GA Audit and Google Tag Manager overview-Review
Conducting a Google Analytics Audit is the best way to ensure you can trust your data. We cannot always have a GA account set up perfectly at the first attempt. We learn by mistakes. To make the data more trustworthy, you need to audit your analytics set up to make sure things are being tracked correctly. There are experts in this field who can audit your GA account for you. But as a Conversion Optimizer, it’s an added value if you know how to audit your organization’s GA account by yourself.
Before you begin, install the below Chrome extensions which you will be using quite a bit throughout your audit:
There is a dedicated and elaborate lesson in the CXL institute’s Conversion Optimization Mini-degree on how to do GA audit. In this post, I’m discussing 6 basic sections to audit on a GA account, which I think are crucial for someone who is just getting started.
#1 Setup
The first thing you need to do is verify the Setup of Property and View. Check if the Default (Property) and Website (View) URLs match. Ensure the default URL doesn’t have any redirects (Use Link redirect trace chrome extension). There should always be one Raw unfiltered View and one Test View. Every valid GA hit includes a document hostname. Make sure there is no hostname filter setup on your Raw View. Check if your Referral exclusion is done correctly.
#2 Accuracy
Sometimes, only after collecting so much data, we will get to know that we weren’t testing the right traffic. To avoid that, you need to ensure the traffic you are getting is accurate. Verify if the pageviews are over/undercounted. When a visitor lands on a page, only one pageview has to be sent. Check for the IP filter and see if it is excluding the right IP you want. You can use the Tag Assistant recordings to see if it’s working or not. Remember the IP address field is not a RegEx field. Also, it’s better if you learn how to decode the CIDR IP address.
#3 Data Sources
Once you have the accurate traffic, you want the data to fall in the right bucket in your GA. We know that Google has default channel groupings like Organic, Direct, Referral, etc. The ones that don’t match those mediums will come under ‘Other’. Adjust your tracking so that it matches the default definitions. For example, traffic from your email campaign is going to show up as ‘Direct’ by default. In that case, you need to set up UTM parameters (should be lowercase) for the traffic to show up as ‘Email’ in your GA account. Understand the default definitions so you can better organize the incoming traffic.
You need to do Site Crawling to ensure the relevant snippets (Analytics.js, GTag.js, GTM) are added on every page of the website you are auditing. Using site crawling tools (eg. Screaming Frog) you can even see which pages have UTM parameters integrated on them. This should be done at the early stage of the audit. I’ll tell you why with an example: If you are running an email campaign with UTM parameters, the parameters will fail to append if the page already has UTM integrated on it. In that case, your session will end and you will lose that referral.
#4 Content grouping
Same pages should be grouped in GA. You may see a single page being broken into multiple lines because of the Query parameters. This might often skew the results. Exclude them in the view settings or globally on the filters. Also, check if the same page does or does not contain a trailing slash.
#5 Interactions
Try to understand what user interactions are most valuable to track on a website and why. If Events and Goals exist, check if there are too many and whether they make sense. To avoid confusion, it’s better to have not more than 15–20 event categories. Make sure there are no duplicates in the EEC labels (Enhanced Ecommerce).
#6 PII
Collecting Personally Identifiable Information in GA could lead to Google permanently deleting your account and information. Hence it is important to search and mitigate them. You can use RegEx patterns to search it in the segments.
Google Tag Manager Overview
The Tag Management System helps you gain visibility and track specific behaviors on your website. Understanding GTM can help reduce your dependency on Developers. GA and GTM are entirely different platforms. GTM is just there to collect information. It doesn’t store and report the data as GA does. To get started with GTM, you need to understand the below terms.
Tags: A tag is a script that sends data to platforms. When you’re using a platform like Hotjar, you can fire the snippet on your website using GTM without having to modify the code. You are giving that power to the Tag manager to do that for you. The tags in turn report back to the platforms with data when an event happens. There are two types of tags, built-in and custom HTML. If you are using a platform that contains built-in tags in the GTM, you can directly use that to fire the script.
Triggers: It answers the question: When do you want your GTM to take action? For example, only when a pageview occurs, I want my tag manager to fire these tags. Only when a visitor plays a video, I want my tag manager to go tell the analytics that it has happened. It primarily answers the question ‘when?’. There are 4 types of triggers in GTM: Pageview, Click, User Engagement, and Custom.
Variables: Variables are information that GTM needs to do its job. For example, instead of GTM telling, ‘A’ pageview has happened, using variables it will tell the name of the page. There are built-in and user-defined variables in GTM.
Data layer: It is a temporary storage of details. Every time GTM loads on a page, it creates a new data layer. If the key of the data layer is ‘Invoice ID’, the value can be ‘6378’. It is more like a Key-Value pair for storing the data. It temporarily stores the data that GTM might need to report back to the platforms. When the next page loads, a new data layer will be created.
There goes a short summary of my learnings on the GA audit and GTM overview lessons from the CXL Institute. Thanks for reading this article! Do share your thoughts on comments.
