEC (Enhanced Ecommerce)
Tracking Ecommerce is the most important along with Events. The things to verify here are if all the events related to Ecommerce are tracked fully including the product labels and that the conversion tracking through funnels is configured correctly so that the site properly reports drop-offs between steps.
Another key thing to verify is if the GA implementation matches with what is actually happening on the site to ensure you are capturing and analyzing the right data.
Here the tools, dataLayer Inspector+ and GTM/GA Debug are very handy and often you can identify issues in the implementation by just having a look at the sites with these tools, you’ll learn how to discover if the EEC data is being captured correctly.
Wrapping Up
This is the beginning of your journey and not the end. You’ll continue to learn a lot from the sites you audit. You’ll run into issues you never knew existed. And you’ll have fun figuring out what the heck went wrong. Happy auditing!
Ciao
