In the digital arena, all types of businesses like to make an impact in the online world. They invest heavily on building a website which is a crucial entry point for consumers. Such hefty investment leads us to a very important set of questions revolving around whether it is working for their business or not. So, knowing the performance of your website is imperative to gauge the return on your investment.
Measuring is always a second step, with the very first step being getting the measurement tools in place. Before taking an important decision based on the performance metrics, you should make sure that the measurement tools are properly set up and that the data coming through is accurate. From the web analytics stand point, this means that the analytics implementation (be it Google Analytics or Adobe Analytics) should be comprehensive, flawless and as per the best practices.
In this blog, we have listed down 5 important checks you should carry out to know the health of analytics implementation on your website. You can carry out these checks for any website and not just yours. To keep the explanation simple, I have considered GA implementation here. You might have to convert these ideas based on the analytics tool you use.
First and foremost thing to do is, know the tracking tools deployed on a website: You can see all the analytics and marketing tools deployed on a website without having to look at the page source code. A chrome extension called “Ghostery” helps you do this. Clicking it will show you all trackers deployed on the website. Please note that Google Analytics will be shown under Site Analytics section.
1. Check if the analytics tags are firing in all pages.
Now you know for sure that GA is deployed. The next step is to see if the analytics tag corresponding to GA is firing. The basic tag is a pageview tag which usually fires on every page of the website. The analytics tags are passed as network calls which can be observed using default browser tools or third party debugger tools like Charles Proxy, Fiddler, HTTP Fox etc. Here I’m using Chrome developer tools.
Open the webpage in a chrome browser and Right Click > Inspect (keyboard shortcut: Ctrl+Shift+I) or F12. This opens up the developer tab. Choose Network tab from here as shown below. Refresh the page once you open the network tab. And type “collect” in the search bar. You can see a list of tags below which are Google analytics tags. “collect” is the keyword used to filter only google analytics tags.
There should be at least one “collect” tag firing for the pageview firing on page load. You can click on the tag to see the tag variables and values. This confirms that a page view is recorded in Google analytics. Repeat this for multiple pages on the website.
2. Look for event tags
Event tracking is a very important part of analytics tracking. We cannot get deeper insights only with page view metrics. The aim here is to understand the behavior of visitors within the site. So, we need to identify the micro and macro conversions within the site and implement analytics tracking for all those interactions. For example, an ecommerce site’s macro conversion is “order conformation” and the micro conversions are steps towards the macro goal which is adding products to cart, clicking on check out etc. which have to be tracked.
To test this, out in your website, look for key CTAs which can be a micro goal or a macro conversion and click on that. Clicking it should trigger a Google Analytics “event” tag. As shown below, the event tag will have value “t=event”. This tag will have details about the interaction under ec, ea & el which is Event Category, Action and Label respectively.
If you don’t find any event tag on click of a key interaction, then it means that you have to implement the required code track it. Without it, you cannot derive useful insights from your analytics data.
3. Check the custom dimensions
Custom dimensions in GA are variables which can take much granular information about the page or event. The use of custom dimensions is very much required to make the tracking comprehensive. With this data, we can report the numbers with a much granular breakdown. It comes handy when you need to do a deep dive analysis to find the root cause for a problem. For example, if your site has a lead form in all the pages, then while a form is submitted, the page name can be stored in a custom dimension. This help us to know which page had contributed more towards form submission. Though we can get this information through other means, this is the easiest and most accurate way.
4. Look for duplicate firing and junk values
Look for pages or interactions where the tags are firing twice. This could be due to multiple reasons. Most common reason is that the pages will have a hard-coded GA container and the same is fired through Google Tag Manager as well, which could fire the page view tag twice. This double firing will inflate the data and hence the numbers reports are not accurate. It will also impact the bounce metrics.
Also, look for any erroneous values passed on to event variables (ec, ea, el) or custom dimensions. This is could bring down the data credibility.
5. Finally, see where the data goes
Last but not the least, make sure the data goes where it is intended to go. Check the “tid” value which carries the Universal Analytics property id in the format – UA-XXXXXX-XX. It should be exactly the same throughout the site. Also the GTM container id in the format GTM-XXXXXX should also be consistent across. This is to ensure the data is gathered at one place which is key while reporting.
After running through steps 1-5, if you find that your site has multiple issues then they need to be addressed immediately, before you report and analyze the data.