#41 Littledata’s GA4 journey (with Edward Upton)
This week Dan and Dara are joined by Edward Upton to talk about how Littledata are approaching GA4 with all of their ecommerce clients across Shopify and BigCommerce. They discuss the GA4 measurement protocol and the nuances that come with moving from the UA version, as well as some of the exciting features that GA4 is bringing to the table for the ecommerce space.
See Littledata’s website for more on who they are and what they do with GA4 on Shopify and BigCommerce on https://bit.ly/3NWuAY6.
Find Ed over on LinkedIn at https://bit.ly/3mX1FaA.
In other news, Dan witnesses a lot of love, Dara sees some sailors and Ed gets a green thumb!
Follow Measurelab on LinkedIn for all the latest podcast episodes, analytics resources and industry news at https://bit.ly/3Ka513y.
Intro music composed by the amazing Confidential (Spotify https://spoti.fi/3JnEdg6).
If you’re liking the show, please show some support and leave a rating on Spotify.
If you have some feedback or a suggestion for Dan and Dara, fill in the form https://bit.ly/3MNtPzl to let them know. Alternatively, you can email podcast@measurelab.co.uk to drop them a message.
Transcript
[00:00:00] Dara: Hello, and welcome to The Measure Pod, episode number 41. Dan and I are very glad to be back after a couple of weeks off we’re feeling recharged and ready to go. As a reminder to anyone who’s listened before, or as a little intro for anyone new, The Measure Pod is an analytics podcast. So we talk about topics, opinions, news, all related to the analytics world, especially Google Analytics. I’m Dara, I’m MD at Measurelab.
[00:00:40] Daniel: And I’m Dan, I’m an analytics consultant and trainer here at Measurelab. Before we jump into chatting with our nice guest this week. So Google Analytics 4 have just released a pretty big feature actually, and that’s their consent mode, behaviour modelling. So this is something they’ve been teasing for quite some time, but a feature called consent mode that’s available in GTM and Gtag basically enables you to collect cookieless data when users haven’t consented for analytics tracking. So what this is enabling you to do is to kind of fill in those gaps, or at least enabling Google Analytics to fill in those gaps of all the data you can’t collect because no one has consented for cookies.
[00:01:12] Daniel: It uses clever machine modelling stuff to kind of fill in those gaps. So it’s been something they’ve been promising full at while it’s finally rolled out and yeah spoiler alert but we’re going to be talking about this on our next episode as well. So I won’t say any more about that big deal in terms of rolling out a feature this big, but so much so that, Dara, me and you’ll go through it in a bit more detail next week.
[00:01:30] Dara: Yep, definitely watch the space. Okay so back to today though, we are very excited to be joined today by Edward Upton, founder and CEO of Littledata. So first and foremost, thanks for joining us in The Measure Pod Ed and welcome.
[00:01:44] Edward: Well thanks for having me. Excited to talk about GA (Google Analytics) and all things analytics.
[00:01:48] Dara: Yes so are we, but just before we dive in, why don’t you tell us a little bit about your analytics journey?
[00:01:54] Edward: Sure so, I think I started using Google Analytics actually probably pretty early on maybe 2008. My first start-up, when I was, I was running an edtech business, helping teachers share curriculum resources. And I think I just got frustrated with not being able to get the insights from the database and wanting more detail. So I’d come from a fairly analytical background and I decided to sort of again be my own analyst. Kind of fast forward to 2015 and I found myself working as a sort of freelance analytics consultant.
[00:02:19] Edward: So I was doing so GTM setup and just general GA (Google Analytics) audit for some bigger brands, and I think that was the spark of founding Littledata that we realised a lot of the stuff we were doing as consultants as many of you will be in the audience. We could productise and basically build into a sort of self-serve platform for scaling direct to consumer brands specifically on Shopify.
[00:02:39] Edward: So yeah, I’ve been doing this stuff quite a long time as an analyst and then moved into running a software company.
[00:02:44] Dara: That’s really interesting and that actually leads us nicely onto our topic because we are obviously going to be coming at this very much from the GA4 perspective. It’s a hot topic for us as well as I’m sure it is for you at the moment, but we are really keen to get your perspective as a vendor in this space, in terms of how the news of GA4, or should I say the news of Universal Analytics being deprecated and the kind of fast tracking of Google Analytics 4. So we’re very keen to get your perspective and thoughts on this big, big change in the industry. Just to kick things off, I guess, can you give us a little more and give our listeners a little bit of a background on Littledata? So what kind of product are you offering? What problems are you trying to solve for people?
[00:03:22] Edward: Yeah, sure. So Littledata works with growing eCommerce brands that want to own their own web channel. And therefore they tend to have a high value on, measuring the customer engagement, the channel. And so when I said web channel, that’s typically their own eCommerce store run on Shopify, or we also support BigCommerce. And what we do for those brands is basically provide a plug and play way that they can set up all of the tracking, both a sort data layer driven browser tracking, or clients side tracking and also server-side tracking from the cart, checkout, order, and post purchase events. So effectively we are a connector between their eCommerce store and their customer experience more generally and their analytics platform, typically Google Analytics.
[00:04:03] Edward: Yeah, really we just make it a lot simpler for them, ridiculously easy in our own language to connect all the data points and get started in a way that, on other platforms they’d need to use or hire a sort of consultant and engineer to get that all set up.
[00:04:15] Daniel: And is there something, am I right in thinking it’s been a little while since I’ve dipped into Shopify, sorry Ed. The checkout funnel itself is kind of off limits, even if you could deploy something like your own GTM container or GA (Google Analytics) tag natively on the header of every page of your site, there still is kind of black box, which is the checkout funnel unless you go to the super high premium tier when you can customise that right. Is that the kind of thing where Littledata comes in and helps you kind of tie all that stuff together?
[00:04:37] Edward: Yeah, exactly that. As you say on Shopify plus brands it is possible to add some custom HTML, but it’s actually highly not recommended and probably likely to be blocked by Shopify in the near future. And so what Littledata is able to do is basically use the server-side events to build the same picture so we don’t need to add any tracking to the checkout and therefore you can track check out steps even for stores not on Shopify plus.
[00:04:59] Daniel: Amazing, it sounds so simple, right? It’s a deployment of GA (Google Analytics) across the site, but then you’ve already thrown in things like server-side tagging, client-side tagging, the kind of continuity and the issues we’ve got with that. Yeah, it’s really interesting, I’m really keen to dig into that and to see like how that works, but I mean something I’ve been curious about this whole time is the whole, obviously it’s on the front of our mind and probably a lot of our listeners mind is that GA4, as Dara said, it’s kind of part of every conversation whether you like you or not right now.
[00:05:23] Daniel: So, how have you found this transition to the Google Analytics 4 perspective? Has the kind of demand driven the development, or have you found that you are getting in front of the curve with your GA4 connector? How has that kind of reaction been around GA4 with these kind of Shopify stores?
[00:05:37] Edward: Yeah, I think we’ve probably like bit ahead of the curve. I mean, as analytics geeks we are probably got excited about GA4 before most of the brands. We actually, you know, obviously from like 18 months ago when it sort of really came out, we realised it was still lacking a lot of the eCommerce functionality, which we can carry on another question, but it still is lacking that, but what we could see is it’s obviously the future, so we need to integrate with that in the same way that we integrate with Universal Analytics. It was actually a problem on Google’s end with, we can talk a bit more about measurement protocol, but ironically, even though they said GA4 is ready, measurement protocol for GA4 was not ready and was in alpha at the same time.
[00:06:09] Edward: So until they fix the problem with the actual traffic attributional stitching the measurement protocol events with the previous data from the browser, we couldn’t really launch our solution. After we launched it a couple of months ago, we’re starting to see uptake of that integration. I would still say though that a lot of brands are still in the kind of what is GA4 and why do we need it phase? So we are just trying to push the same message I’m sure you guys are that, you know, ultimately you need to get on it in the next month or two, because otherwise you’re going to be lacking the historical data in GA4 when you come to 2023.
[00:06:38] Daniel: Yeah Google has decided and cast their verdict over the Universal Analytics and it’s, you know, it comes to us. I think people like us the vendors or the kind of the consultants around it in a sense we’re trying to kind of sell them on the idea of GA4, but actually not trying to sell anything. I mean, if only we got commission, right. But the idea is just that we, you know, have no choice, Google have made a decision. This is just going to help you out in the long run. A sentence that I keep coming back to I say to a lot of my clients actually is “your future self will thank you for this”. Because you know, by the time you get to next year, you would’ve wished you have had this data in place now, even though it might not feel like it’s super relevant or urgent right now.
[00:07:12] Daniel: So a lot of the time is like, we know that now is the time to do this, even though the date on the piece of paper that Google have released about deprecating Universal Analytics is next year.
[00:07:20] Edward: The three reasons I’m excited about GA4 for our eCommerce customers. And one reason I’m a bit concerned. First of all, there is no Shopify out box connector for GA4. So even the brands that are kind of happy with the almost good enough data for Universal Analytics. Just Littledata can offer, a very simple you know, just get it set up in a minute experience. The second thing is because we’re basically using the same data pipeline we’re using for Universal Analytics, we can kind of be sure that you’re tracking like for like. Because obviously there’s a lot of differences anyway, with GA4 and I think data inconsistency, you don’t need to throw on the pile to complicate the trust in the new platform. And yeah, the third thing is I think we can talk more about, but the opportunities to then use that data in GA4 to feed into other destinations, particularly BigQuery.
[00:07:58] Edward: The challenge though is really the fact that as I said, they haven’t built out a lot of the eCommerce reporting and in fact, even collection functionality from Universal Analytics. So they’re still missing product scope Custom Dimensions. They’re still missing any kind of out the box, sort of, you know, sales performance, product list performance type reports. So I think it’s very hard for the brands that we work with who’ve come to rely on some of those Google Analytics or Universal Analytics reports to their standard daily checks that they’re not there. They have to build them from scratch, or they maybe have to plug in another reporting tool or Data Studio or something on top of that.
[00:08:32] Daniel: So aside from the kind of lack of reporting infrastructure, which I’m fully behind, the kind of e-com and the enhanced e-com reporting in Universal Analytics was way more fully fledged than the GA4 equivalent of like one or two reports where you can switch something out for now, you know, who knows if they will change that down the line. Is there anything technical that they’re missing from the Enhanced Ecommerce model they had in Universal Analytics to the GA4 model. Is there anything that’s fundamentally lacking in that, from your perspective that you’re just going to have to figure something out or just move on from having is there anything more technical not just the reporting side that’s missing.
[00:09:02] Edward: Yeah, I think on the data collection side, the only one I think of is really the, the product scope Custom Dimensions which to be fair, are a bit of an edge case, like the, you know, most of the dimensions you want to send about products are out the box. So things like variants, you know, quantity, size, etc. But I think in a way they’ve added, you know, the, the great thing about GA4 is the ability to send way more kind of event properties. I love their, the way that it made it easier to just track a lot of stuff out of the box and then use what you want in the reports later. And that really fits our sort of philosophy when it comes to tracking, which is let’s maximally track all of these different properties that we know and then let’s figure out later what we want to report on because you know, that the problem challenge with reporting is always like, I wanted to cut the report by that dimension, but it’s not there yet, we can’t populate that historically.
[00:09:44] Dara: Are you suggesting to your customers that they continue running. Have they basically got two connectors running in parallel now? So the old Universal Analytics.
[00:09:51] Edward: So we’re basically saying, look track in parallel, duel tag for, for the next year. Because you’re doing it service side there’s not really much overload on doing that. Obviously, you know, if you’re doing all from the browser, you’d have to ping twice to Google. I know since we are doing that direct from our servers really know sort of big overhead, that means you have the flexibility to use Universal Analytics or have people in your organisation use those Universal Analytics reports up until next year. And it means you’ve got a sort of a year to get familiar with the new reports and frankly, probably wait on a bit of Google’s product development to make it a bit easier to build those eCommerce reports.
[00:10:22] Daniel: So you mentioned there Ed the server-side deployment of this, all running through your servers to kind of do that. So let’s dig into that little bit, because I know that the, the measurement protocol for GA4 is somewhat lacking in some aspects, but it’s kind of, from what I understand, it’s fundamentally a different approach than Universal Analytics had whereas you almost could replicate the entire schema on the whole data collection with Universal Analytics through measurement protocol. Whereas in GA4 it’s not quite designed to be a one-to-one replacement. It’s almost like an embellishment, which kind of opens up some gaps in some of the things you can do with measurement protocol. So is there anything that you are kind of hitting up any edges you are pushing up against or any things that you are having to adapt to with this new approach to measurement protocol that GA4 have got.
[00:11:00] Edward: Yeah, I think the problem for a long time was about sessionisation so I think Dan, you and I talked about this a few weeks ago. In Universal Analytics you know, basically if an event came through in the same 30 minute window or whatever, it’s configured as your, as your session duration. And you sent the same Client ID, same cookie ID with that measured protocol event. Google would say, aha, this is all one session. So, you know, when you looked at the reports, you could see the whole customer funnel, the customer journey. The client and server-side events were basically linked into one customer journey.
[00:11:30] Edward: The challenge we had for a long time, the reason we were sort of delayed in launching it was that GA4, you actually have to specify a session. You have to say this event is linked with that same session that had happened with the client and that’s not to do with a time period it’s to do with a Session ID.
[00:11:44] Daniel: That was a good talk by the way, so that was at MeasureCamp London a couple of weeks ago, or at least at the time of recording, and I managed to jump in and watch that one, it was a good chat. But yeah I didn’t realise that the whole Session ID is now like almost a requirement with GA4’s measurement protocol. I think it’s at least in my head it’s because Universal Analytics, it did all the sessionisation backend. It kind of did all the calculation of sessions all in their server. And so you can kind of just collect hit after hit with a User ID or a Client ID, and it just processed it there, whereas weirdly, and I’m sure there’s some logic of why this is better, but for me it feels like a step backwards. Like in classic analytics, Google Analytics, where they process and calculate all the sessionisation on the browser. And then, so you have to provide this Session ID and it just feels, it feels odd. It feels like a backward step in my brain, but like, there must be some justification of why this is the step forward, I don’t know.
[00:12:29] Edward: Yeah I think you’re right. I think the idea of GA4 is this last sort of aggregation and sort of sessionisation done at collection time and more of it has left us the reporting time. It actually reminds me the other thing we’re still grappling with is that User ID. So as you know, in Universal Analytics Google never really incorporated the idea of User ID into reporting. They had these sort of separate reporting views it was a bit unsatisfactory.
[00:12:50] Edward: So as I understand it at the moment, although I do know it’s on Google’s roadmap this quarter, you can’t send it. If you send the User ID it kind of breaks the other traffic acquisition reports.
[00:13:00] Daniel: Wait in the measurement protocol or in, just in general?
[00:13:02] Edward: In the measurement protocol specifically yeah. So that’s still a kind of, it’s still a limitation measurement protocol at the moment. If you do send User ID it kind of mucks with the normal session attribution.
[00:13:11] Daniel: Wow. And they’ve just removed the beta flag, right? It’s now in production.
[00:13:15] Edward: They may have fixed that before, I haven’t checked in the last couple weeks, but certainly as of a couple of weeks ago it was broken. So I think that’s been the frustration.
[00:13:22] Daniel: Yeah it’s just classic Google-ism isn’t it? When they say beta they mean alpha, when it’s production, they mean beta. And then about a year or two after it becomes full feature ready.
[00:13:30] Dara: One of the other benefits Ed to end users to businesses is that they can now get access to the data in BigQuery for free. Whereas obviously with Universal Analytics you had to have 360 to get that export. Are you seeing advantages for Littledata in that respect with the likelihood that BigQuery usage is going to increase and it’s going to be more accessible to more businesses?
[00:13:51] Edward: Yeah, definitely. I see two actual advantages. So the first is that I feel like a lot brands want to have a data warehouse. The kind of brands we work with are a mid-size eCommerce brands. So they’re typically doing between maybe a million and a hundred million online sales a year. They don’t have huge engineering resources in house, they probably outsource a lot of their development to an agency. They kind of want a data warehouse almost to future proof it, they don’t have an immediate use case for it, but they kind of feel like if they start collecting the data, at least at the point at which maybe they have enough resources to hire in an analyst or something or hire the like of you guys, they will have the data there.
[00:14:24] Edward: So I think for me BigQuery provides a really low cost way. In fact, a free in lots of senses, so they can just start collecting that data permanently. And then the second opportunity for us which is super exciting is that there are definitely brands who are choosing to use other BI tools and some more eCommerce specific reporting tools rather than Google Analytics itself as a report. And so this provides us with a way we can actually integrate and feed into these other reporting tools, because they’re all very used to building reports from SQL queries. So whether they use Snowflake underlying or BigQuery or whatever, they can easily adapt the format of the data.
[00:14:59] Edward: So I think both the kind of brand owning their own data warehouse much more cheaply than additional solutions, and also integrating with other eCommerce specific reporting tools.
[00:15:08] Daniel: And probably the fact that you can join it to other data sets and then activate it in something that you know, marketing tools they believe or not don’t have Google written on them. I think outside of the walled garden of Google is the only real way to get any kind of activation from it, especially these kind of brands you’re talking to. So, you know, digitally driven a lot of kind of activation through things like CRM and email, which again will never be supported through Google Analytics, those kind of things really.
[00:15:30] Edward: Yeah, and I think that’s my only concern actually is can we get enough identifiers into BigQuery that makes it easy to stitch in? So absolutely, email marketing, massive channel for a lot of these brands, what they get from email marketing tools like Klaviyo is they have all the sort of, basically all the campaigns attributed to an email address. Obviously we can’t send an email address to Google Analytics, it’s a breach of their PII conditions. So, you know, the only way we could then stitch an email behaviour is by hash email. Is that allowed, Google’s always been a little bit, let’s say circumspect when it comes to pseudo anonymous identifiers.
[00:16:03] Edward: So I think that’s the limitation, if you go via GA4 measurement protocol into BigQuery obviously you are limited by Google’s terms and conditions of what you can send to GA (Google Analytics) and does that give us enough richness to then stitch on the other data set?
[00:16:15] Daniel: That’s a really good point. It’s hard, right? Especially those users stitching, because you end up having this kind of like, I mean, Google have got their own kind of like device graph and their user identification, but you want to build your own on top of it. And it’s like, I’ve got my Microsoft Click IDs, I’ve got my Google Click IDs, I’ve got my Mailchimp IDs. You’ve got all these IDs and you’re stitching those together to try to do something with it but especially as you said as well Ed, like types of clients and we see the same with these kind of clients, it’s like the resource just isn’t there. They don’t have data scientists and data engineers. So it’s almost futile saying, you can do this because they won’t, or at least they won’t have the time or the head space to be able to kind of crack on and learn how to do this kind of stuff.
[00:16:48] Daniel: So interesting to see what kind of pops up, you know, especially as a Universal Analytics disappears and everything is GA4, a lot more people are going to be using BigQuery even if they’ve just been told to use it and they don’t really know why just yet. I think there’s going to be maybe like a boom of a plug and play like reverse ETL tools in a way where that you can just kind of, I’m sure there is already, but more so where you’re going to find the kind of marketer friendly version of connecting your BigQuery data to Mailchimp and some other system. I think that’s really where that kind of value might be.
[00:17:14] Edward: I think it’s our view of the market that these kind of reverse ETL tools, so that tools where you can basically from the data warehouse, then send the, yeah the attribution back to marketing platforms or figures back to marketing platforms. That is a common theme in the market and as I said, I think my only hesitation right now is whether the GA (Google Analytics) to BigQuery data is sufficient for that.
[00:17:33] Edward: So some of our bigger customers are using our Connect Segment. So Segment is in itself our data platform has hundreds of different data connectors, particularly to lots of data warehouse and CRM type tools. And so yeah, they’re already, for example, sending it to a Snowflake instance or something to do this. And then adding a reverse ETL tool onto that. So I know brands already using it, I’d like to be able to support that kind of use case with BigQuery. I think we’re still early days, whether that’s going to be possible technically.
[00:17:59] Dara: Ed are you noticing any of your customers actually consider moving away from GA (Google Analytics) now? Because one of the differences, obviously this time around is it’s not like before where Google moved from classic to Universal or from Urchin to classic. In this case, you’ve got to effectively start from scratch, even though obviously you can get your data out of Universal and into BigQuery, but this might prompt more customers to think about whether now’s the time to move away from GA (Google Analytics), have you noticed any sign of that?
[00:18:25] Edward: Yeah, definitely. I would say that’s that sort of behaviour we’re starting to see, as you say, when brands come to have a major re-evaluation of the tools they use, it’s obviously an opportunity to say yeah, but do we use this tool? And I think, Universal Analytics while it’s hugely, widely used, it’s not always hugely loved. I think there were lots of cases where people found it hard to use. So I do feel with some brands who are not fully bought into Google’s family of products, they may take the opportunity to change. You know, frankly as I said, there’s nothing that’s not going to need a lot more cost and implementation time. So I suspect for most of them they’ll come to the conclusion that GA4 actually does present the cheapest most robust way to do this kind of measurement. That said I would expect some of the bigger brands as I said to look at this whole like data warehouse and then specialist sort of BI tool on top of that as a viable option.
[00:19:13] Edward: But as I said, our data platform strategy is to try and support them with that as well. You know, GA (Google Analytics) is our most popular data destination, but it’s not our only one. And I can certainly see us partnering with a number of those other sort of BI tools in the future.
[00:19:25] Daniel: It’s really interesting that around that whole GA (Google Analytics) or no GA (Google Analytics) kind of bypassing it or going a completely different route because especially with things like server-side tagging, even if it’s server-side GTM, you know, this is a concept where you can take the data directly and put it into a warehouse. You don’t need to go through an aggregator, like even Google Analytics. So even sticking within the Google product suite, you know, you use server-side GTM, you use client-side GTM maybe. You might even use GA4 as the feeder into the client, into the server-side. But the thing is you can just take that data directly and put that into a warehouse and then, like you said, layer on reverse ETL tools or dashboards.
[00:19:59] Daniel: I think for me, it’s a really interesting shift with GA4, which is that for me that Universal Analytics was almost like a data tool first and foremost, and it became like a marketing tool secondary. It almost had these modular pieces you can bolt on to become more marketing friendly, whereas GA4 is a marketer’s activation tool first and foremost. And it even doesn’t focus too much on the data, it has a lot of issues with cardinality and things like Custom Dimensions. You can only set so many, but all of them get fed through to BigQuery.
[00:20:26] Daniel: It is almost like a subset of the data you’re seeing within the GA4 UI. I’m thinking that if you just want it for a data collection, there’s actually a big question now of like, do I need to put it through this kind of marketing engine? Because there’s lots of question marks around Google and what Google does with the data and the kind of the US nature of it and where it’s stored. I’m sure we’ve all seen all those things flying up over the last couple of months. But the point is, do you need GA4? No, but you might do if you advertise through the Google ecosystem because you have to get into that wall garden some way, and actually it’s a really good way of doing that.
[00:20:53] Edward: I think that that kind of, it’s really interesting to see it as a marketing activation platform first. I see that, Google have built this out with the very first things they built with the connectors for Google Ads and I see that’s kind of their thinking. Because if you think about strategically, Google Analytics never really made sense for Google, you know, why are they offering this? So I think them trying to make it much more of a tighter overlap with their valuable Google Ads audience and that’s why if you are spending on Google Ads, and obviously the majority of the brands we work with are spending heavily on Google Ads, you’d be crazy not to send data to Google Analytics because it’s the very tightest and most automated way to build audiences and other things in Google Ads or, or DoubleClick.
[00:21:28] Edward: I think brands should use Google Analytics 4 just for that one reason. Whether or not you say they want to use it as their core measurement tool is a slightly different question. And I guess, again, that’s really what I see the data in the business is doing is we want to support brands, whatever their main customer acquisition channels are. So as you say, if it’s distributing email campaigns or understanding triggering email campaigns, then let’s do that. If it’s helping Google Ads let’s do that. If it’s Facebook Ads, you know, we just launched the connector for Facebook Conversions API (CAPI). And the interesting thing about that is Google Ads does not have the same server-side endpoints as Google Analytics or Facebook Ads.
[00:22:00] Edward: So I’m not sure are Google going to add server-side Google Ads connectors, or are they just going to basically say use GA4?
[00:22:08] Daniel: That’s the, that’s the million-dollar question, right? But yeah my theory with this, for the, you know, for the longest time has been that ever since the announcement came for Universal deprecation, is that they’re going to deprecate the Google Ads pixels and the floodlights as well, because why wouldn’t they? They’re deprecating non-compliant technology that quote unquote they’ve solved with GA4. That is also the same thing for the Google Ads pixels and the floodlight tags too, within the DoubleClick world. So for me, it’s like almost inevitability that they’re going to do the same thing with Universal Analytics that they they’re going to roll out and turn off flood lights and Google Ads, and then all of a sudden Google Analytics 4 is the centre point and the only measurement solution for the Google marketing world. But look that’s my perspective on what’s going on, but it’s like when you kind of think of it that way, it’s almost obvious because you know, if Universal Analytics is crap and non-compliant, and they don’t want to do anything with it, how can they justify the Google Ads pixels and floodlights, all based on third party cookies, the floors disappearing under their feet, what are they going to do?
[00:23:00] Daniel: Ed, last question from me, just really what what’s got you interested in GA4, like what’s the exciting part? What are you excited about your clients getting to use GA4 properly without having to think about it from a migrational perspective?
[00:23:11] Edward: Yeah so I think once the data’s all there and accurate, and obviously that’s what we help with. I think there are a few features that are new or newly accessible. I think the first one is the data-driven attribution models, so this was there in Universal Analytics, but it was limited to GA360 and it’s basically the ability for Google to use machine learning to figure out what is the actual additive value of that campaign in the user journey. So obviously a lot of the brands we work with it’s, you know, it’s not, they multi touch attribution. They didn’t just convert through one campaign, customers are having three to ten clicks before they buy something.
[00:23:43] Edward: So Google’s insight on which those clicks is actually the important click, I think is cool. Obviously you need quite a lot of volume of data to do it, so it’s only for the bigger brands. I think the other thing is Google’s introduced some really interesting predictive metrics. So learning again, making use of their machine learning with GA (Google Analytics), learning about what previous customers have purchased based on their previous behaviour, gives every user effectively a prediction score of how likely they are to purchase, but also how much they’re likely to purchase if they do purchase. And obviously we can help with that because the more data points you feed in the better prediction’s going to be, more insight into how they’ll eventually buy. I think that’s hugely valuable for retargeting and stuff. So you can basically build audiences of people who are likely to buy from me in the future, but haven’t yet.
[00:24:25] Daniel: Yeah I love those. The one thing that I hear all the time is basically just really want that to roll out for the non-ecom purchase events as well. I think that would be the kind of cherry on the cake really, but yeah, I love those, predictive audiences, those predictive metrics are fantastic. Again, it’s just, if you’re not e-com then what, and I think if they can make that B2B applicable or even just lead-gen applicable, I think that’s going to sell it for the last couple of percent and make it a hundred percent of everyone will be able to use these.
[00:24:49] Edward: Yeah, I guess that’s why we’ve always been excited about focusing on the eCommerce market is that if customers can get the event data collection, because the shopping journeys is so standard, there’s just a lot of insights that they can get and a lot of accurate predictions you can make based on behaviour to eventual purchase.
[00:25:05] Dara: Brilliant, well it’s been really interesting getting your perspective on this Ed. Dan and I talk about GA4 a lot, so it it’s been really good getting a very different perspective in terms of like how this is affecting your customers and where you see the opportunities and maybe where there’s still a little bit of room for development with the product.
[00:25:21] Dara: So to switch gears completely now, this is the bit in the show where we ask you what you’ve been doing outside of work lately to wind down when you’re not worrying about GA4, what are you doing to chill out?
[00:25:31] Edward: Yeah, it’s a good question, and I do worry about GA4 a lot. Yeah actually this time of year it’s been gardening for me. We bought a new house with a garden, well, just at the end of lockdown as a lot of people decided they needed that. And the funny thing about gardening is the kind of more you do, the more you have to do. So last year we spent ages kind of basically clearing out the garden and putting these stuff in, but now you obligated to do all this maintenance, perhaps there’s analogy for analytics right there.
[00:25:56] Daniel: I could say something similar, I’m growing my first set of tomatoes this year, but that wasn’t what I had written down, but I’m very excited about my tomatoes. I’ve got two tomato plants Ed, not quite a garden. I’ve got a tiny little box that we made out of a pallet, but nevertheless, I’m going to eat a tomato that I’ve grown for the first time, I’m very excited. What I’ve written down is I’ve just come off the back of three weddings on the trot. So three weddings within one month, and I’ve got a total of five this year that I’m going to, and it’s just been kind of like post-lockdown, delayed weddings, all kind of like, arriving at the same time, just like buses. And so, three weddings, one month and it’s been, it’s a been amazing for each of the couples and seeing three completely different weddings, but I’m knackered of traveling and doing all this kind of socialising. I’m looking forward to a couple of weekends of just sitting at home doing absolutely nothing. What about you Dara, what have you been up to?
[00:26:41] Dara: I think I mentioned on the show before, but we do a monthly work social, and we had Junes a couple of weeks ago. It was just before we finished up for the Jubilee, long bank holiday weekend and we went to see a band called The Old Time Sailors and they put on a really amazing show. So this was in Brighton, it was part of, I don’t know if it was part of the Fringe or the Brighton Festival, I’m not sure what the difference is between the two, it was in the Spiegel tent if anybody listening knows that and it was really good. We danced, we clapped, we had a few drinks. It was really, really good fun.
[00:27:13] Dara: So if anyone gets a chance to go and see them, I would recommend it. It’s The Old Time Sailors. One more question for you, Ed, and then your work here is done. Where can people find out a bit more about you or about Littledata if they don’t know already?
[00:27:26] Edward: Yeah, sure. Best place is our website, littledata.io, and also our blog, blog.littledata.io. Follow us on social, all the usual channels. We’re pretty active bloggers, we write stuff probably every week on this and that topic of analytics. So obviously GA4 is a big topic at the moment so please follow that.
[00:27:42] Dara: Definitely, and what about you Dan? Where can people find out more about you?
[00:27:45] Daniel: For me, it’s LinkedIn and my website, danalytics.co.uk.
[00:27:49] Dara: And for me, it’s LinkedIn. Okay, that’s it from us for this week, to hear more from Dan and myself on GA4 and all things analytics related, all of our episodes are available on our archive at measurelab.co.uk/podcast. Or you could just find them in whatever app you’re using to listen this.
[00:28:08] Daniel: And if you want to suggest the topic, or a guest we should be speaking to, we’ve now got a new form in the show notes, or you can find it at the website or just email podcast@measurelab.co.uk and yeah, we’ll be super happy and grateful to get any feedback and suggestions.
[00:28:22] Dara: Our theme music is from Confidential and you can find links to their Spotify and Instagram in our show notes. I’ve been Dara joined this time by Dan and also by Ed. So on behalf of all of us, thanks for listening and see you next time.
Daniel Perry-Reed
Subscribe to our newsletter:
Further reading
Customer Data Platforms (CDPs)
Grand Designs: “The Warehouse to CDP conversion”
