Sign up to save your podcasts
Or
Share 17 ways to improve the quality of data tracked by web analytics tools - Part 1 | Ep. 6
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
17 ways to improve the quality of data tracked by web analytics tools - Part 1 | Ep. 6
Hello, and welcome to another episode of growth detectives this episode is going to be slightly different than the other ones, because it will be divided into two episodes. So today we are going to talk about ways to improve the quality of data that is tracked by web analytics tools. And this is plural web analytics tools, because most of the ideas here will apply not only to google Analytics, but also to tools like Matomo, PeeWeek Plausible Analytics, or many other ones. Okay, so let's start.
The first way to improve the quality of data is simply to check your setup of your web analytics tool. To make sure that all the pages that should be viewed that should be tracked are actually tracked as easy as that.
Number two, use server side tracking. Not all tools let you employ server side tracking, but if you have a proper integration that allows you to do it and your Tracking tool has this option as well. You should definitely use it Why is that server side tracking allows you to collect data irrespective of ad blockers?
So if users have ad blockers these ad blockers will block your tracking scripts. However, if you implement server side tracking then you will still be able to get this data. And another idea. Dear to improve the quality of data is to incentivize people to log into your website.
Why is that? Well, as you, as you know, visitors can come to your website from different devices. At first somebody can, google your website on, on their mobile phone. Next, they can visit it using their laptop or company computer. So, you never know if somebody who comes to your website is a new visitor or a returning one.
However, if they log into your website, then you will be able to send to the tracking services the ID of this logged in user. And the tracking service will be able to match the sessions from users of the same ID. This way you will be able to more accurately say whether someone is a new user or a returning one.
Okay. Another way to improve the quality of tracked data is to rename and proxy the tracking scripts. This is another method to, kind of avoid being blocked by ad blockers because ad blockers, they, as I've already said, they block tracking scripts. So, if you download the tracking script directly to your server and rename it, then ad blockers will no longer see a Google analytics script that gets downloaded from Google analytics website, but they will see some randomly named script that comes from your own website.
So the chance is now much smaller that this script will be blocked. Another way to improve quality of tracked data is to use UTMs. Whenever, wherever you can. UTMs are parameters that are added to the end of the URL A URL that can be shared or pasted in your, I dunno, social media posts.
They can be put in your newsletter in PDF files , things like that. Yes. And information that is saved in the UTM can be read by your tracking service.
And this information can for example, say that this particular user clicked a link in a PDF.
So another way to improve the quality of track. Data is not to use the default form tracking in Google Analytics.
The thing is that Google Analytics 4 introduced a few automated ways of tracking some user actions. One of which is form tracking or actually tracking form submissions. The problem with this solution is. That is incredibly bad and it gives a ton of false positives. So it's much, much better if you simply stopped using it and started using some customer made solution.
And the last way to drastically improve quality of tracked data that I'm going to talk about in this episode is to fix the incorrect labeling of traffic that comes from Android Applications. And this is a big one. I, I left the, the, the biggest and the most interesting one until the end.
Okay. So let me, so let me explain what I mean by that. The thing is that when somebody clicks a link in an Android application, for example, for the LinkedIn or Facebook or Facebook messenger or Snapchat, yes. Then their browser opens the link. However the information that says what or who referred the traffic to your website is not recognized by by tracking tools.
It's because this information, this referral information does not start with the usual HTTP, but it starts with Android app. The text android app and now depending on On the profile of your visitors you may be getting a ton of traffic from Android applications. And all of this traffic may be labeled as direct traffic.
Yes, and I will give you an example here from two of my clients. In the case of Juan, about 14 percent of the traffic he was getting came from a Google search widget in Android phones. Yes. 14 percent of traffic. I mean, of direct traffic labeled as direct traffic. That's huge, right? So this information could have been labeled as Google search traffic, but it wasn't.
It was labeled as direct traffic, which was of course. incorrect. And another one of my clients has quite a lot of traffic from social media applications and especially Pinterest. So it's turned out that also about like 14, 15 percent of his traffic was coming from a Pinterest application.
So after adding a small script to his website I was able to correctly label all this traffic to Pinterest. Right? And what is interesting is that now we are talking only about Android applications. Yes, but we are not talking about iOS. Why is that?
The thing about iOS is that applications on iPhones do not send any referral information. So We are now able to get the information from Android applications, but we are not able to get any referral information from iOS applications. However, with very simple maths, you can more or less accurately say how much traffic you are getting from the equivalent iOS applications.
Simply. By looking at how much traffic you're getting from iPhones in comparison to Android. So for example, if you are getting twice as much traffic from iPhones than Android, and you are getting like let's say 10 percent of traffic from Pinterest Android application, then it would be safe to assume that about 20% is from iOS, so in total.
You will be getting 30, 30%. So this is, this information is very, very important. And there are actually two easy methods to get this information and to have it visible in Google Analytics, in Matomo, in Plausible, in Pwik in other systems.
However, it's impossible to correctly describe the. process, how to get this data in a podcast. That's why I will soon be recording a YouTube video about it. And I will add the link to this YouTube video, to the description of this podcast episode.
So this is it for this episode of growth detectives. Tomorrow or in two days, I will be posting another episode telling you about all the other methods to improve quality of data tracked by web analytics tools. Thank you.
View all episodes
By Krzysztof PlanetaHello, and welcome to another episode of growth detectives this episode is going to be slightly different than the other ones, because it will be divided into two episodes. So today we are going to talk about ways to improve the quality of data that is tracked by web analytics tools. And this is plural web analytics tools, because most of the ideas here will apply not only to google Analytics, but also to tools like Matomo, PeeWeek Plausible Analytics, or many other ones. Okay, so let's start.
The first way to improve the quality of data is simply to check your setup of your web analytics tool. To make sure that all the pages that should be viewed that should be tracked are actually tracked as easy as that.
Number two, use server side tracking. Not all tools let you employ server side tracking, but if you have a proper integration that allows you to do it and your Tracking tool has this option as well. You should definitely use it Why is that server side tracking allows you to collect data irrespective of ad blockers?
So if users have ad blockers these ad blockers will block your tracking scripts. However, if you implement server side tracking then you will still be able to get this data. And another idea. Dear to improve the quality of data is to incentivize people to log into your website.
Why is that? Well, as you, as you know, visitors can come to your website from different devices. At first somebody can, google your website on, on their mobile phone. Next, they can visit it using their laptop or company computer. So, you never know if somebody who comes to your website is a new visitor or a returning one.
However, if they log into your website, then you will be able to send to the tracking services the ID of this logged in user. And the tracking service will be able to match the sessions from users of the same ID. This way you will be able to more accurately say whether someone is a new user or a returning one.
Okay. Another way to improve the quality of tracked data is to rename and proxy the tracking scripts. This is another method to, kind of avoid being blocked by ad blockers because ad blockers, they, as I've already said, they block tracking scripts. So, if you download the tracking script directly to your server and rename it, then ad blockers will no longer see a Google analytics script that gets downloaded from Google analytics website, but they will see some randomly named script that comes from your own website.
So the chance is now much smaller that this script will be blocked. Another way to improve quality of tracked data is to use UTMs. Whenever, wherever you can. UTMs are parameters that are added to the end of the URL A URL that can be shared or pasted in your, I dunno, social media posts.
They can be put in your newsletter in PDF files , things like that. Yes. And information that is saved in the UTM can be read by your tracking service.
And this information can for example, say that this particular user clicked a link in a PDF.
So another way to improve the quality of track. Data is not to use the default form tracking in Google Analytics.
The thing is that Google Analytics 4 introduced a few automated ways of tracking some user actions. One of which is form tracking or actually tracking form submissions. The problem with this solution is. That is incredibly bad and it gives a ton of false positives. So it's much, much better if you simply stopped using it and started using some customer made solution.
And the last way to drastically improve quality of tracked data that I'm going to talk about in this episode is to fix the incorrect labeling of traffic that comes from Android Applications. And this is a big one. I, I left the, the, the biggest and the most interesting one until the end.
Okay. So let me, so let me explain what I mean by that. The thing is that when somebody clicks a link in an Android application, for example, for the LinkedIn or Facebook or Facebook messenger or Snapchat, yes. Then their browser opens the link. However the information that says what or who referred the traffic to your website is not recognized by by tracking tools.
It's because this information, this referral information does not start with the usual HTTP, but it starts with Android app. The text android app and now depending on On the profile of your visitors you may be getting a ton of traffic from Android applications. And all of this traffic may be labeled as direct traffic.
Yes, and I will give you an example here from two of my clients. In the case of Juan, about 14 percent of the traffic he was getting came from a Google search widget in Android phones. Yes. 14 percent of traffic. I mean, of direct traffic labeled as direct traffic. That's huge, right? So this information could have been labeled as Google search traffic, but it wasn't.
It was labeled as direct traffic, which was of course. incorrect. And another one of my clients has quite a lot of traffic from social media applications and especially Pinterest. So it's turned out that also about like 14, 15 percent of his traffic was coming from a Pinterest application.
So after adding a small script to his website I was able to correctly label all this traffic to Pinterest. Right? And what is interesting is that now we are talking only about Android applications. Yes, but we are not talking about iOS. Why is that?
The thing about iOS is that applications on iPhones do not send any referral information. So We are now able to get the information from Android applications, but we are not able to get any referral information from iOS applications. However, with very simple maths, you can more or less accurately say how much traffic you are getting from the equivalent iOS applications.
Simply. By looking at how much traffic you're getting from iPhones in comparison to Android. So for example, if you are getting twice as much traffic from iPhones than Android, and you are getting like let's say 10 percent of traffic from Pinterest Android application, then it would be safe to assume that about 20% is from iOS, so in total.
You will be getting 30, 30%. So this is, this information is very, very important. And there are actually two easy methods to get this information and to have it visible in Google Analytics, in Matomo, in Plausible, in Pwik in other systems.
However, it's impossible to correctly describe the. process, how to get this data in a podcast. That's why I will soon be recording a YouTube video about it. And I will add the link to this YouTube video, to the description of this podcast episode.
So this is it for this episode of growth detectives. Tomorrow or in two days, I will be posting another episode telling you about all the other methods to improve quality of data tracked by web analytics tools. Thank you.