#83 GA is dead, long live GA!
In this week’s episode of The Measure Pod, Dan, Dara and Bhav delve deep into the transition from Universal Analytics (UA) to Google Analytics 4 (GA4). They reflect on their personal experiences with UA and express their mixed feelings about its departure. They also discuss how GA4 is perceived as an upgrade, but in reality, it’s a completely different product! The conversation touches on the fear of change and the initial challenges associated with learning GA4. As the sun sets on UA, hear our hosts discuss the implications, opportunities, and challenges presented by GA4, and shed light on its nuances and what it means for the analytics industry.
Join Dan’s BrightonAnalytics community and come along to the next in-person social event!
Intro music composed by Confidential – check out their lo-fi beats on Spotify.
Master Google Analytics 4 with Daniel Perry-Reed on the next GA4 Immersion 6-week cohort training course. Early bird, charity and group discounts availible!
Quotes of the Episode:
- “One of the biggest revelations and kind of changes I went through in my career working with sort of web analytics and Google Analytics specifically, is the reality of how unimportant it all is, and just acceptance of that.” – Dan
- “The people who are going to be most surprised with it are going to be the ones who sporadically see the data… They’ll be like, hey, why have these numbers changed?” – Bhav
- “UA is dead. Long live GA4” – Dara
The full transcript is below, or you can view it in a Google Doc.
[00:00:15] Daniel: Hey Dara, hey Bhav. I have brought you here today. Thank you for coming. This is a big moment. But it’s not an analytics podcast unless we talk about the elephant in the room, which is the sunset of Universal Analytics. So this isn’t our ‘why you should care about GA4’ kind of podcast. And if you’re playing the drinking game, the measure pod drinking game, please don’t because I feel like the terms like Google Analytics 4 and Universal Analytics and attribution and all that kind of lovely stuff is going to come up a lot during this.
[00:00:38] Daniel: But Universal, at the time of recording, at least, we’re a week after the 1st of July. So we are a week after the official sunset of Universal Analytics. But what I wanted to talk to you today about is not about why Universal has gone, what GA4 is and why it’s not so bad or crap or great or everything in between, but more around like what does this actually mean now that we are the other side of it. Small print there that, you know, a lot of UA properties are still tracking it at this moment in time, at very least. Now we’re the other side of it, what’s your thoughts? What’s changed? Has there been any reaction? Have you seen anything happen? Has the world stopped revolving around the sun? Has your world come to a grinding halt or is everything kind of business as usual still?
[00:01:15] Dara: I mean, I’d be more interested to gauge how general users are finding it once the dust has settled a little bit. Because we do tend to be a little bit inside our kind of filter bubble where the people we follow and the people we talk to there are other people like us. And Dan, we’ve been talking about GA4 and UA being sunset for so many episodes now it’s very easy to believe that it’s the biggest thing in the world. So I’d be, yeah, I’m going to be keen to see what the general consensus is and I have a feeling it’s going to be okay, great, it’s GA4 now. What was Universal Analytics? Oh, yeah that thing from the past, I’ve forgotten about that.
[00:01:48] Bhav: I mean, I think from my point of view, I’m sad, right? For a lack of a better word. I’ve loved Universal Analytics. I’ve not had to deal with it in a professional sense for a number of years now, because the majority of the companies I’ve worked at and the teams I’ve been using, the more product analytics based tools, but there’s always been some element of Google Analytics that’s stayed part of my life. We’re running in parallel, an instance of Universal Analytics along with product analytics tools, just so the marketing teams can capture their marketing data better. But also on a personal note, like all of my websites have always been Universal Analytics based, and I always dig into the data. So part of me is feeling kind of sentimental today, I don’t have any anger towards GA4, today I’m just reflecting quietly and calmly on how much I’ve loved UA and how much it’s meant to me personally.
[00:02:34] Dara: You know, it’s funny because I feel like this might surprise Dan, but for so long a kind of internal joke here was about how, you know, I was so in love with Universal and it was, you know, a bit like what some of the blogs I’ve read recently from people who’ve been working in this space for years. A bit like them, GA kind of made my career, but weirdly, I don’t know if my grief period is just shorter, or maybe I’m just not sentimental. But as of Friday when the date officially kicked in, I’m now like, right Universal’s gone. Let’s stop talking about it, let’s stop thinking about it. It’s GA4 now, that’s the only thing we have.
[00:03:08] Dara: So maybe I’m just admitting to being very unsentimental here and maybe I just am cold and clinical. Heartless, yeah, heartless. I feel like my grief was very brief and it was probably when GA4 was very new and scary and I didn’t know anything about it, and I thought, oh, this is all going to change. But we’ve had, I mean, what is it, two years, two plus years now, and I think, you know, get over it. Universal’s gone, good riddance.
[00:03:36] Daniel: We’ve had it for over three years technically. It came out of beta, quote unquote beta. Who knows what that actually means in hindsight, but came out of beta in October, 2020. So we’ve had almost three years of playing with it. We’ve had a little bit of early access when it was App + Web. I mean, the thing is, and without going into the detail with this, I think there’s plenty of content out there to talk about this, but you know, the features are always rolling out and it’s playing a bit of catch up in terms of feature parity, but it also kind of goes beyond in some other aspects. I think this is one of my relatively recent understandings of this, that GA4 is not an upgrade to Universal Analytics. And I’m not saying it’s a downgrade either. It’s just a different product.
[00:04:08] Daniel: It’s the equivalent of going, implementing I don’t know, Adobe on your website and saying, oh, I’ve upgraded from Universal Analytics. It’s like, it’s not quite the same. You’d never say it in that sentence. The fact that the logo is the same and it’s still called Google Analytics, or at least it’s in the name somewhere, means that people consider it as an upgrade and obviously they’re switching one off, which makes it feel that way.
[00:04:24] Daniel: And I think a lot of the collateral says, have you upgraded yet? But an upgrade feels like it has to be better in some way. And it’s only, again, I don’t want to go into the nitty gritty. I think we’ve got a lot of content content on this already, but it’s only better to Google, the only people that benefit from us using GA4 is Google, because it’s a way to make sure that the measurement stays consistent from an investment in marketing, right, that’s the key thing. It’s not intrinsically better, I don’t think anyone was going Universal is so shit, let’s have a new product, please. Yes, it might be better in hindsight. We talked about that with you Bhav, as well, about it being more product based, event based and blah, blah, blah, blah, blah.
[00:04:57] Bhav: I mean, I was going to say like, it’s often the case. We cling onto the past, right? So maybe I’m entering into the philosophical view of this. We cling onto things from the past. We know that we’re going to adapt and we’re going to just get used to it and we’re going to get on with our lives. I think there’s been so many instances, like if I go back to time, when I was a Facebook user over the years I saw, you know, you saw Facebook change so much and the timeline changed a lot and going from the first version to the next was always a bit like, oh, I’m never going to use this. And you kind of just suck it up and you just kind of like adapt to it, I think people like fear change.
[00:05:28] Bhav: I think the thing I’m most worried about, and I think this is maybe the general concern that I’ve basically seen in the industry based on my LinkedIn feed, is that where Google’s always been powerful is that it’s so intuitive to use as a platform, and I think the move to GA4 requires a lot more brain power to initially get started. And I think that fear of taking that step and becoming used to this new tool is probably what we’re seeing. People feel like they’re going to lose the thing that they’re comfortable with, they know how to use and they’re going to have to like develop, you know, in some ways it’s kind of like a career change, right? Like changing your career is a scary thought. You’ll be fine eventually but the initial weeks and months are always a little bit scary and they’re always full of imposter syndrome.
[00:06:14] Daniel: Is it just intuitive because it’s been around for so long? Because I think the instance of Google Analytics in whatever form has been around for what, 10 plus years. And I think it’s only intuitive to use because it’s been there for so long and a lot of people have built, like you said, their career using that tool and I think anyone that’s new to Google Analytics, I mean, speaking from experience, just doing Google Analytics training, even back when it was Universal Analytics and people were fresh into it, it’s overwhelming as fuck. Like there’s so much in there and it’s so hard to comprehend, especially this idea of like, sessionization, users, hits, events. Teaching Google Analytics 4, maybe it’s just me being better at teaching it now, but I always found it’s an easier access point. So I don’t think it’s any less intuitive.
[00:06:53] Daniel: Like change is difficult and I think if you’ve learned to natively just work around things or to know how to handle things or know where to go to get something, we’re basically saying that has to change. But I don’t know if it’s inherently less intuitive.
[00:07:04] Dara: We need to get a group of people that have never used Universal or GA4 and train them in both and find out which they think is more intuitive. I mean, I’m joking, but you’re kind of right Dan, this is where I wonder about the filter bubble as well, because pretty much anybody working, not just in analytics, but in marketing, maybe a lot of people from product as well, people with web dev backgrounds, they know GA, they know what they see as GA, which is Universal Analytics. So then when you’re asked to use this completely new tool, it’s that whole thing of, it’s like, you know, you talked about Facebook and every time they do make a change, everybody goes absolutely crazy. And then two weeks later, they’re fine and they can’t remember how it used to be.
[00:07:43] Dara: It’s a bit like that, it’s familiarity and it’s it’s fear of this new platform. But I know you’ve said this before, Dan, like when you train people, it’s easier to train them on GA4 when they haven’t used Universal before, because your mind is automatically thinking of how it’s different from Universal. Whereas if you haven’t had that experience, then you’re kind of a blank canvas and you can just learn it from scratch and not have some of the hangups that us relics that have been using UA for so long have.
[00:08:09] Daniel: Just this year to date I’ve trained over, believe it or not, over 700 people on Google Analytics 4 I know, I was just doing, I just had another tab open I counted, and I honestly couldn’t have guessed that. But anyway, we’ve been doing a lot of Google Analytics training, and you’re absolutely right, like teaching people Google Analytics 4, I find a lot easier because it’s simpler and you don’t have to be like, oh, this works this way just because, just understand and move on. Like you don’t have to teach a kind of, oh, I don’t understand it properly or Google’s decided this and I don’t know why, let’s move on. Yes, there’s different aspects like that you have to teach them, but as a process, it’s like, oh, events, everything’s events, here’s your events, you know, and here’s users. Users aren’t people.
[00:08:44] Daniel: You know I think that’s the kind of the two concepts, and I kind of move on. But in terms of change, I think this is the key thing, and I don’t blame Google because Google’s trying to keep the lights on from their perspective, right? Yes, they’re a big behemoth company, marketing agency, you know, they could be good or evil it depends on what sci-fi future you predict happening. But the thing about Google, they’re trying to keep the lights on and they’re seeing a change in the industry, they’re seeing a change in legal changes, you know, policies, laws that are coming into place like GDPR, ePrivacy, but also changes in like the walled garden ecosystem like Apple building their walls higher and higher and granting less trackable stuff.
[00:09:15] Daniel: And they’re just saying, well, we need to adapt, we need to change. And you know, who knows if GDPR didn’t come into play, if Apple didn’t make all the changes that they’re making, who knows? We might not have seen a GA4, at least we might not have seen it being so dramatically different in a lot of different ways. I think it’s a reaction rather than an action. I mean, Google doesn’t really act first in any way, right? Like they’re always reacting a couple of years behind everyone. And yet, because it’s Google, it’s the biggest company ever that people think it’s the, you know the first company to do it. And actually they’re just, you know, following shortly behind.
[00:09:46] Daniel: We talk about like the event schema being like wildly new and exciting and different for every web analyst or marketer to use. But I mean, Bhav you can speak to it. I mean, how long has the idea of event-based data been a thing right? In the product world or the app world, it’s been there since almost day dot. It’s just for this subset of people, it’s new and they’ve got loud voices and there’s a lot of them, right?
[00:10:07] Bhav: Yeah it’s really funny to say this today when we’re actually recording this. Facebook have just launched Threads, so I don’t know if you know, Threads is a direct competitor to Twitter, and most people who have Instagram accounts and Facebook accounts and are part of that Facebook ecosystem, they’re probably going to adapt to, you know, they’re probably going to start using Threads, having always, you know, and having never used Twitter and suddenly, and you’re absolutely right, you know, they’re going to suddenly be these Twitter users and Facebook aren’t innovating in this space. Twitter’s been around forever, but for the users who have never used it, it’s going to feel fresh, it’s going to feel new, it’s going to be sharing thoughts with our images and videos. It’s just like, let me put some thoughts out there. Like Facebook aren’t first movers either, right? Like they’ve done their first moving many years ago and now they’re happy to just play catch up, wait to see what works, and then just do it and just like rip it off.
[00:10:51] Bhav: And because they’ve got such a massive audience like Google, they can just tap into the fact that, okay, we’re this big company, we already have millions of users. We’ve launched this, and everyone’s like, oh, wow, this is amazing. Schemas, Threads, posting stuff that aren’t images and videos. So, yeah I agree. One thing I will say is actually if anyone is listening and fearful about GA4 and the migration to GA4 and, and you know, like what the world’s going to look like, I always try and break problems down into their sub components. So right now, GA4 probably seems like this absolute mammoth of a problem that you need to overcome, break it down into a subset, that’s how I get through life basically.
[00:11:30] Bhav: If you think about like the fact that, forget the tools and the changes and tools and the events and you know, all of this new terminology and stuff that you need to know, think about what you want to do, right? Your use cases for GA4, you know, from UA to GA4 aren’t going to change. You want to build funnels, you want to see how your marketing is doing. You want to see how your pages are doing, you want to see how many visitors you’re getting to the site. You know, all of these things aren’t going to change. If you break it down one by one. Say, okay, let’s figure out how to build a funnel in GA4. Check, done that. Okay, now I need to build a segment on top of this. Check, done that. And I always find that if you can break down the problems into much smaller parts, it’s a lot less scary.
[00:12:09] Daniel: For sure, and I think, you know, I would hazard a guess in saying like, at least 70% of all users of Google Analytics are there just to look at traffic by day, by channel at most. Right? And I think, you know, the reality is looking at that very simple, from our perspective at least I’m not looking down upon it at all, but it’s like a very simple use case that’s not going to change. And actually any tool you pick up is going to do that. So just go with the one that’s free, easy to use and accessible to you in any meaningful way, which predominantly is probably going to be Google Analytics because, they’ve cornered the market by going free back in 2005, right? They completely wiped that out. And GA4 is still free and it still does fundamentally the same thing.
[00:12:42] Daniel: I think going into those intermediate and more advanced use cases, that’s where the amount of blog posts and the conversation, yes, they’re within our bubble. But, you know, that might extend outside of that is like, you know, why go GA4? Why not go to any other analytics tool, you’re quite right. Why not? Like there’s no reason you have to go to GA4. For me, I think there’s three reasons why even if you move away from GA4, you’re probably going to end up back on GA4 eventually at some point, even if it’s down the line, even if it’s not immediately.
[00:13:08] Daniel: One is like you said, it’s free, right? You can’t get away from something that’s free. The value of it being free is immeasurable actually, because trying to get budget signed off in any company is going to be anything from slightly tedious to a nightmare. And so if you can do something for free, then you can do it straight away and you’re not sort of beholden to gatekeepers, budget holders and things like that.
[00:13:27] Daniel: Number two is that you can’t get away from it being really good in terms of connecting to the Google Marketing Suite. So a lot of people, a lot of companies out there, their biggest marketing channel spend wise is going to be Google Ads and the Google Marketing Platform. And so the way that’s going is that you’re going to have to implement GA4 to do all the measurement of that, so you’re kind of stuck if you don’t have GA4. And the third thing is that it won’t happen straight away, but it’s something that will happen down the line is that everyone will know how to use Google Analytics. So right now we’re all struggling to learn GA4, but that’s temporary, right? Like Universal Analytics is a crazy hard thing to learn, but everyone knows it, right?
[00:13:58] Daniel: Whereas I think going into GA4 at some point in the future, everyone’s going to know it. So you’ve got this easy access point, huge amounts of community resources and documentation and videos, free training, paid training loads of stuff out there to help you learn it. Everyone’s going to have an understanding of it, it’s free and you kind of have to use it anyway if you’re marketing within the Google Stack. So like I say, those three reasons are the key reasons you come back. I highly recommend looking if you are not thinking of this from a marketing perspective, if you’re purely thinking it from a data perspective. For sure, there are hundreds of other avenues for you to go down, opportunities, you can even just do your own data collection, use something like server-side Tag Manager, and send those hits straight into BigQuery or AWS or something like that. You don’t need an analytics vendor, you can just do it yourself.
[00:14:39] Daniel: But if you’re there for the marketing analytics, so you’re there to connect it to your Ad platforms and to activate through audiences, A/B testing tools, if you’re using any of these tools and you want to plug and play, then you’ve got an easy, free and a quick route to do that.
[00:14:50] Dara: The thing is as well, if anybody tries to argue with that logic, all you have to do is look back in time, like the evidence is already there that that’s the case because all of those same three reasons applied for Universal as well. So when people would use Adobe Analytics or another analytics tool on their site. Very often they would also have GA on there and you’d have people within the organisation who would say I don’t really know Adobe, so I’m just going to use GA for that. So those kind of three points, it being free, it integrating with the Google Ads platforms and everybody knowing it. It’s going to be a little lull maybe when that’s not the case for GA4, but very quickly that’s going to become the norm again. And even if somebody decides, at least someone being a business decides to move away from GA4 and use a different product. It’s only a matter of time until people change within that organisation and the new people say, we know GA so we’re going to put GA on there, even if it’s just in parallel to begin with.
[00:15:48] Dara: But I wouldn’t be surprised if those companies end up relying more and more and more on GA4 again as time goes on, because we’ve seen this, it happened exactly the same way with Universal. So I don’t see why it would be any different with GA4.
[00:16:02] Bhav: I think we can all agree like the underlying data collection is for a lack of better word, better than the actual Universal Analytics methods, right? Like moving to an event-based model is a lot more powerful. So I think the big hang up or concern is really around the main platform and I’d love to hear from you guys, like what do you think, like is the new platform in a good enough state or is the idea that you send that data into more powerful platforms like Looker Studio, BigQuery, and you utilise the power of those platforms to do more stuff with, and actually, Google have built the interface for GA4, you know, just to be a good enough state.
[00:16:40] Dara: I’ve often wondered about that. Dan, I know you’ll have probably a lot more to say about this even, but when GA4 was first on the scene, I really strongly thought that that was the case. That they were intentionally making it very basic so that you would actually use well, what was at the time Data Studio now, Looker Studio, or even get the data into BigQuery and work with it in there and then feed it into other BI tools or whatever. I’m less convinced now and I’d love to hear what you think about this, Dan. Because I think with things like the explorations, it’s almost like, maybe, I don’t know, are they hedging their bets where the core interface is maybe a little bit basic, but then they are going to build out more analysis tools for you to use if you do want to do it within the interface itself?
[00:17:19] Daniel: Well they’ve recently coined a term for these different places of accessing data called the reporting surfaces. And I kind of have a little bit of a sick come up in my throat when I call them these, these kind of terms. There’s a thread on the MeasureSlack, and lots of other people have jumped in with all sorts of funny terms, but let’s just use the terms. They’re reporting services and basically there’s three reporting services you can report via the UI. So they log into GA4, and that’s the same as the API, by the way. So the UI and the APIs, the, the same surface, whether you’re doing that through Looker Studio, Tableau or whatever.
[00:17:48] Daniel: The other surface is the exploration workspace as you said Dara. And the third surface is the BigQuery data set. Now there are three versions of the data in GA4, and they’re going to give you slightly different numbers, and I honestly find it an uphill battle to try and reconcile between those three. They’ve all got very different aggregations, they’ve all got very different sort of data structures and it’s a bit of a nightmare. But the key thing for me, and this is, believe it or not, I do have an opinion on this, but the thing around these surfaces, these reporting services is they’ve got different audiences in mind and it’s changed from Universal Analytics.
[00:18:18] Daniel: So Universal Analytics I’ve always explained as a data tool with a plugin for marketing. So you can build segments and then you can publish, quote unquote, publish them as an audience and sync those to your Google Ads property. Whereas now, the UI and the API is designed for marketers. It’s designed as a marketing tool. Everything you do in there directly affects your Google Ads and your Google Marketing Platform products like SA360, DV360 you can even now create audiences from Google Ads in GA4 without logging into GA4. So it is like the UI side of it, and I suppose through an extension, the API is a marketer’s surface, it’s the way to access bits of the data. Like you don’t have all the data there, you have to set up custom dimensions if you want to keep stuff there. It’s not going to be complete, it’s going to be modelled, estimated and lots of different modelling and things going on there, but it’s basically like a marketer’s view of the data.
[00:19:04] Daniel: Now the BigQuery data is the analysts version of the data. So this is the raw data you get, even the unconsented data out there through Consent Mode, you’re going to have data in a table that you can play around, do all that stuff. Now, it doesn’t give you the nuances of the modelled data, the joinings that the Google kind of layers on itself. It doesn’t do the predicting, it doesn’t do a lot of the other stuff, but you do that yourself. And so if you want the raw data, you’re an analyst, you can build that stuff, connect it to your own data. You’ve got that, but you lose a bit of the stuff that you get from Google’s kind of layering on top of it over here.
[00:19:31] Daniel: Now the exploration workspace is almost like the bit in the middle. I think of it as a line, you go from the basic sort of standard reports to BigQuery and you meet in the middle. And this is a way of looking at the raw underlying data within GA, not the BigQuery table that you have, but this is the raw data behind the scenes in Google Analytics. And this is where we are kind of dependent on things like data retention. So potentially for some properties you’ve only got 2 or 14 months of data in here and we can do all sorts of clever stuff there. And it’s like where you were saying Bhav, building funnels and you can do fancy data viz without having to learn SQL. You can do some data segmentation without having to be kind of like an expert in sort of python or R. And so this is a way where you can dabble, you can dabble with this stuff and start building some cool visualisations. Now I think they’re going to expand that over time, but I think this is where you’ve got those three audiences.
[00:20:15] Daniel: You’ve got the kind of the marketer’s view, which is the the UI and API. You’ve got the analyst’s view or the data scientist’s view, which is the BigQuery dataset. You’ve got the kind of bit in the middle, which is for both sides to kind of meet and kind of dabble. Like for example, me, I’m not a SQL analyst, but I know how to use the explorations really well. It’s a really good tool for me, but I couldn’t say that for some of my colleagues at Measurelab that are SQL analysts, they probably would never go in and use the UI, right? It’s just you kind of stick to what you know, and then you kind of have this overlap in the middle. Now, I’ve spoken for a long time about this and I just wanted to kind of state that, you know, I can’t deny that I, and Measurelab, have profited from the changes to Google Analytics 4 Right. Because we are there as an agency using things like Google Analytics, training it as well, but we’re also not affiliated with Google. You know, we’re not like in their pocket and we don’t have quotas to fill and things like that, which I think is really important to state.
[00:21:01] Daniel: So I’m no way saying, oh, it’s a great tool, it’s perfect and it’s better than everything else. But I think it’s an important tool and I think everyone’s going to be using it because of those three things I said earlier. And I think the biggest change is this idea of who’s the user, who is the target audience of GA4? And actually in Universal Analytics, it was like a bit of a data tool, right? You know, not very many marketers would log into UA, let alone, you know, use it to its full extent. They’ll log in, get the bits they need, often they’ll just sit in a Data Studio dashboard for the most of the time, whereas that’s changed. Now, logging into GA4 is really important for marketers to be able to do, because you almost need to be a dual admin now between Google Ads and Google Analytics 4 just to get the most out of it.
[00:21:36] Dara: I’m going to hit you with a straight yes or no question. Actually, before I do, I want to say I’m glad you proved me right, I said you’d have a lot to say about that and you absolutely did I teed you up nicely. Right, straight yes or no question. So you said earlier, it’s not an upgrade it’s a new product. And you just said there, you made the point about, you know, not being kind of beholden to Google and you’re helping people when they decide that there’s a tool they want to use. I’m reading that as you are, you’re skirting around something, which is, is GA4 better than Universal Analytics? Yes or no?
[00:22:08] Daniel: Ooh, I was really hoping you wouldn’t ask me outright.
[00:22:11] Dara: The big one.
[00:22:13] Daniel: Look I’m not going to answer it straight away. I mean, I have to say that it depends is a shit answer here and it’s not satisfying at all. But I think reacting to changes in browser habits, technology, legal policies and things like that, yes. I think it is doing a better job at representing modern behaviour in data and providing access to that data via these different surfaces, you know, like the BigQuery dataset, the event schema. And I think, yes, it is a better version. It’s an upgrade, GA4 is better. Now it’s partly due to that, but also some of the nuance around you can’t validate data anymore, it’s a big old guessing game. Every single fucking number in GA4 is modeled and manipulated and touched by Google in some way. And for that, I would say that, no, it’s not better, because there’s no possible way of doing any data validation anymore. I can’t ever tell you if something is correct, I just have confidence that it’s set up correctly and then that’s the data. And I can’t tell you, Google doesn’t make it very clear if they’re actually modelling data, like conversion modelling or behavioural modelling. So like all of this stuff that’s going on behind the scenes I think is quite, not disingenuous, but I just feel like they’ve approached this from a very black box perspective and that is not the case in Universal Analytics as much.
[00:23:22] Daniel: And now my tune might change on that over time as it becomes more normal. But I would say that the kind of raw quality, viewable confidence in the data, I would say no, it’s not an upgrade at all. But I think, like I said before, I think that’s where the BigQuery stuff comes in, just I should be using that if that’s my problem.
[00:23:36] Bhav: And I think I was going to say this before Dara’s question was if you’ve got these different views of the data, you’ve got your marketing view, you’ve got your data science and analytics view. And you know, if the analysts and the data scientists are required to build their own models, they’re going to end up in a place where they create different numbers right? Now historically, this has always been such a painful point in any organisation, which is, which number do you use? Do you use the one that Google’s provided or do you use the one that’s, you know, your data scientists have built or your analysts have built and, okay in the grand scheme of things, you know, like you’d have to like get it completely wrong to have the numbers be completely different, but even those minor nuances when you’re talking about things like CPA (cost per acquisition), you know, if you’ve got a, let’s say, you know, you’ve got a target of a £5 CPA, right?
[00:24:22] Bhav: If your data science model creates a CPA that is £5.50. And then you’ve got your Google model, which shows your CPA as like £4.50, right? The differential is quite large. And you’re going to have people like, we’re paying more than we should. Or, actually, no, we’re not, we’ve still got room to spend more. Like these nuances are the things that I worry about. Like I much prefer to have a world where everyone’s looking at the same number for better or for worse, as long as it’s the same number, everyone’s dealing with the same degree of error.
[00:24:58] Bhav: So yes, you know, GA4 is, and by the way, I’m just playing devil’s advocate. I like the idea that you’ve got data for the marketers, you’ve got data for the analysts, but is that not just causing more problems?
[00:25:10] Daniel: Yes. There you go, simple answer. Yes, it will do. Especially in the short term as we are not used to it and we think that GA4 is an analyst tool where it’s not, right? So I think that’s, that’s the thing. But I don’t think this is a new problem, Bhav. So first of all, you know what’s going to happen, the marketer’s just going to pick the number that has lower CPA, right, and just use that to their advantage. So I mean, this is not a new problem because we could take that exact example with attribution modelling. Like which attribution model do I take? I could build an attribution model that tells the story I want to tell. I can pick another attribution model that attributes everything to me.
[00:25:40] Daniel: I can go into Google Ads and pick that attribution model, and that’s making my PPC look really good. I can go into Google Analytics for social and make that look slightly better. So I don’t think having one number here is going to be the answer. I think what it’s moving towards is as an organisation being aware that there’s variation, there’s infinite kind of possibilities and just agreeing as an organisation, we use this number here to report internally. Now, that’s not going to be useful for things like bid optimization or budget adjustments. Let’s say you need to shave 20% off your investment across all channels, that one number’s not going to tell you where to do that, but I think what’s going to be really useful here is just another piece of data, another piece of the puzzle to build that story or narrative.
[00:26:15] Daniel: And I think because data is, you can’t tell what’s trackable, what’s accurate, what’s modelled, what’s estimated in GA and stuff like that anymore. But because the idea of having a 100% accurate, trustworthy, provable data set is kind of long gone. I think all we’re doing now is building narratives. We’re building stories with data and having variation and various different sort of perspectives is going to be key to that. So it’s saying, okay, well we’ve got about a £5 CPA, and it depends on how we look at that. It varies between £4.50 and £5.50, but it’s about £5, so it’s all good. There’s no one answer and I can’t be confident that it’s exactly right either. So I think moving beyond that, it’s just, okay, for my bid optimization, we’re using this methodology, but as a narrative I’m putting together as the channel owner, I can start telling that story and the numbers will be less important than the story, the narrative, the text I suppose that we put behind that.
[00:27:00] Bhav: That’s a very good response to my question. So I’m bought into your view.
[00:27:05] Daniel: That’s it, maybe I should do this for a living.
[00:27:09] Dara: Do you think, here’s another question. So do you think that when Google bought Firebase, do you think they knew, do you think it was a plan or was it a happy coincidence that they were able to kind of use it to become the new version of GA.
[00:27:23] Daniel: I’m speaking again just from having learned experience from the platform and being in the industry for a while. I don’t have any inside scoop, or I’m not on the inner circle, but I think they wanted to enter the app world. They saw the rise of app and obviously app marketing, you know, key part of this is from a marketing perspective. They needed to get their foot in the door. They had AdSense which is their ad delivery mechanism for apps, but they needed to have some kind of ecosystem for app infrastructure to be able to kind of do more right?
[00:27:49] Daniel: This was after things like ePrivacy, but before GDPR came in, so this was around, I’m not even going to fact check myself here, but let us know if I’m wrong, but I think it’s around 2016 they bought Firebase, and GDPR came in in 2018, so they couldn’t have seen that early. They might have known that things were going on, but they couldn’t have saw that as a reaction. And the thing is, they didn’t buy it with Firebase Analytics wasn’t a thing when they bought Firebase, believe it or not, Firebase Analytics was built by Google. It was actually not an acquired product, they built it themselves. So I think they bought an ecosystem. They built an infrastructure, and they also bought a brand’s name, which is key. They bought a brand’s name like YouTube, buying YouTube, they bought a brand, you know, they had Google Video, which no one fucking used and so they bought a brand, and I think that’s more of the thing there.
[00:28:29] Daniel: I mean, I’m sure they had like analysts predicting the market trends and stuff, but I can’t imagine they kind of foresaw GA4 being based on the Firebase stuff as they were building it. But maybe they thought of it, who knows.
[00:28:40] Dara: Because I’ve got a photographic memory and I’m so smart, and I definitely didn’t Google this, it was 2014.
[00:28:46] Daniel: Ah, right. I can see it in your glasses, Dara.
[00:28:50] Dara: You saw the reflection of the Google logo in my glasses.
[00:28:55] Daniel: That’s the burden of three people that wear glasses. We can instantly tell when we are going between different tabs.
[00:29:02] Bhav: I mean, I’ve definitely got a screen open. I’m trying to see what people are saying in real time, just so that I don’t feel like I’m completely wrong but I think the general census is going to be the people who love Google and know it well enough are going to be yay, GA4. It’s really going to be like the passive users who are mostly going to potentially struggle with the move to GA4.
[00:29:21] Daniel: But it wasn’t like a midnight strike and everything broke, like, we are still getting data coming through our UA property from the Measurelab website. And I think this is the other thing is that I don’t think it’s really had an impact just yet. I mean, it’s been like an oddly quiet but also a bit of an anti-climax actually, because the banner has now just changed to say, oh yeah, your property will be downgraded soon. Imagine like the millions of properties out there existing, they’re all joining some queue. I almost wish that it’s like a, like a gig ticket site or like Glastonbury or something like that, where they give you are number X in the queue or something like that. I think that could be quite interesting. We’re still in the queue and we haven’t been downgraded just yet or turned off, so who knows what actually that means. I would imagine that well, no one knew that until the 1st of July, but I haven’t really seen any kind of violent reactions or you know, throwing the toys out the pram just yet, because I think it’s just, like you said at the beginning Dara, it’s just like, get over it, move on. Change is here, like it or not, it’s not your choice. It’s happening whether you like it or not, and you just kind of move on with your life.
[00:30:17] Dara: I’d love to be a fly on the wall as there’s going to be people who have somehow managed to avoid this news completely. And they just won’t know that it’s happened, and then they’re going to log in one day and it’s going to be a completely different product that’s been auto migrated to GA4.
[00:30:34] Daniel: So if you know, if you listeners know anyone that’s like that, then that’ll be interesting. Let us know, it’ll be interesting to see their reaction. Send us a link to a LinkedIn post.
[00:30:42] Bhav: I don’t actually think it’s going to be any sort of like practitioners. The people who are going to be most surprised with it are going to be the ones who sporadically see the data and they’re going to see like, let’s say for argument’s sake, like a monthly board review meeting or something like that where, you know, where teams have historically reported on users, page views, you know, whatever, like nonsense metrics go into like board decks and they’re going to see it for the first time ever. They’ll be like, hey, why have these numbers changed? And like, and also part of me thinks they’re probably not even going to notice that the numbers have changed. There’s going to be a certain level and above where these numbers will flip overnight and they’re not even going to realise that something’s changed.
[00:31:20] Bhav: So I think the people that probably care about it the most are going to be the ones who use it. I think everyone else will be like, okay, it’s changed let’s just get on with it. Like what’s the new numbers? Tell me.
[00:31:28] Daniel: One of the biggest revelations and kind of changes I went through in my career working with sort of web analytics and Google Analytics specifically, is the reality of how unimportant it all is, and just acceptance of that. Because when you are new into this, if you’re in the weeds of everything and you’re working with lots of people around it, it can feel very important. It can feel the most important thing within a business running successfully is like, we need data. And I saw lots of arguments when GDPR came in saying, the Google Analytics cookie is essential because we need it to run our business and the reality is you don’t. And I’ve worked over the many years, I’ve had tracking issues where tracking stopped, people didn’t notice. We’ve had someone, their site being attacked by a bot scraper that inflated their traffic four times for two months, and we didn’t notice either or they didn’t notice, rather.
[00:32:07] Daniel: And I think the key thing here is, is that, and going back to this idea of when you put together these monthly marketing reports or these monthly dashboards for businesses and organisations, I think the thing that I always go back to is this idea of what is a KPI? What is a key performance indicator? And I don’t think web traffic is a KPI because what we’re not going to get, and that going back to this idea of a KPI is something you can get promoted or fired over. If web traffic was a KPI and it flatlined, you’d get sacked. Or if it quadrupled, you would get a promotion because it, you know, and that’s not always the case. And I don’t think ever I’ve come across someone that would’ve got fired over web traffic dropping to zero. And I think this is the point where it doesn’t become a KPI.
[00:32:43] Daniel: And I think there’s bigger things, more important things, numbers that exist within an organisation that may be based on web analytics data, but thinking of page views and sessions as KPIs and I think the reality there, it’s not really that important. And you’re quite right, I don’t think we’re really going to see those kind of organisations, those kind of people are probably not even going to notice that the numbers changing because it actually doesn’t matter if the definition has slightly changed of how sessions are defined or what a user is and you know, and also we are not going to find out until the 1st of August, at very least right? And I think that’s the next thing is they won’t find out. So I wonder how many people are going to already be using GA4 data without even realising it’s going to be interesting.
[00:33:18] Bhav: You know, I’ve been in situations where we’ve had some type of metrics definition change or something’s changed, which has caused the underlying data to change. And actually in most instances, most people are okay with it if you’re transparent, if you can explain why a number has changed, and especially if it’s something as like big in magnitude as a change from Universal Analytics to GA4 where underlying models have changed. I think the worry will be when people do things like year over year comparisons. I think that’s when it’s kind of like, okay, well how do we do this? And in most instances, you probably have enough data to be able to say, okay, for month one, the year over year won’t make that much of a difference. But once we’ve got this for like 3, 4, 5, 6 months and we’re seeing a consistent %X difference year over year, you can just apply that retrospectively, right?
[00:34:02] Bhav: And you can just say, okay, well if the number is over, you know, it’s 10% lower than it was say last year, and it’s been consistently 10% lower than last year, we can kind of just say, okay, well compared to last year, we’ll say 10% is as a company, we’ll accept that 10% differential between last year’s data or the UA data and the GA4 data as gospel. And if it starts to waver into like 25, you know, 25, 30% difference, then maybe there’s calls to react and act. It’s important to us, most people in the organisation aren’t going to care, especially if you can justify the change.
[00:34:34] Daniel: Well, I think it’s just a pride thing, right? If you are the one responsible for this tool, or you’ve got governance over it, or you’ve done the implementation, it’s a pride thing. I think it just hurts the pride a little bit when that breaks or stops and you don’t notice. But I think in terms of business impact, who knows. I just wanted to kind of wind it down a little bit and just share a couple of links that will be in the show notes and I’ll open up to you guys in a sec as well. But there’s a couple of blogs that I’m going to share from Brian Clifton and Simo Ahava talking about this whole change and their perspective of this kind of like a, an equivalent of what we’ve been doing in terms of talking about it here. There’s some really interesting takes on that, some of them are very Brian and very Simo in terms of their approach to things, and I’ll let you read through that to see what they have to say.
[00:35:09] Daniel: And also just because I just mentioned it there, and I just remembered it existed, the blog I wrote around KPIs and this idea, I think I called them this idea of ‘key KPIs’ really irked me when I heard that being used for the first time and you having a key, key performance indicator, and it’s like if everything becomes a high priority then nothing does, and you have to promote it one higher. And so I’ve got a blog just talking about that, it’s a bit ranty from me, but you know, I’m prone to that occasionally and I’m going to stick a link to that in the description. Is there anything, last call, is there anything you would like to share before we wrap up?
[00:35:37] Bhav: Yeah, does this mean we’re never going to, you know, Dan, you’re never going to talk about GA4 again in this kind of like topic ,vein of transitioning and migrations. Or will GA4 now be the norm of what we talk about? Are we going to stop talking about UA from now on?
[00:35:51] Daniel: I hope we stopped talking about it moving forward. I can’t promise that the drinking game will be back in the next episode I’m sure, but yeah. So we’re back to business as usual I’m sure. You know what? Just on that, I do kind of hope they dropped the ‘4’ term. It’s always confused me, like, why the hell have they started numbering them at four and where’s the other three? Because there’s been way more than three versions before this. I can hazard a guess why this is number four, but fucking hell, why not just call it something different? I hope they drop the numbering system, who knows? Or they’re going to lean into it and they’re going to start changing the number every year, keeping us busy right?
[00:36:20] Dara: In one of those articles you mentioned, the one from Brian Clifton, he mentioned why it was to do with the phases of Google Analytics, and Universal was launched in phase three. So that’s his, I don’t know if that’s fact or theory, but that could be why.
[00:36:33] Daniel: I mean, it sounds good enough, right?
[00:36:34] Dara: Yeah, now looking back, you can call it UA3 or GA3.
[00:36:38] Daniel: Oh God, UA3.
[00:36:39] Dara: I’m confusing myself now, but 4 is not consistent in the marketing and the documentation, like in the platform, in the interface just as analytics, doesn’t it. But then in the documentation, because they still need to have the Universal stuff as well, I guess it’s just to differentiate. But yeah, I look forward to the day where it’s just called Google Analytics again.
Rapid fire questions
[00:36:58] Daniel: Well one last thing is that I did run a small meetup, so we used to run this meetup called Brighton Analytics, and this was, oh, over four years ago. And so this was one good thing to come out of the sunset date is that I used this as a good excuse as any to relaunch BrightonAnalytics and arrange a little kind of social get together in Brighton on the Friday, the 30th of June. So now that I’ve done that, I’ve got the taste for it, and I’m going to be doing a couple more of those over the next couple of months and seeing how that goes, it’s purely a social event. Unlike your very well organised and highly popular CRAP Talks, Bhav, this is purely an excuse to get together and have a pint. I mean, there’s no talks, there’s no nothing else going on there.
[00:37:36] Bhav: Whatever you think right now, like four years from now, someone’s going to be telling you, hey, I’ve just started my own sort of like meetup. It’s just a few drinks, it’s not like you’re well organised, BrightonAnalytics, where like speakers and things like that. CRAP Talks started in the exact same way. There was about 12 of us in a room, we were just discussing analytics. We had a presentation, but it was mostly about just drinking and having a chat.
[00:37:57] Daniel: I’ll tell you what, maybe there’ll be a meet in the middle in a couple of months if it does ever get there, it’ll be like CRAP Talks Brighton or something like that, maybe that will be the case.
[00:38:05] Bhav: Oh, we should do that, we should do that.
[00:38:07] Daniel: Cool, well nice chatting with you. It’s good to just kind of react to this kind of stuff and there’s lots of opinions flowing around and just wanted to kind share our thoughts on the matter.
[00:38:14] Dara: Okay, no guarantees we’re going to stop talking about Universal Analytics and hopefully we’ll at least reduce it. This has been a really good chat. UA is dead, long live GA4.
[00:38:25] Dara: That’s it for this week, to hear more from me and Dan on GA4 and other analytics related topics, all our previous episodes are available in our archive at measurelab.co.uk/podcast, or you can simply use whatever app you’re using right now to listen to this, to go back and listen to previous episodes.
[00:38:44] Daniel: And if you want to suggest a topic for something me and Dara should be talking about, or if you want to suggest a guest who we should be talking to there’s a Google Form in the show notes that you can fill out and leave us a note. Or alternatively, you can just email us at firstname.lastname@example.org to get in touch with us both directly.
[00:39:01] Dara: Our theme is from Confidential, you can find a link to their music in the show notes. So on behalf of Dan and I, thanks for listening, see you next time.