#93 GA4, the CDP of the future you didn’t know you had (With Rick Dronkers @ Data to Value)

Intro | Topic | Rapid fire

Intro

[00:00:00] Dan: In this episode, we’ve got Rick Dronkers from data to value back on the show returning for a second time, which is an achievement held by not very many at the moment. So we’re welcoming him back to talk about things like CDPs, GA4. I think I jumped off the starting line a bit quick in this. So thanks Bhav for pulling us back in, but we do talk about a lot of stuff. We talk about CDPs and GA4, as we mentioned, but everything in between, I think there’s a really, really insightful conversation with Rick in this episode. Really kind of digging into the evolution of GA4, what it means for marketing and this whole kind of theme of the year at the moment of a composable CDP. What do you guys think? 

[00:00:49] Bhav: I thought it was a really, really interesting episode. I think CDPs are a fascinating space. I remain a pessimist and a realist about everything when it comes to data including, especially around data technology and adoption around it. So I think I put some difficult questions out there to Rick. Hope you didn’t mind. 

[00:01:05] Dara: That’s what you’re here for, Bhav. You’re gaining the reputation as being the one to ask the difficult questions. I really liked it, my memories probably letting me down here, but I think this was the same the last time we spoke to Rick, but there’s always a little hint of what Google is up to when we talk to Rick.

[00:01:23] Dara: It’s like Google are arguably, or at least I think have become, so Google Analytics has become a better product analytics tool than it was without meaning to be, it’s becoming a CDP without, maybe without meaning to be, and we never know what Google are really thinking or what their strategy really is, but I, that’s twice now with Rick where I’ve ended up thinking what exactly are Google up to? Are they meaning to do these things or is it all just figure it out as we go along. 

[00:01:51] Dan: Probably the latter, to be honest, I am wondering if more unified strategy, I mean, having worked with Google for many years, I think knowing that there are two different departments don’t talk to each other, like the Chrome team deprecating cookies and Google Analytics, not knowing what to do about it, it’s just kind of interesting perspective. 

[00:02:06] Dan: Anyway, listen to the full episode it’s coming up right now. A couple of quick plugs, join us over on the CRAP Talks Slack community where you can chat to me, Bhav, Dara, and many, many other people. Talk about analytics, CRO, testing and everything else, there’s a link in the show notes. And lastly, I’ll keep plugging this for this season, but we’re giving it a go, we’re a vodcast and I still can’t bring myself to use that term seriously, but we are on YouTube. And if you’re listening to this on Spotify, you can actually click a button and watch us do this. So if you want to watch us, I don’t know why you’d want to do that, but please do. The more people that do it, the more likely we are, we’re going to continue to do this. But if that’s your thing, then it’s there for you as well. All right, on with the show, hope you enjoy. 

[00:02:44] Dan: It’s a warm welcome back to Rick Dronkers for his second appearance on the show. Welcome back, Rick. How are you? How’s life? How’s business? How’s everything going?

[00:02:53] Rick: Life is amazing. Father, so busy on the private front, but really enjoying that experience. Business a little bit less as you can imagine, focused a bit more on the household, but yeah, really enjoying.

[00:03:07] Dan: Awesome, congratulations, Rick. Yeah, that’s awesome news. And yeah, well, you must be doing something right. Or maybe I’m doing something completely wrong because not only if you had some time off and become a father, but you’ve just been everywhere. You’ve been publishing content. You’ve been active on LinkedIn, on everything. And one of the things that got me really excited and asked you to come back onto the show to kind of do the second round of this is, and there’s a couple of by the way, I’ll be linking all of this in the show notes of the podcast. So anyone listening, you can have a read of those obviously after we’ve heard from the man himself. 

Topic

[00:03:37] Dan: But there’s a couple of things you’ve been talking about specifically around the evolution of Google Analytics 4, and the way that it’s developing and the kind of tinfoil hat of like GA4 is changing, it’s happening. It’s all coming to a bit of a head. So I just wanted to kind of say awesome work on that. I think the writing what you’re portraying and the way you’re explaining it is not fear mongering. It’s not scaring people necessarily, but it is something that I think is something I completely agree with, which is GA4 is changing. It’s going into this world of the CDP, well, Google CDP. And yeah, I just wanted to go over like where’s all this coming from? What’s your thoughts around it? And, and I’ll stop talking. 

[00:04:12] Bhav: Maybe before we go deep into this conversation of CDPs Rick and Dan, I think there’ll be a large volume of users who may not be familiar with the term CDP. And given that it is the nature of this episode and we’re going to be talking about it extensively. It might be worth just doing a quick definition of CDP, Rick. And you know, given that you’re, you know, you’re the one that we’re going to be grilling about CDPs. That’s why I’m really here, to learn about CDPs. 

[00:04:40] Rick: I’m actually, I wouldn’t say I’m a CDP expert, but a CDP stands for customer data platform. And it’s basically, it’s in my, in my eyes, it’s like the next step that we went to from like the data management platform, the DMP, I don’t, I don’t know if you guys remember that trend. And then if we’re talking buzzword bingo, it’s the customer 360 view, right? Like the CDP basically has all touch points from the customer, ideally all your systems, right? So every way you interact with customers should be stored in the CDP. That would be the ideal scenario. Oftentimes when I see CDPs implemented, they just only capture web interactions, right?

[00:05:18] Rick: So they’re not that rich, but in the ideal scenario, they would capture like if you interact with your customers via SMS, via email, via obviously via your website, perhaps via in store communication, whatever, right? Like it would all be fed into the CDP, all tied to the user ID, which is, you know, from our measurement point of view, the holy grail, and then you get to make a segment that says, I only want to send this email to everybody who didn’t purchase in the last 30 days, but has an average purchase volume of at least this and, you know, all these kinds of cool segmentation features and then blast it out to your email system or to your SMS system or whatever. And that’s, I think way it gets sold to customers, let’s put it like that. 

[00:06:00] Bhav: So it’s kind of a way to unify all of your data and really enrich it with more information about users. Whereas historically, you would have had the same data about the users, but you wouldn’t have been able to use it. So for example, if I was doing, if I was trying to score leads I could score leads based on their user behaviour on the website and then say, okay, these are warm leads based on the fact that, okay, they’re on a free tier, but they’re, they’ve been extensively using the platform. They’ve created a bunch of stuff they’re in it almost weekly. And then you join that together with, you know, any CRM comms that you have, and then you kind of send them a very specific email that’s tailored to them. Is that right? 

[00:06:39] Rick: Yeah so I think it’s the unification of data about the user from multiple systems, but then tied to that is the actionability, right? So it’s the segmentation and the taking action. So the most standard example is sending an email, but you know, it could also be an SMS or send them a postcard, whatever, right? But the action part tied to it and the segmentation part tied to it in one system, and then I think that also opens up the door to the, to the next term, which is the composable CDP, which is something that has been popularised by Hightouch, which is basically having a CDP, but not buying it from one vendor, but instead having an analytics vendor and then a segmentation vendor, and then a reverse ETL, which is the deactivation layer. And then having an email marketing platform and whatever, right? So then you have separate vendors for all these aspects, but they all work together. And then you have a CDP, but it’s composed sort of. 

[00:07:35] Dan: It’s never really, it’s never really completed. It’s just worked on. And I think this ecosystem is kind of like the evolution of a warehouse, data warehouse, or a CRM and it’s kind of like the love child of the two. It’s kind of the natural evolution of the one product, it’s not one product. And I think this is where you can buy off the shelf CDPs. And I think this is kind of starting to drift towards the kind of opening question really of around where GA4 fits into this, but you’ve got the kind of build or buy methodology.

[00:08:01] Dan: The composable side is this kind of modular Lego brick system where you can swap out your event collection, your activation layers and things like that. But I think this is the other option, which is you can buy or use freely available or cheap software that says it does it all in which you become kind of locked into the vendor.

[00:08:18] Dan: And I think there’s pros and cons, but this is very much where Google Analytics 4, and the Google marketing ecosystem is going because web analytics and now app analytics, I suppose with GA4 is one thing, but what it’s doing is it’s blooming out of that. And it’s basically saying, okay, well, we can activate and again, this is where it’s kind of narrow compared to the idea of the composable CDP. 

[00:08:40] Dan: Like it’s going to activate through the Google ecosystem, but even then it’s starting to make waves so that you can activate beyond that. And I think this is the fundamental shift in my thinking around where Google Analytics 4 fits within that ecosystem, where it fits within that CDP ecosystem, because now it’s not just about Google Ads audiences anymore. Or at least it won’t be, it’s actually now starting to become a centralised audience hub, you know, this segmentation machine, or at least that’s where I feel that it’s positioned itself to. 

[00:09:09] Rick: I think it’s twofold, right? So in my opinion, CDPs are not successful because I’ve never really seen one successfully implemented fully. They are very successful on the sales side. And I think the reason is because from a sales perspective, it really makes sense, right? Like the person who has to extract value, like let’s, let’s call this person, the CMO, right or the marketing director, whatever they have budget and they want to create actions that generate value.

[00:09:38] Rick: This is finally a system that covers it all right. So from a sales perspective, like these people understand like where they want to touch their customers with certain messaging and before it was all fragmented and then they had to invest in all these separate steps and now they get sold the dream, right? So the CDP from a sales perspective is perfect.

[00:09:57] Rick: I think the composable CDP makes way more sense from a practitioner’s perspective because every system that can do it all. You know, by default, you will, you know, the practitioners will tell you, like, if it says it can do it all, then it doesn’t do everything at a hundred percent, right? You will be lucky if it does everything at 80 percent and probably less. 

[00:10:16] Rick: So you’re going to sacrifice quality. And the price tags of CDPs are, are pretty hefty. So yeah, the composable CDP makes a lot more sense. And I think that’s what Google also noticed. And yeah, it’s pretty interesting to see that Google actually opened up Google Analytics 4, right? So before Google Analytics has never been able to talk to any other system, basically just outside of the Google ecosystem right. And last or this year they launched the audience API and the user data export to BigQuery, which are both kind of the same thing in a different pipeline which enables you to create a segment and share that outside of Google Analytics and potentially import it into your email marketing system, for instance right. 

[00:11:01] Rick: So it’s like a first, it’s like a first baby step to, to composable CDP functionality. And it’s pretty big that Google is opening up their ecosystem to potentially integrate with other tools. The first use case is A/B testing, right? So they deprecated Google Optimize and the audience API is also what fuels for instance, convert.com to integrate their A/B testing solution with Google Analytics. But there’s obvious other use cases available once you open up that flood gate and they open it up for everybody, not just for them so that’s really interesting. 

[00:11:35] Bhav: It was a strange decision when Google decided to sunset Google Optimize. From your point of view, do you think that was a, there’s a commercial risk because sunsetting Google Optimize, and opening up the APIs to connect to other A/B testing platforms or whatever it is, you know, you want to integrate with, do you feel like they, there must be something that’s, what is the end goal from your point of view about why Google might make it a bit more compatible with other platforms instead of just sticking to their own?

[00:12:05] Rick: Yeah, I honestly don’t know right, so it’s all speculating. I think that the deprecation of Google Optimize kind of does make sense to me because A/B testing and supporting that tool is probably pretty cost them a lot of resources right. And it’s a really specific niche to support. And I think, you know, integrating with others doesn’t really like those, those other companies are not really a danger to Google right so I think that’s fine. So that part of it does make sense to me. I think the opening up it might be related to European legislation about digital, you know, the digital markets act, the digital services act about, you know, more transparency and more, more openness, like a preamble to, you know, what might be coming.

[00:12:53] Rick: But it might just be Google taking a different approach, right? I think Google Analytics is probably still a loss leader for Google. I don’t think GA 360, the paid version covers the bill for all the free versions. So, you know, the end goal is probably still to get more people to advertise on Google.

[00:13:11] Rick: And from that perspective, maybe they figured by making Google Analytics integrated with more tools in the end, we get more people into Google Analytics and in Google analytics, we can highlight how well Google Ads does for people. So they spend more with Google Ads, right? Maybe that’s the line of thinking.

[00:13:28] Dan: I think it’s not just even Google Ads now. And I think this is where, for me, that the kind of shift has happened is that by, by making this pivot so that it opens up its measurement capabilities and integrations to other platforms, we know that Salesforce Marketing Cloud, for example, is another example of the A/B testing tools that you mentioned, Rick, and others that will come out down the line is that in a sense, they’re enabling any new technology to be spun up to without having to deal with the measurement side of things, you can just plug and play into the Google Analytics 4 ecosystem, which most websites already have. And especially with the, as of the time of recording, the very recent news around things like enhanced conversions and encrypting things like email addresses, names, and addresses, and send those into Google Analytics. Because what they’re now doing is they’re now making this a kind of user level reporting tool, which then powers their ecosystem across ads.

[00:14:16] Dan: And so it’s, I think it’s quite clever because it’s not just about spending money in the Google ecosystem. It’s like, well, can Google as a business continue to track and measure users across sites, across websites and cross whatever. And they’re in a uniquely prime position to do that because every website, every app that uses Google Analytics, but it, all it has to do is enable this thing called enhanced conversions.

[00:14:37] Dan: All of a sudden it’s capturing unique IDs, like email addresses everywhere. And it can continue to do that, right? It can continue to do this stuff that it’s always been doing. And it needs to make the tool look sexy and alluring and make people want to use it or force people to use it or make the other technologies dependent on using it so that it never goes away. Because I think this move like a lot of this stuff they’re doing now is like you don’t need to implement Google Ads pixels you can do that all through GA4. You don’t need to implement, for example tool X, Y, or Z pixels, you can just do the GA4 stuff and sync it over here. And so now, like they’re making it like literally at the centre of this digital kind of ecosystem. And I’m, I’m thinking that they’re playing the very long game here, that’s my thoughts anyway. 

[00:15:20] Rick: Yeah I think I agree with you that I think that everything seems to be going towards right? So everything seems to be consolidating towards one tag, right? The Google tag or it’s really confusing right now. But I do think it might backfire to be honest. One thing that is confusing to me right now. So recently Google announced that enhanced conversions, right which is like you said, is adding first party data, often a hashed email address, but it can be other data points, which you could already send towards Google using offline conversions or using the Google Ads pixel. Google Analytics will now be enabled to also send this data to Google Analytics.

[00:15:58] Rick: And right a week before that announcement, there was a webinar by Google about consent mode and updates to consent mode. And that GA4 from March, 2024 will require consent mode to be fired right, so to be configured. And the whole consent mode thing seems to be tied to like the enforced consent mode seems to be tied to European regulation. If you want Google Analytics to be on every website, then you shouldn’t make it a marketing only pixel in my opinion, you should make it also usable as a purely statistical analysis tool. 

[00:16:38] Rick: It seems like Google is forfeiting on that, right? In my opinion, what it seems like right now is they basically accepted, okay, we’re just, we’re marketing analytics, right? We’re not, no longer analytics, we’re marketing analytics. And you will need marketing consent, like opt in marketing consent in the EEA to fire Google Analytics, which in my opinion, long term will result in less companies using Google Analytics, at least less than before right. 

[00:17:08] Rick: Well, the market share was like 85 percent or more. So a little bit less probably won’t hurt them in the, you know, in the broader scheme of things, like if it drops to 70, so I don’t know, you know, what that result will be, but I do know that I have clients that are actively considering this, right. That are thinking like, should I, do I use the marketing, like the advertising side of the pixel, or do I just use it for statistical analysis?

[00:17:31] Rick: If I just want to understand how people are behaving on my website and what they are doing, is Google Analytics still the right choice for me? Even though it’s free, like people are starting to understand the implicit costs, right, of the free. So from that point of view I’m not sure if Google has another option, right? Because I think it’s pretty clear what the EU is moving us towards, like a consent only tracking, right? And maybe even those other solutions that are fully hosted in the EU, fully privacy minded, maybe even those will go towards like requiring consent before they track. It does seem like a limiting choice by going down this advertising route and adding enhanced conversions to GA4.

[00:18:14] Dan: Is it just an ROI thing then do you reckon for Google? Because as you said, it’s a bit of a loss leader for them. All of their customers that are using the features for Google Ads bidding or the GA 360 customers. So in a sense, their revenue stays the same if that cost base goes down because some people that aren’t using them are going to leave. If anything, their kind of profitability in that this tool goes up. And I honestly wonder if they just don’t care, like you said, it’s not going to make so much of a difference for them. It’s like, okay, well, the customers that were only going to kind of unpick it to use it as a pure analytics tool that’s not what they’re for. I’ve been kind of like saying this for a while. And this is why it kind of resonated that the readings of your writing, Rick, is that this is more than ever a marketing tool. 

[00:18:54] Dan: This is a marketing technology and for example, one of the more recent updates they’ve done is that you can now create audiences. You can create GA4 audiences from within the Google Ads interface. And there was, they pulled it quickly after I think they accidentally released it before they meant to, but they also released this idea that key events are going to replace the concept of conversions in GA4 and thus they’re going to keep the term conversion, but it’s only going to be relevant for stuff that you activate through your advertising campaigns in Google Ads.

[00:19:19] Dan: So now a conversion is only things you use in Google Ads, let’s say they roll that out down the line. And all this other stuff that’s happening, I’m just wondering whether or not this is becoming a you know, they could rename it, for example, Google Ads Analytics. Do you know what I mean? Like it almost doesn’t matter what the tool is. The interface of GA4 is a marketer’s tool for building audiences and measurement capabilities in the advertising ecosystem and beyond if it becomes kind of more of a CDP for other technologies, I’m just wondering that, you know, the GA4. BigQuery export kind of scratches a bit of the itch from the kind of pure analytics people. But you’re right there is many other technologies now I’m wondering this idea of the, the composable CDP nature, maybe GA4 fulfils the audience building list stuff, not the event collection, maybe from an analytics perspective, this isn’t the analytics tool anymore. 

[00:20:08] Bhav: I mean, I’ve always felt like, I think we’ve covered it before if you do use Google Analytics as a composable CDP, you’re able to bring in multiple data sources, which you otherwise wouldn’t have brought in before, and then, because everyone sees the value in like connecting your CRM data or your product analytics data with your marketing data, and then you push that into GA4 because it’s, you know, effectively it’s free, but then you realise that actually GA4 is limited what you can do from a analysis perspective, yes, but also, you know, for most people, they want to take some action at the back of it, which means you then have to rely on BigQuery. And I know I keep going down this, I keep coming back to this route.

[00:20:46] Bhav: And Rick, you actually called it out in your, in your post that I was reading Google Cloud is probably the end goal for them, having more, and the CDP becomes an enabler for that because you connect more and more data sources to it, you need to store that data, it’s too much for Google Analytics to handle, therefore they, you know, they sample the data and they do all those annoying things that we hate as analysts, but the idea is that we push it into Google Cloud, and that becomes the you know the thing that pays for everything else, in my opinion. And Rick, I’d love to hear your thoughts about it, because you called it out on your post. 

[00:21:21] Rick: I think on the, the point that Daniel just made, like, I think maybe Google Analytics already is slash already was the tool for marketers, right? Like I think it has always been used by mainly by marketers, but also analysts. And I think that is just, you know, it’s moving a bit further towards the marketing side, I would say. I think the Google Cloud part, I’m not sure if it’s like you know, a pre-aught strategy, right? Like okay, this is, this is the way we’re going to push people into Google Cloud, but I definitely know that, that Google was happy with the amount of marketing organisations that are now starting to leverage Google Cloud.

[00:22:00] Rick: And you see, there’s now a Google Cloud Marketing analytics certification, right? So they’re, they’re like, you know, they’re also catering towards that, like building marketing use cases on Google Cloud certifying agencies for that so that’s obvious. But I think the issue I have with that is that in GA4, a lot of decisions have been made on, I think, reducing the cost of the tool, especially the free version right. And we see that in other and not assigned, right. And all the issues there are in GA4. 

[00:22:37] Rick: And that makes sense, right? If you’re going to run a free tool on all the websites on the internet, I don’t even understand how like the data set of Universal Analytics, how big that must have been right. So I think like from a cost perspective, cleaning that up is already like yeah, that’s probably saving them some money right. So I understand why did the new GA set up you know, from a building perspective, they’ve perfected it to handle more data in a more cost effective way, which turned out to hurt analysis, right and gave us some features that we didn’t like. 

[00:23:08] Rick: If we’re going to use Google Cloud, right? So if we’re all using GA4 as a segmentation engine, right? So, so let’s say Hightouch has like this product called Hightouch personas. I think they call it, and I think Census has the same thing, which is basically like the audience engine in GA4, right? So you can select, I want to have a user that has done this and this in the last 30 days, and then I want to reach them somewhere you could argue that if GA4 becomes a composable CDP, then the interface of GA4, you know, you would only use that for the audience engine, right? For segmentation purposes.

[00:23:42] Rick: But then I think the downside is if we’re going to use GA4 to get the data into BigQuery, I don’t see why we would use those limitations of GA4 to get the data into BigQuery, right? Because then we’re still dealing with the fact that it’s a marketing tool that we need to consent for everything like, right. So there’s a lot of limitations at the front end. So then I would rather look at Snowplow or, you know, something equivalent to get the data in, get higher quality, higher fidelity data, all first party, perhaps even on a fully owned infrastructure to get it in. And then after that implement something else on the backend for like you know, like something like high touch or, or census to the users that did give consent to use their data to, to send it to other platforms 

[00:24:29] Dan: Do you think that’s a maturity thing though? So I’m thinking that there’s an awful lot of businesses out there that will be like 100%, yes, I agree, Rick. You know, let’s go down the Segment the data warehouse route and the activation or the reverse ETL activation through something like a high touch. But then I’ve got, you know, and I can imagine because your agency, Rick is very similar, I think, to how Measurelab operates in terms of if they can do that, they’re not coming to us for help like they’ve already got that sussed. And actually a lot of the people I end up working with are going to be the ones that are like, we don’t have the money for an internal analyst or an analytics team or things like that.

[00:25:02] Dan: And actually having the plug and play nature of the Google ecosystem is, is kind of where they’re at. And although they might be aware that there’s a better way of doing it, it’s just realistically just not possible in the short term. So I’m wondering if there’s enough of that market there for Google to be like, cool, plug and play, spend with us, measure with us and do that. They’re going to lose the higher maturity end to, to the kind of the kind of composable CDP market. But I’m wondering if they’re just kind of thinking that’s okay there’s still a crap ton of people in this bucket that we’re going to make money from. 

[00:25:36] Rick: Yeah, a hundred percent right. I think the market is so big, right. There’s these statistics, right of like the, how much of the internet runs on WordPress and then Google Analytics, and it’s like, I don’t know, I can’t even pronounce the number, right. How big it is, how many websites have it. So from that point of view, right. Obviously yes. On the flip side of that, I do think you see technology being invented at the higher end and then it always drips down the market right. So reverse ETL is becoming cheaper and cheaper and easier and easier right. 

[00:26:07] Rick: So all these technologies are also coming downwards basically and I think it’s a maturity thing, but it’s also a, what’s the value of your data? What’s the potential value of your data thing, right? So how big of a lever is data in your organisation? Right. What kind of actions can you influence with data? What kind of moments can you create based on data, based on having the right data at the right moment? Right.

[00:26:35] Rick: And if as soon as that potential value goes up, then the potential investment in it also goes up right. So it’s more about unlocking and identifying what that can be rather than the actual maturity where the organisation is at right now, I would say. 

[00:26:50] Dan: So it’s more of like a marketing maturity because it’s kind of going to be like, you need to be able to action this enough to be able to justify the spend. And it’s the age old conundrum of the profitability or the return on investment for analytics or this kind of data work, right. Which is like the impossible question or the one that also has too many answers of like, you need to be able to, it’s something we always say at Measurelab is like, you know, we can dedicate five people full time and throw everything at the wall to work with you on your data ecosystem, your MarTech solutions and all that stuff.

[00:27:20] Dan: But their side, they could only feasibly pick up one email a day. So how useful is that going to be for them? And so in terms of this profitability, it’s about scaling that and saying, okay, well, from maybe this marketing ecosystem, we go down that path with this CDP route and saying, if we can make, if we can make a 5 percent difference on your revenue then that justifies this spend plus some and it makes it a no-brainer. 

[00:27:40] Dan: But if you don’t have that end goal in mind, or if you don’t have the marketing maturity, or if you don’t have the organisational maturity to be, oh, cool, if you can’t even shift a budget from search to affiliate to display, then there’s no need for this kind of stuff because it’s not going to be useful. But if you’re fighting for your own budgets and to keep them next year, and it’s like, oh god, we need to spend it or we lose it, and that’s the wrong kind of place to be, to make use of this right? And we’re going down the right path there. 

[00:28:03] Bhav: I’m curious just to understand from your point Rick, if customers start using Google Analytics as their CDP and it’s done in this like composable way, you kind of eliminate so much of the need for everything else that Google offers, right? Because you can have your data sources coming from wherever you like, and you can have your end points being wherever you like, and they don’t necessarily have to be Google. So this, which is why I think the BigQuery endpoint is such a important point for them because ultimately, otherwise it becomes a free tool, right? If you give customers a free CDP and you make it composable, I’m really struggling to understand why if BigQuery isn’t the end goal. So I know I’m kind of like taking it back to BigQuery. 

[00:28:56] Rick: I fully agree and I get what you’re saying. I think currently it’s far from a CDP, right? To be clear. I think they’re making, they’re making moves that could, that could end up in that situation. But, to be clear, currently you cannot activate anything other than a Google tool from Google Analytics, right? The only thing that has changed is that now you can export a list of users with their ID to BigQuery or via the API to wherever you want to, and you could enrich that ID with the same ID, hopefully you store the same client ID as well in your CRM, right? And then you can enrich it with whatever you have about that user. And then you could use that list and send it to Facebook for the conversion API or whatever right? 

[00:29:47] Rick: Like, so there’s currently still a step in between and obviously Google would like it if you do that step in between on top of GCP, right. But to be fair, like you could, you know, like using the API, you could get it out to do it somewhere else. Also the API is not real time right now it’s still on a daily basis right. So like if you want to do a, if somebody ordered, I want to email them within one hour, you know, you’re out of luck, right? It’s not that kind of thing yet, you know, perhaps. But yeah, it’s heading that way well, so the paid the GA 360 accounts, their BigQuery export of the user data is I believe now they have a checkbox to go to streaming. So you know, that’s a paid feature, you know, GA 360 costs, costs money. 

[00:30:32] Dan: A lot of money.

[00:30:34] Rick: Yeah well, less than it did before right? Like there, there are now a lower, lower tiers available I think. 

[00:30:39] Dan: Sorry, I’m going to jump in, but yeah, absolutely. I think that’s the way it’s going Bhav, but I think that what my tinfoil hat, and this is, I think I’ve been predicting this for the last two years is that they could make such a big impact as if they turned the monetization of GA4 into a similar model to Firebase Analytics, or sorry, Firebase, as the ecosystem, as a kind of pay as you go model. So at the moment, it’s kind of like 360 or free, there’s nothing in between. And we had a lot of issues when it came out with like the API quota, you know, building dashboards in like Looker Studio and that breaking because we run out of these magical tokens. 

[00:31:11] Dan: The amount of people I know, the amount of businesses I work with, if they could spin up a hundred or $200 a month direct debit to Google just to get their dashboards working again they would jump at that opportunity. They don’t need the thousands of dollars a month, 360 licence with all the bells and whistles they just need that one thing. And so with this as well, like for example, imagine if you could just, you know, use the free version. It’s like, I’m only using it for the real time streaming to BigQuery. Okay, well, let’s just pay for that feature. If they started going down that path, they’re going to piss off an awful lot of sales partners for one. So I don’t think it’s a very easy and quick decision, but they could, they could make such a momentous difference I think just from that one update, I think, and also I’m wondering if overall they might net positive in the kind of the ROI for the products, because if everyone’s paying 50 dollars a month, how’s that different to paying 50, 100 dollars a month for piano analytics or Matomo or something else. You know, it’s just, it just becomes another paid tool. 

[00:32:03] Bhav: Okay I’m going to change the conversation slightly to something you said right at the beginning, Rick, and it caught my attention and I was not going to let it go. You said that in your experience, you haven’t been able to get CDPs actionable right. And it’s just, it’s failed every, every step of this in every step of the way. I’m a listener and I’ve been sat here talking about CDPs and how great they are and yet, you’ve never been able to make it work to it, to the, in the way that it’s been sold. And I, by the way, for what it’s worth, I’m in the exact same shoes. I’ve seen all of the potential a CDP has to offer, so the ones I used that previously was Snowplow and Segment, but I’ve never seen that utopian endpoint that was sold to me. Why? 

[00:32:47] Rick: Yeah so to clarify my statement, I’ve never seen a CDP at a client that was ROI positive right? And I think the main problem with that is that they are sold at over here, right? Like they are so expensive, like the price tags I’ve seen for CDPs, like, for instance, TreasureData. It’s wow I didn’t know software could be that expensive right?

[00:33:12] Rick: So you mentioned Snowplow and Segment, they don’t really fall into that category, so I think, I think they are, first of all, they are often like cheaper or they can be cheaper right. Well, obviously Snowplow can be free it’s open source. Well, not really free you have to host it. And Segment, you know, you have levels to it. And I think after they got acquired they’re moving more towards the CDP, but like those full, full fledged enterprise CDPs I think the big problem is they get sold to large enterprises and they get sold on the dream.

[00:33:39] Rick: And then people start paying and usually it’s like a two year contract to get started. So then you know, you’re down a lot of money and you need to get a return on investment on that right. And okay the enterprise is big, right? So hopefully you can leverage a couple of touch points across millions of customers and then, and then get a return, but it is really challenging to make the return mostly because getting an enterprise to adopt something is really hard because everything needs a stamp of approval and legal needs to approve. And so before like you implement the tag and then six months later, somebody is able to start using it right. That’s I think the main problem with selling something with such a hefty price tag and then getting a return on investment. It’s more about the price versus the way of it’s being implemented. Obviously there are going to be success cases somewhere, but I haven’t seen them yet. 

[00:34:35] Bhav: For what it’s worth, I don’t think it’s just an enterprise issue. I was working at a startup and we were, I think we, you know, we were thinking about Segment as You would CDP, you know, we had all ambition of being able to trigger emails off the back of some actions that users have taken on what they were doing on the website. And, you know, we were enriching data that we were collecting. Admittedly, it was a very expensive data pipeline but the potential to, like, operate and use it as a CDP was there, and I think that’s what we were aiming to get to, but it just, I don’t know, it just doesn’t manifest, and I can’t figure out why. Dan, this is the technical equivalent of my self-serve pitfall, I think. If there was another failure in life, I think it’s real actionability. 

[00:35:22] Rick: But I think the problem is it’s really hard to work backwards from that desired end state, right? The step is too large, right? To bridge all at once. So let’s, let’s say an organisation already has a decent analytics setup. They already doing some attribution, re reallocating of budgets. They have actionable people on like email and on social media and on paid search, whatever. And then they introduce a CDP or like a CDP layer, like to the reverse ETL layer, so to say, then it’s going to work, right? Because they already took step one, two, three, and then they’re going to add that activation step on top. I fully believe in that step. But if you start out with like, okay, we’re a bunch of people and now we’re going to adopt this huge system and we have to go through implementing analytics, trusting the analytics, building reporting, getting people to take action on a daily basis in whatever they do, whether it’s email marketing or you know, something else.

[00:36:13] Rick: And then we’re going to add this advanced segmentation engine to it, to optimise it. It’s too much to do all at once I think. I don’t think we as people can organise well enough to jump that whole thing. 

[00:36:27] Dan: It’s the people, exactly as you said, Rick, it’s like, people are slow and and things like people have like annual budgets to kind of align to. And so things like financial years are quite often quite locked in, in terms of changing processes or flows or budgetary decisions. And I’m not saying that’s a good thing. It’s just, I think the technology might be able to be spun up and used. And I think everyone can agree that end state is going to be better than the current state, but actually the human beings that need to use these technologies and we have to change the, we have to get IT, legal, marketing, the data or analytics teams, the data scientists, maybe all need to be behind this. And I think that is the slow thing. 

[00:37:03] Dan: And I think if, if you’re holding out for this one golden holy grail at the end, that’s where you see such a kind of a difference between start and kind of value that, like, as you said, Rick, if you can make quick decisions it’s truly working agile, right? In terms of like failing fast, failing often, or succeeding fast, succeeding often, because I don’t think you can hold out the investment or the time or the patience. And people might even move on from the company in that time. 

[00:37:27] Rick: People will leave, you know, right before you get to that point where you need to be the, the analyst leaves. You can start over on that front again, right? Those kinds of things. 

[00:37:34] Bhav: But I’m going to challenge you because I have to, but then what you described of having all of those things in place, right? You’ve just described an enterprise where they’ve got the analyst in place, they’ve got the social media in place, they’ve got the marketing, but you know, the people who are ready to pull the triggers but we, you know, we all agree that it takes six months to even just get from the point the contract is signed to the point when the first person can even take some action.

[00:37:59] Bhav: It’s not as big a step, but the way I think about it is if a startup can’t do it, and they’re the ones that can be lean, they can be agile, they can operate quickly and make these big decisions. If they can’t do it, what hope does a big organisation, which has all of these established things already in place, but trying to get organisational changes, like trying to turn an oil tanker in the middle of the Thames.

[00:38:22] Rick: Yeah, I think two things, right. For a big project, any project, not even if it’s just marketing, you need top down ownership, right? So you need like a CMO who is really involved and who is driving change, who is going to be at legal’s desk saying, I need this. I need this every week. I need this, I need this, right? So I think, so they get it signed without that. If it’s a flat organization might work out, but once it’s layered organisation, you’re going to fail because you’re going to get stuck in one of the layers. And besides that, I think the total approach to like a CDP basically solves multiple problems in one solution.

[00:38:57] Rick: And I think that is a, like I said, it’s from a sales perspective. It’s great in theory, but I don’t think that’s how it will work in practice. So I think a composable CDP or just calling it in the separate steps as they are, that makes more sense. So year one or Q1, depending on how fast you are and how many resources you got, focus on implementation, next step, focus on making sure analytics dashboarding is great next step, right?

[00:39:24] Rick: So building it up in layers and then in the end you have a CDP, right? But you’re not calling it a CDP, but you’re, you’re still there, but then you’re chopping it up into these small actionable pieces that people can commit to and drive instead of committing to this, you know, end goal somewhere that nobody will see because everybody left the company already.

[00:39:45] Dara: Are a lot of businesses not pretty close to that already, though, the idea of this composable CDP. Like I guess there’s something missing, maybe, maybe there’s still too much manual process involved, but I get your point completely about the end to end, big cost CDP’s and how difficult they would be to get implemented and bought into. Would you not say most businesses, I’m saying most, maybe not most, but many businesses are close to having a composable CDP in place, even if that’s not what they’re calling it. They might have it without even realising it. 

[00:40:15] Rick: I think especially with GA4, right, because they have the BigQuery export, whether they know it or not, they are closer than they might think. So I think, you know, what do you need? Let’s say you’re advertising on Google Ads and Facebook Ads, right? So you’re going to need to export the data to BigQuery, right? So you use Supermetrics or Stitch or whatever kind of ETL tool, right? You get your Google Ads and Facebook data into BigQuery and you have your GA4 data into BigQuery.

[00:40:40] Rick: So you basically have your cost part of everything. You have your behavioural analytics, part of everything. The last thing you need is your reverse ETL tool, right? So look at something like Census or Hightouch. And then you obviously need something where you’re going to send people, right? Where you’re going to take action, right?

[00:41:00] Rick: So it could be your remarketing list in Google Ads or something, you know, the equivalent in Facebook Ads, but it could be your Hubspot email marketing or whatever, right? You’re going to sync data towards that. Now, obviously there’s steps in between and, you know, you could group that as like data engineering, right? So the data you pull from Facebook Ads and Google Ads and the data you pull from GA4, you’re going to have to make that usable and tie it all together to make it usable for the reverse ETL part. But yeah, I think most companies have a lot of like a basic CDP already, I would agree yeah. 

[00:41:38] Bhav: The one that springs to mind from a use case perspective, we talked about you know, the challenges of implementing a CDP, we’ve talked about you know, what the future would look like, potentially who’s doing it, what are the major use cases from your point of view of a CDP?

[00:41:53] Bhav: So the one that obviously springs to mind for me is personalization, but are there any other ones? Because I think personalization is one that CDP’s get sold largely on and it’s usually the hardest to achieve. 

[00:42:07] Rick: I think if, like, broadly speaking it’s why I called my company data to value. I think it’s the action part, right? So what I ran into is like, we collect a lot of data, then we transform a lot of data, then we analyse a lot of data, but if that analysis is not turned into action, then there will never be any value like it’s just impossible. Like even like we might get smarter, right? But there’s no actual value for the business at it.

[00:42:31] Rick: So it’s really hard to actually bridge that gap I’ve noticed. Like to make the full loop from like collection to actually taking action and for a variety of reasons, right? We always want better data, more perfect data so we get stuck in that circle. Or then when we’re doing conversion optimization, we get stuck in like, what’s the perfect way to determine if a test is valid or not right. All this kind of stuff where actually we should be taking action, right? Make more repetitions. And I think that’s where a CDP shines because it is closest to the sources of action, right? It is closest to the places where we can take action as marketers, right? So send a better email, target a better audience, not target this audience because they’ve already purchased or they, they’re unlikely to purchase or right, whatever.

[00:43:20] Rick: So these actions are the actual valuable things that we should be influencing with our analysis and with our data. And I think that’s the value of a CDP or a reverse ETL, whatever you want to call it. And I think personalization is obviously, you know, part of that, like, okay, for this user, the website should look like this because that was more likely to make them spend more right? So yeah, the value is really obvious and then sometimes it’s really hard to do, right? Because if you know what the website should look like, then making the website look like that is the next step, right? And I’m not a developer, but obviously it’s not as easy as we would like it to be. But I think that it’s close to taking action, that is the value.

Rapid Fire/Outro

[00:44:00] Dara: Okay, Rick, we’re going to put you in the hot seat now for our rapid fire five questions. So question number one is what’s the biggest challenge today that you think is going to be gone in five years time?

[00:44:12] Rick: I think consent, weirdly enough. I’m not sure if it’s five years or 10 years, but I think consent is like a really big headache right now, but it doesn’t make any sense whatsoever. And I think we’re going to have to find out a different solution for it. 

[00:44:25] Dara: Okay so what will be the biggest problem then in five years, once that’s all sorted and consent isn’t an issue anymore, what’s going to take its place?

[00:44:34] Rick: Data activation. Yeah I think that will remain the, the not necessarily problem, but the highest leverage thing we can achieve which will always be the hardest thing. So how to actually turn all that, all that wonderful data into something valuable. 

[00:44:50] Dara: Okay next one, Rick, what’s one myth that you’d really like to bust?

[00:44:58] Rick: That data is the new oil.

[00:45:05] Dara: A slightly similar question, but if you could wave a magic wand and make everybody know one thing, what would that be?

[00:45:12] Rick: I think that there’s not, that there’s not a truth to be found often it’s more about taking action, less about being right. 

[00:45:22] Dara: So instead of single sources of truth, it’s actually, there is no truth. 

[00:45:27] Rick: Yeah, or there is your truth and that’s fine, right? Like stop losing time trying to find, trying to perfect your truth and instead take more action and figure out what happens once you take those actions.

[00:45:38] Bhav: I think this kind of leans on that deterministic versus probabilistic conversation we were having. We didn’t go into any real detail of it, but all of the data only, you know, the best you can ever get to is like some type of high probabilistic outcome that you have to make decisions off of, right? You’re never going to get the perfect truth. 

[00:45:54] Dan: It’s about being happy in that world, right? Rather than trying to have 100 percent accuracy of data of like, you know, sessions and users that’s irrelevant, but change and effect is the important side. 

[00:46:05] Dara: So the last question, I’m going be a bit mean with this one. So the last question is, what’s your favourite way to unwind? But because you already mentioned that you’re a dad now, you can’t just talk about being a dad. You have to give us a bit more than that as well.

[00:46:17] Rick: Right now it’s just like, whenever there’s a possibility to sleep, it’s sleeping. Yeah, no I’m a bit boring that way. I think what happened is I turned my hobby into my work and I think you guys can probably relate to that. And I haven’t really found a new hobby and now that I have a kid, I think that my hobby is sleeping.

[00:46:36] Dara: Yeah, fair enough. 

[00:46:39] Bhav: As a dad of two, Rick, I can say categorically that sleep will now remain your hobby for the rest of, at least for the next like, I’m up to eight years and it’s still a hobby. 

[00:46:49] Rick: Yeah, I can see how that would turn out that way. 

[00:46:53] Dan: Amazing. Well, thank you for the talk, Rick. Thanks for the rapid fire. Thanks for all the insight you’ve given. And we’re going to be putting links to your medium articles and LinkedIn profile and everything in there. So if anyone wants to follow Rick and read more about what he has to say, then please head over to the show notes and click those links there. Is there anything else you’d like to plug, talk about, or promote or share while you’ve got the airtime, Rick? 

[00:47:16] Rick: Come to Superweek if you can. I know it’s expensive, but it’s amazing. It’s end of January in Budapest. I’m going to be hosting the live stream at Superweek. And I think this year is going to be even better than last year. So if you can’t attend, look at the live stream. I think we’re going to be broadcasting the last week of January. First couple of days of February, and that should be awesome. I think there is a lot of really valuable analytics content over there. 

[00:47:40] Dan: It’s the one, it’s the one to go to, isn’t it? And I unfortunately can’t make it this year, but it’s already in the calendar for next year. So I’ll be watching the live stream. 

[00:47:48] Rick: Nice, nice. Good to hear it. Thanks for having me guys. 

[00:47:49] Dan: No worries Rick, look forward to part three soon.