One of the most curious things we find in our web analytics agency is the ease with which serious errors are repeated when measuring a website. For this case there is an old saying that today comes to us as a ring to the finger:
“The creique and the penseque are brothers of the tonteque”
In the world of web analytics implementations we have seen many things, others have been told to us and others we have suffered personally (I will not clarify which, out of modesty). Fortunately, once you are aware of these blunders, they become part of a basic checklist that anyone with a minimum of interest should take into account so as not to fall into the most egregious errors that occur in web measurements.
Table of Contents
The top 10 of “penseques”
Here are the 10 most common “penseques” in measurements with Google Analytics:
I thought that putting the Google Analytics code on the web measured everything
The customer asks you to analyze the data of his ecommerce and catalog downloads, and you have to tell him that his Analytics doesn’t measure transactions if you don’t implement them, and nor does it measure downloads,nor objectives (unless they are urls), because it has not made any implementation… Ah, but doesn’t google analytics give that? Yes, but only if you implement it.
How to avoid it:do a basic audit of the account and inform the customer of what your Analytics measures and what it doesn’t.
I thought installing the Google Tag Manager code was enough
This applies to some clients, with supposed marketing knowledge, who ask you to generate a Google Tag Manager code for a new site, which they already take care of, because they already know how to put it alone in the WordPress template. A month later they protest because Google Analytics is empty…They didn’t put a tag inside the Google Tag Manager!
How to avoid it:Whenever a Google Tag Manager code is generated, request notification when it is deployed, and check with the Chrome Tag Assistant extension if the code is working and is not empty.
I thought that after putting Google Tag Manager I didn’t have to remove the Google Analytics code
Usual case. There are people who after hearing the wonders of Google Tag Manager put together the manual to install a tag and launch page views, but forgetting to remove the Google Analytics code on the web. The consequence: duplicate page views (we will see them in the next point).
How to avoid it:Always validate the hits of page views with one of the extensions that exist in Google Chrome: Google Analytics Debugger or GTM / GA Debug and check that the Pageviews do not jump twice.
I thought the measurement of my website was well done
The customer is delighted with its number of pages per session and its very low bounce rate… that’s why he is very happy with himself and with his digital marketing agency… until you tell him he has the duplicate Google Analytics code, and that with your fix you’re going to have to report to your boss that actually, last year they had half as many page views, and that after the fix you’re going to have a lot fewer page views and your bounce rate is going to go up from 1% to more than 60%. Drama assured.
How to avoid it:in the same way as in the previous point.
I thought after putting a payment gateway on the web there was nothing to do in Google Analytics
A client hires a campaign in media, a lot of platforms, Facebook Ads, Linkedin… with a payment gateway by card or PayPal, and contact you after finishing them to know if it has been profitable … and when you analyze it you have to tell him that all his conversion has been generated by the payment gateway (sis.redsys.es or paypal.com usually), with the source of acquisition totally lost because they have not excluded it from the reference traffic of the payment gateway.
How to avoid it:Exclude the payment gateway from the reference exclusion list
I thought I could put in Google Tag Manager a script that I found on the internet
How to avoid it: Google Tag Manager scripts, only from reliable sources, and to be validated by a developer.
I thought that with email marketing campaigns you don’t have to do anything in Google Analytics
You have a customer excited about Mailchimp and the fantastic newsletters that their creative team creates… They are already talking about implementing a “marketing automation”… and when you ask them if they tag campaigns with utm parameters… they don’t know what it is or what you’re telling them. That’s when you enter your analytics account and see that everything is in direct traffic, without any possibility of knowing what traffic the newsletters send, which is the one that has worked best. so much effort to not be able to draw any conclusions because campaigns have not been tagged.
How to avoid this:Send parameter tagging instructions with a link to Google’s Campaign URL Builder
I thought Google would give me the demographics of the users
Google Analytics offers you reports on the age and gender of users, but for this you have to activate demographic data,but many people forget about it and are surprised that then the data does not appear.
I thought that with a filter I improved the measurement of Google Analytics
A customer creates filters indiscriminately in Google Analytics to “improve their measurement” without taking the precaution of leaving a view unfiltered. If the filter is poorly made, the traffic that has been excluded will be lost forever.
How to avoid it:Always have a backup view without filters, so that regardless of the changes that are made, we always have the raw data.
I thought that the developer who made me the web could give me access to Google Analytics
It is very common for web developers to create instead of accounts, properties that hang from their own Google Analytics account, so that they can’t give you full access to your account to, among other things, create filters, or give permissions, because if they did they would expose the data of all their customers.
How to avoid it:Audit accesses and if the problem arises create a double measurement of Google Analytics (making sure to change the Google Analytics tracker)
If you are an analyst, you surely feel identified because surely you have seen with your own eyes quite a few of the errors cited here. One of the things that is clear is that you can never take for granted that the customer knows, even though it seems so. If you have a website, and you are not very sure how you have it measured, it is likely that you identify and can solve some of them … and if you are interested in having a quality measurement, you already know that a digital analyst will have the solution to your problems. That’s what we’re for!