Friday, 30 September 2022

Preparing Your Data Consumers for GA4 — Whiteboard Friday

You know GA4 is coming, and last week Dana took you through some of the top things to be aware of before making the transition to it. In this week’s episode, Ruth Burr Reedy discusses what a lot of marketers may not be thinking about enough: the people besides us who use Analytics data, and what they need to know about Google Analytics 4 in order to continue using Analytics data.

whiteboard outlining four insights into GA4

Click on the whiteboard image above to open a high resolution version in a new tab!

Video Transcription

Howdy, Moz fans. I'm Ruth Burr Reedy. I am the CEO of UpBuild. We are a boutique digital marketing agency specializing in technical SEO, advanced web analytics, and whole-funnel digital marketing strategy. And today, I want to talk a little bit about Google Analytics 4.

GA4 is here

So as, hopefully, all of you by now know, Google Analytics 4 is here. It is the next generation of Google's Analytics tracking software. And what I think at this point, hopefully, most marketers at least know that it's coming. 

The most important thing to know is that on July 1st, 2023, which is less than a year from now, Google Analytics 4 will be the only Google Analytics product that will be continuing to collect data. Universal Analytics, which has been the analytics standard for several years now, will be completely sunset in less than a year.

A lot of people are talking about this. A lot of people are talking about Google Analytics 4, what it is, how to use it. I'm hopeful that the Whiteboard Friday audience already knows a lot of these things. 

But one thing I think a lot of us marketers are not thinking enough about are the people besides us who use Analytics data and what they need to know about Google Analytics 4 in order to continue using Analytics data for the things that they use it for.

Stop! Get GA4 running

So, before we get into that, let's just stop for a minute. Do you have GA4 collecting data on your web properties, on all of them? Do you? Are you sure? If you don't, if there's a website that you own or are involved with in any way that does not have GA4 collecting data, just pause this video. Just pause it real quick and just get GA4 running on that site. It only takes a second. It's actually very, very easy to implement. And all you need to do, at this point, you don't need to configure it, you don't need to do anything else, just get it collecting data. I'll wait.

Who needs this data?

Okay. Ready? All right, great. Here we go. So one thing that we need to think about, as people who use marketing analytics data, is that there are other people besides us who need that data and see that data and use that data to understand what's happening and make decisions accordingly. 

So who are these people? It might be team leadership. It might be your CEO. CEOs the worst, I know. It might be your boss. It might be your boss' boss. Your board of directors, if you're a publicly-traded or a venture capital-funded company or any other kind of company that has a board, chances are you are compiling at least a quarterly report for your board on how the website is doing using Analytics data.

You might also, if you are like me, an agency marketer, you have clients, you have client points of contact. Hopefully, you've already been in close communication with your client point of contact about Google Analytics 4. You've already got it set up for them. 

But chances are your client point of contact has these same people in their work life. So if you are working with a director of marketing, a marketing manager, a CMO, someone in the marketing seat, you need to make sure that that person, your client point of contact has this same information to message to the people above them and the other consumers of Analytics data within their organization who may not be as familiar with how Analytics data is collected but still know enough to be using it and at least know enough to be receiving reports containing this data. All of these people need to understand what is happening with Google Analytics 4.

Implement it now

The time to do that is now. The time to do that is not July 2023. We need to start right now. We have less than a year to get everybody on board with what GA4 is, how it's different, and what that means for the data that they consume and use to make decisions, because it is different. It's going to be different.

I've talked to people who have seen that little alert pop up in Universal Analytics, that strikes a little fear into all of our hearts, Universal Analytics is going away on July 1st, 2023. Some people seem to think of this as when your phone says, "I'm going to install an operating system update overnight." Like cool, great. Phone, you do you. Turn it on the morning. It's not that different. It's fine. That is not the case with Google Analytics 4. And we need to make sure that everyone who uses Analytics data, with whom we interact, understands that so that they are not taken aback when that change happens.

We can also start making that change now gently, iteratively over time, while UA is still collecting data in order to illustrate to our data consumers what the differences are.

Have a plan for historical data

So the first thing that we need to do, when it comes to GA4 and the upcoming transition, is have a plan for historical data. One of the big things about this transition is that historical data in Universal Analytics will not be available via the GA interface after July 1st, 2023. 

So how are you going to get that data? Because chances are you're not going to be okay with just saying, "All right, well, we installed GA4 when it first came out in the fall of 2020. So we've got a little more than a year or two of data, and that's our new data universe." There are companies who have just decided that that's what they're going to do, that's their new normal. KonMari your historical data, namaste, release it into the world. Most people are not going to do that. Most people are going to want to see historical data from farther back than fall 2020, which is the earliest that you could have been collecting this data.

So what is your plan for historical data? Hopefully, you have one. There are many blog posts and videos and articles out there on ways to preserve your historical data. I'm not going to go too deep into that. But whether you're using BigQuery and you're going to port it into some sort of database or data warehouse, maybe you have a small enough dataset that you're just going to export a bunch of spreadsheets and kind of store that, maybe you're going to build some kind of custom SQL database, whatever you're going to do with your historical data, it's none of my business, but you should have a plan to store that data.

Now, at this point, you may already have spoken to these people about the fact that historical data is going away, because that's something that they are probably feeling pretty amped up about. I'm feeling amped up about it. It's a lot, and we need to have a plan. But that's okay, that's what we're doing today.

Don't equate apples to oranges

The real thing that we need to keep in mind, as we're making a plan for historical data, is that the data in Google Analytics 4 and the data in Universal Analytics is collected differently. Even things that have the same name are going to be slightly different metrics behind the scenes. We're going to talk more about that in a minute. But it's important to know that comparing GA4 and Universal Analytics data is always going to be apples to oranges. They're not the same. 

So even as you have a plan to make and store and use this historical database, however you're going to do that, you need to keep in mind that that data and the data you use going forward are not going to be one-to-one. And that's okay. I mean, it is what it is.

Some people, keeping that in mind, are changing what they're doing. Is the expense and effort of data warehousing your historical UA data worth all of the time and expense it's going to take, considering that it's apples to oranges? I can't make that decision for you, but it's something to consider. It's something to ask yourself and really think about what you're going to do with historical data going forward and how you're going to use it.

What's changing? Everything!

Because what's changing with Google Analytics 4? Everything. It's really different. And hopefully, by now, you've gone in there, gotten under the hood, you've played with some of the reports, you've looked at the UI. It's really different from Universal Analytics.

I've been doing SEO since 2006, and this is the biggest change in Google Analytics, especially the front end, the UI that I've ever seen. And also the backend is different. The method of data collection within GA4 is different, because a big part of why Google is making this push for GA4 is in an effort to be more in compliance with data privacy laws. So they're having to change some of the ways the data is collected and reported.

They're also looking at how to better do things like report on cross-domain traffic, cross-device traffic, traffic between websites and apps, when those are the same thing, they're the same and they're different. And now, in GA4, you can look at that data in a more holistic sense.

There's a lot of exciting, cool stuff happening in GA4. But the important thing to know is that things that are called the same thing in GA4 are still fundamentally different and collected at the very least slightly differently than they are in Universal Analytics. This is going to be hugely important when we're looking at this historical data.

A great example is sessions. The session, for many, many years, has been the core unit of Analytics data. GA4, you can tell from their reporting, is really trying to shift everyone's reporting from sessions to users. Both sessions and users are collected slightly differently. They are counted slightly differently. So your session numbers in GA4 and Universal Analytics for the same time period are going to be slightly different.

Now, the degree to which they are different is going to depend on a lot of things. Filtering options in Universal Analytics are a lot more advanced than they are right now in GA4. So if you have a lot of custom filters set up, if you're filtering out a giant known bot network, if you are filtering out data from specific countries, whatever you're filtering out, chances are you cannot implement that yet in GA4. That's going to affect it. But the session itself, how it's counted, when a session resets. So, for example, a session is resetting at midnight, having your time zone configured, hugely important in UA, less of a thing in GA4. So depending on the time period, depending on your filtering, depending on how you're counting sessions now, your sessions data may be a little different or a lot different.

All of this is also going to depend on the scope of your data. Tiny differences become big in bigger datasets. So if you've got hundreds of thousands or millions of sessions in a given time period that you're reporting on, the chances that those numbers are going to be different in GA4 to UA, they're probably going to be different by a bigger percentage.

And, at the same time, if you only have a very, very small number of users, because that sample size is smaller, you may also see bigger gaps. It really depends on your data. 

The important thing is your data consumers don't need to know the ways in which data collection is different. You can tell them and they're not going to remember, and that's okay. They're busy and they don't need to know. What they do need to know is that it's different, it's not the same, and you have a year, at this point, to show them the degree to which it is different so that they can start to understand what the difference between the old dataset and the new dataset is, while you still have those same time periods of data collection to compare. Just to give them an idea of what's different.

So, at this point, you probably have some of these data consumers who are in love with a report. They've got their one report, and they look at it every day or they look at it every week or they look at it every quarter, and you have spent the last, however long you've been reporting to them, refining that report. You show them the report. And then they say, "What about this piece of data?" And you put it in there, and then they never look at it again. You take it out and no one notices. Or you put it in there and it becomes the new normal. Or maybe you have been trying for a long time to get them to look at users, instead of at sessions, but they just love sessions so much as a metric. Whatever it is that your data consumers love about Universal Analytics, chances are it's going to be at least a little bit different in GA4. And it's highly likely, and I would go so far as to say advisable, that that report is going to have to change. So the time to show them that it's different and ease them into that change, like dipping a toe into the Jacuzzi, is now. Not July 2023, now.

Users are another really great example of what is different between Universal and GA4. So in Universal Analytics, we're all used to total users and new users. Those are the two breakdowns of users. In GA4, you have a metric called Active Users, which is the users that have been active on your site in the last 28 days. That is the default users metric that you're going to see in GA4 and in the reporting. Now, you may decide, because you've already been reporting on total users, that you want to report on total users in the future. You can do that, but I would encourage you to look at the ways in which GA4 is presenting and encouraging you to use the data.

It's very interesting to me, this is a little bit of a sidebar, the ways in which Google Analytics, over the years, has taught us what is important to measure based on what they surface up most prominently in reports. And for my career, that has chiefly been the session. Now, increasingly, we're looking at the user, which is great in a world in which most people's purchase journey involves more than one device and certainly more than one session. But it does change the way we fundamentally look at and think about data. And I would encourage you, rather than trying to swim upstream on that, to think about how you are going to change your data reporting in order to mesh up well with the reporting that GA4 is going to roll out, because they're still rolling out new features all the time. You can take a look at what they are surfacing up prominently now to get an idea of where those new features are most likely going to be rolling out, especially in the next year, but even beyond that so that you are reporting in ways that are going to get you the most new cool data soonest. But I digress.

Another thing I want to make really sure that everybody, especially these people, understand is that events mean something completely different in GA4 than it does in Universal Analytics. In Universal Analytics, events are a very specific thing. You collect a piece of data. You have four parameters that you can assign — category, action, label, value. We've all, at this point, used UTM parameters. We know what those are. We're familiar with it. It's comfortable.

In GA4, everything is an event. It's almost going back to like super, super old-school internet days and thinking about hits on your website. At this point, everything in GA4, if you boil it down to a fundamental piece of data collection, is called an event. Could they have called it something different and made this less confusing? Yes, but they didn't, and here we are.

So this is really important to make sure that your data consumers understand that event collection is going to be different. And that's important because of this apples-to-oranges comparison. As you're collecting events data in GA4, it's going to be really, really tempting to try to recreate, as much as you can, your Universal Analytics instance, how you're collecting data, how you're reporting on it. Resist the urge to do that. 

When you're configuring custom events in GA4, resist the urge, and maybe even the pressure from these people, to replicate that category, action, label, value naming convention just because that's what you're used to. Instead, this is a fabulous time to really be rethinking your data collection and your reporting. And we, as marketers, have a huge opportunity here that I want to make sure we don't miss.

Now is the time for data governance

Many of us have come into whatever role we're in now and come into an existing Google Analytics instance. Filters have already been set up. Goals have already been configured. Events have already been set up and have been tracking data for however long before we got there. What this usually means is that things are not set up entirely to our liking. Many marketers, myself included, have come into an Analytics situation and found that the data is incorrect, inaccurately reported. It's double counting things. It's not counting things. We have an opportunity now, with GA4, to make sure that our data collection is complete, accurate, precise, and robust. And we need to seize that opportunity.

And the same thing goes with event collection. Now is the time, for everyone watching this video, to start thinking about data governance. Now is the time for us to seize control of the data and do what we can to not only make it complete, accurate, precise, robust, but also future-proof that data collection for ourselves, for the organizations that we work for, for our clients, and for our data consumers, because we may not be the only people using that data. 

There are often other teams going into Analytics. If you work with a paid search team or a display team or you work with just a general marketing agency who maybe doesn't do anything with Analytics but they look at the data, maybe they don't do anything to the website but they need data about the website because it informs their campaigns, they probably have dashboards configured. They probably have events set up. They may have set up those events in ways that you don't like. Things like, oh, here the label is capitalized. There the label isn't capitalized. Guess what, those are two different events. That's still going to be true in GA4, as of right now at least. Capitalization is still going to make two different parameters. So we have an opportunity right now to enact some data governance, make some rules, and take control. 

So when we think about events in GA4, yes, everything is an event. There are many things that are going to be collected automatically. You do not have to configure GA4 to collect things like page views. They're just going to do that. I don't think you can get them to not do that because it would break the tool. You could, but why would you?

In addition to that, there are enhanced measurement events that Google has available for you to configure. Almost all of those, they're very easy to set up and they're standardized.

The same is true for recommended events. So within GA4, the next level of complexity from automatic and enhanced measurement events are recommended events. And in the GA4 support documentation, there is a large and increasing list of different recommended events and the parameters that they collect that you can look at. I would say, at this point, any recommended event that applies to your site you might as well configure, because you could use that data. 

With both enhanced measurement and recommended events, because they have built-in parameters, Google is going to be using those more to drive some of the ... I know they're wanting to do a lot more with automated analysis and machine learning on datasets. All of that's going to start from the data that is consistent across Google Analytics' broader dataset, which is these enhanced measurement and recommended events. All of the parameters will be named the same thing, so it's very easy for them to collect them and then apply machine learning to them.

You still, as you are setting these up, need to make sure that you're enacting some data governance. You need to make sure the parameters are named the same way, that the same parameter is collected in the same way across recommended events so that you are, going forward, no longer having apples to oranges. That's GA4 to UA. Everything in GA4 should be oranges. Now, get that on a T-shirt, no one will know what it means.

And then the next level up is custom events. Custom events in GA4 are really cool. You can collect data on just about anything. You can pass just about any piece of data that Google Analytics 4 can collect. You can collect as a parameter. There's a ton of functionality, especially if you are pushing things into the data later to collect that as event parameters. We're no longer limited to category, action, label, value. We are limited by the total number of parameters that we can collect per property, which makes sense because data storage is expensive and it's expensive for Google. But we have a lot more customizability when it comes to custom events.

This is very cool. And we really need to apply the Spider-Man principle here. With great power comes great responsibility. Resist the urge to get in there and start tracking everything, partying like it's, I guess, 2099 at this point. Resist that urge. Make a plan. Now is the time for data governance.

As you are thinking about the custom events that you were going to track and the parameters you are going to collect, you might start by just outlining what you know you want to track and how you want to collect those parameters. But then it's time to make some rules, some rules for what you're going to track, how you're going to track it, what parameters you're going to collect, and how those parameters are labeled. You shouldn't just do this for whatever you're going to configure, as you're setting up GA4 now. Think about how you can create and future-proof rules for data collection going forward so that, over time, you get promoted, somebody else is doing your job, you win the lottery and go off to an island and are having a beautiful time. Whoever has your job after you should still have rules so that, when they are setting this stuff up, it is still oranges to oranges and you are creating a dataset that is correct, that is also useful in comparison with itself. Parameters that are useful in comparison across events and event types. This is the time to be doing that. Create those rules, make them clear. Make sure that people on other teams, anybody else who might be setting up events, even outside of the inbound search team or the marketing team, or whatever team you're sitting on, make sure that they have that. If you have clients, make sure you're doing a whole training session with them on what the rules are and how to use those rules to configure events in the future. Make a video. Document it. Share it out. The more you can do now to set yourself up for success in the future, the more valuable your GA4 dataset is going to be from day one and going forward.

So if you get nothing else out of this video, take some time to think about data governance and how you're going to make sure your data is useful and consistent going forward.

Now that we have this beautiful dataset, we're collecting data, it's configured, we've got a year. At this point, I'm filming this at MozCon 2022, we've got a year left to talk to these people about the difference between our apples and oranges and help them fall in love with oranges. So your CEO, your board, whoever it is that loves that report, don't just recreate that report for them with GA4 data. Take some time to talk with them, to understand what it is about that information that they use to make decisions, what it is about that report that they look at that helps them do their job. Find out how to solve that same problem for them with GA4 data rather than just trying to make GA4 look as much like UA as possible, because, over time, it's going to be less and less the case and, over time, people are going to forget about Universal and now you just have a GA4 instance that looks like Universal Analytics for no reason.

Now is the time to not do that. Resist the pressure to do that and figure out what your GA4 install is going to look like in 2023 and 2025, maybe even 2030. Institute those rules now so that you can help your board, you can help your CEO, you can help your clients and their bosses and their bosses' bosses gall in love with your new reports. They're beautiful. They're oranges-to-oranges. They've got new, robust, actionable data that you're using in new, exciting, advanced ways. This is coming. It's happening. Right now is a really important moment in terms of making sure that everybody is on board with what the changes are, in terms of making sure that everybody is on board with how we're going to collect data in the future, and giving everyone a year to fall in love with this new report before they have no other options.

If you have questions about any of this or you just want to talk about Google Analytics 4 and geek out about data collection, please holler at me at Twitter anytime. I'm very friendly, and I love talking about this stuff.

That's my Whiteboard Friday. Thanks, everybody. Have a wonderful Friday.

Video transcription by Speechpad.com

No comments:

Post a Comment

How to Use Chrome to View a Website as Googlebot

The author's views are entirely their own (excluding the unlikely event of hypnosis) and may not always reflect the views of Moz.