Listen to the episode part one >>>Here
Episode hosted by: Robert Webster, Global Vice President – Strategy at CvE
Intro – 00:00:00: Welcome to the Time for a Reset podcast, where we interview senior marketers on the big issues of the day and how they are dealing with those challenges in an ever changing landscape.
Rob Webster – 00:00:15: Hello and welcome to Time for a Reset. It’s me, Rob Webster, and here is the latest technology division. And I’m delighted to invite on my first podcast as host the wonderful Stuart Colman. Stuart Coleman has had over 20 years experience in antech and marketing. Started out bringing digital marketing to the world of print at the FT, one of the pioneers of behavioral targeting, which was audience sites where we first met. And I think at the time you were also the IEB chair for behavioral targeting, is that right, Stuart?
Stuart Colman – 00:00:39: For my sins, yes, I was. Yeah. That was a fun gig, talking to too many MEPs about how the internet works.
Rob Webster – 00:00:44: And then a few other roles after that. You’re a consultant for a period, but then at a long stint as being, I think one of the real pioneers is of the data clean room world and how that applies to marketing at InfoSum. Do you want to talk with me about that, Steve?
Stuart Colman – 00:00:55: Yeah. So a great time at InfoSum. I joined almost as commercial employee number one. There were a couple before me, but when clean rooms didn’t really exist and did four and a half years there, which is fabulous, and grew the company to what it is today, established the space at a brilliant time. But four and a half years in, it was time for a change, time for doing something new in Andrew, the fast pace of consulting and getting back into that. So, yeah, fun times.
Read this episode’s insights >> Here
Rob Webster – 00:01:15: Well, thanks to it, you’re the really good person who introduced me to the potential for clean rooms in marketing, and I’m sure we’ll talk more about that as the chat goes on. But my first question, and it’s the name of the podcast, is what do you think it’s time for a reset within the world of marketing, data and tech?
Stuart Colman – 00:01:28: And that’s a great question. And we could literally spend hours talking about it, and we probably have over the years. I think one of the challenges that we have is the third party cookie made life easy for us. Whatever the question was, the answer was a third party cookie. Whether that’s planning insights, whether that’s activation and various forms, whether that’s measurement and the things that we do at the end, it made life simple. It was a bit like a Swiss Army knife for the Internet. So what needs a reset? I think the whole industry probably needs a reset, but a reset around the fact that we need to think differently about almost every single part of how we do what we do and look at what technologies provide the best solutions and the best services on those individual elements. What are the best measurement, what’s the best planning insights, what’s the best activation and reset, how we think about how they work together and how we need to build out a holistic solution across them all without just relying on the cookie to underpin it and essentially solve everything for us.
Rob Webster – 00:02:21: No, I quite agree and I think this is the whole world that needs to change. I think the world that you and I have both relent into the last X number of years. But to kind of bring that conversation forward, let’s look at the basics first, then we can kind of get onto the nuts and bolts. For me, identity is all about there’s the world of first party density, which is if you run a company’s website, they need to know what you’re doing so they can A, make their service better and be able to sell you stuff. It doesn’t work if there isn’t that first party world. And by and large, that’s staying more or less the same. It’s being tightened up, but it stayed the same. So we can park that. The third party cookie is all for companies who are not the owner of that service. And that’s mostly advertisers, but not exclusively. And that’s the world that’s changing so quickly. And it’s changing really in terms of two or three ways for advertisers. One is the world of how do you target people, how do you understand which people you want to show your adverts to. Then there’s also the world of how do you measure whether or not that advertising works. And then there’s also some nuts and bots in between about how do you interact with ads. But that’s really the fundamentals of what’s changing and it’s led to some really weird dynamics. But first of all, you agree with that? Would you build on that at all?
Stuart Colman – 00:03:21: Absolutely, I agree with that. And I joke about it being a Swiss Army knife of the internet, but it really was. It made everything accessible and easy. And for those that didn’t have either a direct relationship with consumers or more specifically, a direct ability to engage with consumers, I think the third party cookie was a great way of being a proxy for that engagement. So things like measurement you and I talked long about measurement approaches and techniques, but it really was the third party cookie that allowed that singular view, that singular opportunity to understand someone across multiple environments. And that bred so many ways that we can think about how we create value and how we measure that value and how we understand that value. Yeah, that changing is seismic.
Rob Webster – 00:03:58: But a lot of people think that this is a challenge of the future, that the old world still works and something to worry about in a year, two, three years time, and that it will sort itself out. Now, you can tell that my view isn’t that, but what do you think?
Stuart Colman – 00:04:11: Yeah, I think if anybody is genuinely still sat there thinking that the third party cookie will be around for a while and things won’t really change that much and we can just carry on as we are. I think they’ve probably had their head down and not looking around them for quite some time because the world is already changing, the world is already shifting and we’re starting to see people think about and engage with new technologies and new ways of working that are fundamentally transforming their business. Not just how do I do what I did before with a different technology, but also how do I just do this differently, how do I think differently about what I’m trying to do and how I’m trying to understand the outcomes. I would argue anybody that isn’t seeing that and isn’t engaging with that is really not aware of what’s going on around them.
Rob Webster – 00:04:52: And the example I always use to sort of bring this home to people is the world of the Apple audience. So today the third party cookie doesn’t really work. Otherwise it’s nor does the app version, the IDFA, or the app ID. And everyone who’s doing marketing online today I think is experiencing that. And I think 99% of advertisers don’t get it right. What I was asked them to do is log into your measurement system of choice, look at the amount of money you’re spending on different browsers and realize how low it is on Apple, then have a look at how visitors your site or app gets and do that same test. And for almost everyone, it’s about half or less of the ratios it should be. So let’s say 40% of your customers or Apple users. I’ll have a wager with almost all of you advertisers out there that you’re only getting 20% max of your spend away to those devices. And that’s really, really important because Apple users I’m an Android guy myself, but Apple users represent the richest half of our market. 40% higher median average salary for iPhone users. And to not be able to target to that group is for me a sign that the online mappers in the world is broken right now.
Stuart Colman – 00:05:51: Yeah, it’s an interesting one. Again, we could talk about Apple for hours about what’s their motives, what’s the driving force behind some of the ways in which they’re limiting access or changing access for the broader world. Is it for consumer privacy? Is it for their own commercial benefit? I think the jury is probably still out a little bit as to which one it is, but there’s a whole world there combo. Yeah, both. I mean they’re a commercial business so you can’t really blame them. But I find it fascinating exactly what you’re saying there about how people perceive it. And I wonder whether the old adage of I know 50% of my advertising works, I just don’t know which 50%. I wonder whether people adopt a little bit of a well, I’m kind of just guessing that Apple’s working, because I can see it working on other platforms, and I can just kind of project forward and say, hey, if I’m getting this kind of return on Android or Chrome or anything, maybe I’m getting the same or better on Safari. And this kind of I can’t do much about it, so I’m just going to accept it.
Rob Webster – 00:06:40: Except in the reality it’s worse than that, right? Because what actually happens is the algorithm that you’re using will push all of your spend towards the bit it can measure. So actually I work with companies that literally sell iPhones and they’re not running advertising to people with iPhones. It’s that dramatic. They do a bit in branding but on the direct response, they don’t because the direct response budget for things that you can’t track is zero. And so people think they’d be very smart using all these automation AI Algos and actually what it’s doing is making their marketing worse because they’re actually moving it away from their target audience. And of course it’s not just Apple either. It’s also things like the world of CTV and out of home has different some of the challenges and I think that you’re on the same page. And what I’d love people to get across is the fact that it’s a problem for now, which is only going to get worse.
Stuart Colman – 00:07:22: It absolutely is only going to get worse. And I think for me, and I may be a lone voice in this, I don’t know, but it speaks to probably two areas that people should start are thinking about quite seriously. The first one is going back to some more traditional ways of targeting and understanding the internet. So the rebirth of contextual and first party DMPs and all those kind of fun things that we used to have and then we shifted into a more a third party world. I think that perhaps we’ll start to see more value come from some more traditional ways of thinking and ways of metrics and that can apply to some of the measurement techniques that we use as well and some of the old school ways of doing that. But the other thing I think, and I hope it’s something we’re going to get into as we go through this, is the role of an alternative identity approach, whether that be based on clean rooms and the one to one matching of first party data, whether that’s around the role of some form of universal identifiers in the space, but something that exists, particularly in Europe in a privacy safe, robust format that can provide some form of continuity of identifying a user across the internet, if never going to be to the volumes that you see with third party cookies. But if it can provide a bit like a portfolio, if it’s part of the answer, I think that’s a positive step forward.
Rob Webster – 00:08:28: And to kind of repeat what you just said, I think what we’re trying to say is that there’s no one solution, right? The cookie was so easy and it did all of these things from an advertiser’s point of view at least. Really, really well. Whether it did it from an individual’s point of view is a whole different question.
Stuart Colman – 00:08:41: Yeah, absolutely.
Rob Webster – 00:08:42: And as you say, some of the solutions are around going back towards more traditional forms of marketing and measurement, as you say, contextual, but also going, when I ran these adverts, did I sell more stuff? And then on the other hand, it’s the how can we use modern technology to be able to be more effective in terms of replacing the people with a more robust and privacy friendly solution? Which still helps advertisers. Because we have to remember, don’t we, that this is also where your world is faster because you also run a publisher business for my sins.
Stuart Colman – 00:09:07: Yes, I do. Which is an interesting experience. It’s a relatively small kind of long tail publisher, but it does give me the opportunity to play with technology. It gives me an opportunity to kind of see some of the outcomes that publishers experience with regards to what browsers using, what return rates you’re getting, TCF opt in rates and all that kind of interesting, fun stuff. So it’s a good insight into kind of what’s happening underneath the bonnet of it.
Rob Webster – 00:09:27: Well, indeed. And I think if you were to say to people on the street that advertises are struggling because of this loss, I don’t think you get too many violins being played,
Stuart Colman – 00:09:34: Probably not now.
Rob Webster – 00:09:35: But will you also point out that this has a material impact on publishers and their ability to survive and thrive? Absolutely. And if you think about all the money that used to go to paying for a publishing really now is going to the big four or five mega publisher giants, your Google, your Metas, your Amazons, that’s really starving the publisher of all sizes their lifeblood. Is that what you’re seeing on that site?
Stuart Colman – 00:09:55: Without question. Mine’s a relatively small long tail publisher that I do as bit of a hobby, but it’s a good kind of barometer of the space, I would say the last 18 months or so. We’re having to work so much harder to generate good revenue. We’re having to work so much harder on understanding and optimizing the partners we’re working there seems to have lost that natural kind of progression that the publishers saw with using good technology and using good partners. Now we have to really work at that much, much harder. And revenues are down without question. And particularly you were alluding to earlier with the challenges with Apple across our Safari base, our CPMs are down quite significantly. We’re running about 55% to 60% of the CPMs that we see across other browsers which is a significant hit.
Rob Webster – 00:10:37: Yeah, given, as we said, it’s the richer half of the market. Right. So they should be higher.
Stuart Colman – 00:10:40: Absolutely. And then you throw in all the challenges around CCF the consent framework. Yeah, the consent framework and the global vendor list is ridiculously long. So that breeds kind of a little bit of suspicion with consumers, and therefore your opt in rates kind of fluctuate a bit. And the whole thing just puts more and more pressure on publishers to be able to generate a decent return on the work that they put in. And that is definitely harder. We’ve had to work much, much harder at it over the last probably 18 months, two years, I suppose.
Rob Webster – 00:11:05: There’s a number of things we should talk about here, but we’ve already talked briefs about clean rooms and about different IDs, and now we’re onto consent and the TCF. So we’ve already talked about all three of those. So why don’t we start with the TCF? And for those don’t know, the TCF is the IAB’s effort at allowing individuals the right to provide consent or not provide consent and be informed. There are the six tests, and I think TCF is ubiquitous. It’s used everywhere. Every time you see one of those pop ups that comes up, that’s almost always a TCF, they will pop up. Now, again, if you ask most consumers, they will probably think that if all GDP items achieved was do that, that probably wasn’t a net positive. Right. Yeah. Because do people feel any safer? They probably feel more annoyed.
Stuart Colman – 00:11:42: Yeah, absolutely.
Rob Webster – 00:11:43: And there’s a question as well as are they generally more informed? Because what proportion of consumers do you reckon have actually looked at the vendor list? It’s either an automatic yes or automatic no. Right. But that aside, it’s probably worth saying as well that the validity of the TCF has been called into question. Right. And I think there’s a case next year that’s going to be decided on that.
Stuart Colman – 00:11:59: Yeah, there’s been some backwards and forwards with the Belgium DPA. I think that’s resolved to a point. I think it’s yet to go up to a higher council for review, but it’s certainly been challenged the EU courts.
Rob Webster – 00:12:10: Go to, isn’t it? And that’s happening roughly. They expect the same time as Google has announced that it will, next time, at least remove the third body cookie from its setup. I always think with Google their dates ar⁵⅝g e more not before that date, at any point after.
Stuart Colman – 00:12:23: Yeah, think you’re right just on that. Just as a segue. I mean, Google, they’re a commercial business. They’ll do what works for them. They’ll get rid of the cookie when it’s right and relevant for them to get rid of the cookie. And I don’t think we have any insight as to when that is and how it will manifest.
Rob Webster – 00:12:36: No, let’s dive into that. So what would make up that choice? So I guess for me, three things, right? One is how much money would their ad products lose? And their ad products have lost money already from the situation with Apple B. How does it impact their broader products? So they wouldn’t want to have people not use Chrome, for example, and they won’t do anything that would hurt paid search. And then third, as to what size of legislation or fines could they get if they don’t do it.
Stuart Colman – 00:12:58: Yeah, I don’t know. People will see enough at Google decision making around this to speak for them in any way, shape or form. But I wonder whether the real catalyst for change at organizations like Google, but not just Google themselves, will be when the US catches up on privacy legislation. Europe is important to organizations like Google, and GDPR in its general form is probably a bit of a frustration for those kind of businesses because of what it makes them do, but it can kind of be fudged and worked around and manipulated a bit to make it work. I think that’s a different set of parameters and a different set of pressures when US really kicks in with privacy legislation. So I wonder whether we won’t really see massive change until that happens.
Rob Webster – 00:13:35: And when do you think that will be, just to give people an idea of what we’re talking?
Stuart Colman – 00:13:38: Yeah, I’d love a crystal ball and make a load of money by telling you an exact date, but I think you’re seeing the progression there already. You’re seeing of all states bring out versions of CCPA, which is the US version of or one of the US versions of GDPR.
Rob Webster – 00:13:51: CCPB California?
Stuart Colman – 00:13:52: Yeah, it’s the California one. So I think New York and Ohio and a couple of others are kicking in there. I know there’s some conversations at a federal level, these things will move at the pace they move at. I think for me, the real tells are when you start speaking to perhaps private equity firms who say, we’re really not looking at any companies at moment that don’t have a privacy story or don’t have technology that can be privacy applied because they’re starting to see that this is going to be critical.
Rob Webster – 00:14:13: Well, yeah, I think if you look across the US and globally, it’s a trickle becomes a flood. Right. So GDPR started fired the first shot, if you like, and you now see GDPR-like functionality around 50, 60 states globally. And then if you look at the US. As you say, you’ve got three or four or more states that have it in place now, a whole load more that are pending until you get to the point where you’re going. Right? Well, if this doesn’t work in California and New York and Ohio, do I not just need to apply everywhere because it’s too much effort to have different setups in every single state?
Stuart Colman – 00:14:41: Right, absolutely, 100%. And going back to what we started with talking about the technological change, that will be a big catalyst for changes when America catches up on privacy. We’re in a really positive state in Europe, I think, in terms of understanding privacy and understanding what’s going on and we’re getting there in terms of adapting our technology to meet that. We’ll go back to talk about TCF in a minute, but I think we’ll see the real acceleration of that when the US starts to really catch up.
Rob Webster – 00:15:05: And actually, I think that’s a real competitive advantage for EU businesses. Right? And you must have found that InfoSum that rolling out Cleveland is not an easy task, but actually, for the first time ever, really, in history, that Europe’s ahead on something, right?
Stuart Colman – 00:15:15: Without question, yeah, 100%. But not just on the legislation itself, but on the understanding of it. I’ve spent a lot of my time in for some talking to DPOs and lawyers and infosec security teams that have a privacy remit. There’s a lot of talent there now in those organizations that really truly understand the depths of the characteristics and the ways in which privacy needs to be developed and delivered, for brands, for tech companies, et cetera. There’s a lot of talent there, and I think it gives European businesses an advantage. I think it gives European tech businesses an advantage, and for some is one that has thrived, being UK based and having that strong privacy foundation built off a detailed understanding and a framework, privacy framework that underpins it.
Rob Webster – 00:15:52: We’re starting to see it already, but hopefully these dynamics will mean that European businesses start to have an advantage globally and can get that much of a leg up.
Stuart Colman – 00:15:59: I really hope so, because there’s some great businesses out there.
Outro – 00:16:03: We hope that you enjoyed this episode of Time for a Reset. Thanks for tuning in. We’ll be back talking to another senior marketer very soon. Make sure to leave a review and we’ll catch you next time.
Part two
Listen to part two >>>Here
Intro – 00:00:00: Welcome to the Time for a Reset podcast where we interview senior marketers on the big issues of the day and how they are dealing with those challenges in an ever changing landscape.
Rob Webster – 00:00:15: So just go back with the TCF then. So hopefully things will be sort of more or less started in the next twelve or 18 months. What areas do you think that might change for individual, for the sector?
Stuart Colman – 00:00:26: TCF is one of those weird things where it’s not really good for anybody, but it’s not really bad for everybody. Does it do exactly what consumers want it to do? Probably not. Does it give them the right level of information? Probably not. Is it the best vehicle for companies to gain a legal basis for processing data? Possibly not. Are there some question marks about how it works and who sits where in the process? Probably we can sit here and pick it apart all day. What has it done? It has given us a vehicle that we have been able to go out and ask consumers for consent or ask them to agree to legitimate interest when that’s relevant. And it’s done so in a way where that information relating to it is there for consumers to engage with if they want. But it isn’t perfect. It really isn’t perfect. I do think it needs to change and I think the two key areas for me that require change is firming up who has responsibility for what in the process. So the Belgian DPA review that they went through and IAB have been working with one of the primary things within that was are the IAB a processor or controller? It’s a bit of a technical distinction within GDPR, but it’s important as to who controls the data and what responsibilities do they have, and consumers will care about that. And I think that needs to be resolved. I don’t think it’s yet fully resolved as to where that sits.
Rob Webster – 00:01:37: Yeah, I think there’s really sort of a few things we just said there. I mean, firstly it’s progress, ever perfection, no first draft. Everything is ever perfect. Right?
Stuart Colman – 00:01:44: Absolutely. Yeah.
Rob Webster – 00:01:45: And for all the negatives of the TCF, you will say it’s better than nothing, right. And it is moving in the right direction.
Stuart Colman – 00:01:50: Yeah, absolutely.
Rob Webster – 00:01:51: And I think, as you say, just on that point of who’s the controller? The issue for the IAB is the IAB is a relatively small business in the grand scheme of things, but it’s not funded like an internet company and therefore it can’t be responsible for a multibillion dollar business and the large fines that Google or Meta would laugh off would not work so well for the IAB.
Stuart Colman – 00:02:08: I’d agree with that, but I counter slightly with but that doesn’t change the law. And if the way it works requires him to be a controller rather than the processor of data, which carries the higher degree of responsibility, that’s the law. And I think then the question becomes, rather than trying to say, well, let’s try and bend the rules so that we don’t have to apply to the same law, maybe you need to look at how is the IAB funded or how is it underwritten to manage that? The Podcast is about a reset and a rethink. Maybe that’s one of them is we have to rethink about how we actually address some of those universal uniform requirements as an industry. The second thing I’d say on TCF in terms of where I think it needs to go, is on the consumer side of it. I run a TCF on my site. I use one of the paid for services, so I get the best technology to run it. And I’m a relatively well experienced person in industry. It’s a difficult thing to set up. It’s a difficult thing to understand when you’ve got a global vendor list that’s 400 long, and the language that’s used to describe and explain processes and what people are agreeing to is very, very technical, that’s not a good experience for consumers. I wonder whether, if we were to simplify it down a bit, if we were to make it more I was going to say challenging, but the process of being on the global vendor list to be more robust, I think they could be good things for the overall process.
Rob Webster – 00:03:16: I think it’s what needs to be made much more easy for consumers to understand by actually having names on it that people recognize. So what I mean by that is that if you’re on a given newspaper or a given TV channel online, because the whole world of digital TV CTV is massively involved in this, if you’re engaging with that content, you expect that company to have some data on you. In the modern world, I think what you don’t expect is 400 names of companies, most of which the average person doesn’t recognize, also have access to that data. You think about something famous like the Amex BA card. The fact that that data might get Amex and BA two parties is expected, but they wouldn’t want other people involved in that. Right, but I think that really applies to if you see an ad for a pair of shoes on a newspaper site, you don’t mind those two companies having the data.
Stuart Colman – 00:03:57: Absolutely.
Rob Webster – 00:03:58: And that actually sort of brings me onto the world of clean rooms. Yeah. So are clean rooms a solution and how can they help resolve some of these issues?
Stuart Colman – 00:04:05: Yeah, so I think it goes back to what I was saying earlier about there being a portfolio of answers to what the future looks like. I think they are part of the solution. I mean, clean rooms, they exist to allow two or more parties to understand connectivity between their data in a privacy safe way. Essentially, if you’re a brand and I’m a publisher, we can work out that we know the same person. They’re a customer of yours and they’re a viewer of my content, we can work out that we know the same person without actually having to share that data directly with each other. As long as the Identifier that we have on that person is something that is unique to them, like an email address or a phone number or Apple or Android device ID or something of substance that identifies you as an individual. And they can play a really, really great role in creating an understanding of that connectivity. I, as a brand, want to be able to learn how many of my customers are on your website and then I can make some decisions about what I do with that. Whether that’s we can understand those consumers more and that can help influence the type of campaign that we plan and run with you. Whether it’s about activation of those people actually showing them ads, or perhaps not showing them ads, that’s another way. Whether it’s about creating a base or a seed for a modelled audience and I’m sure we can touch on that. In a bit, or whether it’s about measurement, clean rooms could have really, really important part in that. Where clean rooms struggle a little bit is the scale that they can then deliver because they’re working off absolute identity and that is a subset of their entire audience of a brand or a publisher.
Rob Webster – 00:05:26: Let’s try and also make that a bit more digestible. So I think of clearing as a bit like plug sockets. Now, bear in mind they work really, really well if you’re using that plug socket. And also I think they work really, really well in that world we spoke about where you have two partners, you have the publisher and you have the advertiser, and you only want those two people to know. And if you need to, you can also add in the data companies was a third. But for me, the challenge of clearance has been someone that’s tried to roll these out. I know you have two for your own previous company, but different markets use different clear rooms, different publishers use different clean rooms. I mean, Meta claim not to use clean rooms, but in fact they kind of do. Amazon are just rolling one out as well at the moment, but they all have their own different standard. As an expert, would you agree?
Stuart Colman – 00:05:59: Yeah, I would. Yes. Clean rooms are individually unique in their approach to how they do things. And that’s driven a lot by the technical development around how they manage privacy and how they manage an interaction or a connectivity in a privacy safe way. And there are some broad brushstrokes you can put on homomorphic encryption or differential privacy. You can throw out kind of different things. But fundamentally, every player in the space has a different approach to it. And in these early stages, and we’re still in the early stages of clean rooms interoperability between them standards across them is quite difficult to achieve because they’ve got a commercial advantage in the way that they do it.
Rob Webster – 00:06:35: So should the EU ask for a single plug socket?
Stuart Colman – 00:06:38: It’s an interesting one. I think partly, yes. And you could argue something like a Google pair as an approach could be a way of creating certainly a universal or a single universal adapter. Not necessarily a single plug socket, but maybe an adapter that plugs into multiple. So I think, yes, there is definitely commercial benefit and functional benefit to better interoperability across these platforms. I don’t think you’ll ever reach a point where they all work off the same standard and do the same thing because they lose some of their commercial benefit as a result.
Rob Webster – 00:07:10: And innovation.
Stuart Colman – 00:07:12: And innovation. Absolutely. I think if clean rooms are to start to scale and impact the industry in a way where they’re more than just a nice idea and a promise, they do need to get better at that interoperability. And I think you’ll see companies like Google and perhaps Amazon and others offer up these kind of pair based solutions, which is a way of creating interoperability. I think you’ll see more of the orchestration software, something like Ahabu play a role in perhaps bringing together more of these companies. And I think you’ll start to see some of these companies just want to work together a bit more to drive the whole thing forward.
Rob Webster – 00:07:43: Absolutely. And I think as well, you’ll have consolidation, right? Because at the moment there’s a number of these different things. Normally over time they merge into a few and that will bring that interoperability piece.
Stuart Colman – 00:07:52: I also think and I was chatting to somebody about this earlier today I also think you’ll end up, as we see in other areas of the industry, you’ll end up with a spectrum of offerings from simplistic, low level, privacy based clean rooms right through to highly technical, high end clean rooms with really strong privacy elements to them. And I think two things you’ll see, one is that use cases will fit into that spectrum at some point, and the decision you make as a brand or a media owner will be based on your needs. So some people might need an InfoSum style high end privacy solution. Some people might need a solution that is less about privacy and ticks the GDPR box, but not necessarily pushing the boundaries. So I think you’ll see that. I also think you’ll see companies start to work with multiple providers and I think that’s just a given that a brand or a median will end up probably working with two, three or maybe four providers in that space. That will naturally require a shift, perhaps in the economics that underpin the space, you can’t pay for everybody at a premium rate. But I see a world where I, as a media owner, might work with three different clean room providers based on three different solutions or three different requirements.
Rob Webster – 00:08:52: And to sort of set the dream on this. As to why, I think you obviously also myself have been very keen on this world and I think the reason is because they’re clearance your data is safer. Whether or not it’s perfectly safe is a whole different question we’ll we might come on to in a second. It does allow for the advertiser to then be able to target auditors and be able to measure. So you solve that problem. Problems we talked about, I think, have been about fragmentation and that’s led to, I think, a challenge that it’s really been a solution for the giants so far. Right? So if you’ve been a giant advertiser working with one of the big TV channels, for example, works really well. But what we’ve described hopefully should mean it can open up to the longer tail. So there is a danger, isn’t there, with these sorts of systems because they’re relatively expensive and relatively difficult to use. You yourself to run a small publisher, I guess you don’t use a clean room today, do you?
Stuart Colman – 00:09:29: I don’t know. I think two things to consider there. One is clean rooms generally are going to be a really good springboard for all the other things that we can do as an industry. A clean room based solution for a brand and a publisher, as a foundation for understanding connectivity, understanding profiles of audiences that you share, be able to derive insights from that, that then feeds into how you plan across that media. A whole bunch of different activation underpinnings, but forget activation for a minute, then be able to measure that in some form at the end. It’s a really good springboard into other solutions. And we can talk about retail media measurement within retail media that can springboards from clean rooms. We can definitely talk about insights and planning with things like channel Four brand Match IQ, which is about TV planning derived from connected audiences, a whole bunch of things you can do and that’s great. But to your point about the longer tail, the access point is quite difficult. It’s quite a challenge. And that challenge isn’t just money. Clean rooms aren’t cheap right now, but that can change and business models can change. The entry point for a clean room with regards to things like privacy and infosec reviews is still quite high and there’s no real getting around that at the moment. If I’m going to give my data to a clean room and allow for collaboration with somebody else, I need to do a DPIA. I need to have a DPO look at that and be comfortable with the process. I need to have the right warranties, liabilities, controls in place to underpin it. I need to be confident from a technical point of view that it works. And I don’t think today there is a solution in market that makes that as mainstream and as beyond the big players as is probably going to be required long term.
Rob Webster – 00:10:57: But probably that will emerge as because what we’re doing really is rebuilding the way data works and building those pipes takes time, right?
Stuart Colman – 00:11:04: I actually do think it would emerge. I think it’s still going to be challenging, however creative and actable you are, to get past the fact that I can’t just give you my data. I have to check the things that I’m going to do. So small businesses are going to have to have a degree of is this the right thing to do?
Rob Webster – 00:11:18: Well, actually that’s there by design, right? We don’t want people to be as free with data as perhaps they once were some time ago. Those barriers are there. So if this data is as, it is really valuable and very important. You don’t want it to be that kind of fruitful. But it does have a question, doesn’t it? I think one of the implications of GDPR and all this regulation has been it’s favoured the giants. Now, we’ve just described how like a clean room can work for TV companies and big appetisers. What it’s really done is favoured the mega giants, your Google’s, your apples, your Metas, your Amazons. And that’s been an unintended consequence. How do you think? Because obviously I don’t think the EU is particularly big fans of those giants. If you said to them that what this will do is help them dominate even more, that wasn’t what they intended, is there anything that you talk about there that might sort of help move that forwards?
Stuart Colman – 00:11:59: It’s a really difficult one because it definitely wasn’t the intention of GDPR to favour Facebook and Google. The work I did at the IAB with the Oba Council was actually at the time predated GDPR. It was to do with the Cookie directive. And I spent time in Brussels talking to MEPs that were making this law. And with the greatest respect to them all, they do a fantastic job in many ways. They have a fairly limited understanding of the Internet. And I spent a lot of time just explaining how the basic internet works. So it’s not unsurprising that we end up a point where a law intended to curtail perhaps some of the power of the Big Four, big Five has end up actually giving it to them. As for how we fix that, it’s a really challenging one because I don’t think there is a simple answer. And I think part of the answer isn’t about how do we simplify the process for consumers or how do we simplify the process for longer tail. It’s about how do we rethink the way that we do things. And we may need to say things like I’m not saying this is necessary, true, but maybe clean rooms aren’t for the longer tail. Maybe we just can’t square that circle in terms of the requirements on the GDPR for safe data connectivity and transfer in a world where those organisations don’t have the resources to make that available.
Rob Webster – 00:13:02: That makes a lot of sense if we say that claim rooms are a great solution for the companies that hold a lot of data, right, be it publishers, be it advertisers, a solution that might work for the mid to long tail might be the universal ID. And I know that that’s kind of related to the world of clean rooms, but different. What do you make of that space and how is that going?
Stuart Colman – 00:13:18: I think it’s a really important space. First thing I’ll say off the bat is I don’t think identity or universal IDs and clean rooms are mutually exclusive. I think they can actually work really closely together. I think there’s some connectivity that can be created there. Where universal IDs can be really successful is they do have a lower bar of entry for participants, whether that’s brands or media owners or other participants, data companies, et cetera. It’s got a lower bar of entry. You don’t have to do quite so much to be comfortable to work with them. And they do offer a greater degree of scale. They are able to offer a broader access to a broader internet, cross publisher, cross channel, et cetera. So they’re a good solution for that. The trade off is you don’t have so much of the one to one connectivity that you get from a clean room. incredibly exciting
Rob Webster – 00:14:35: And can it also be good for the individual?
Stuart Colman – 00:14:37: Yes. Let’s break that down a bit. I think there’s probably two elements that we should consider. One is, is it good for consumers that their internet experience or their digital experience can be adapted to their interests? Is that a good thing? Yes. Years ago there was an IAB study done that said that 74% of people didn’t want to be tracked online, but 84% of people wanted advertising more relevant to them. Now they’re made up numbers because I can’t remember the actual numbers. But the point remains is that people want an Internet that actually makes sense to them, an Internet that actually provides value to them and is tailored to their needs. So would a universal ID approach on top of a clean room solution that does the same thing give consumers something they want? I. E. a personalised Internet. Absolutely. That’s a good thing. There’s still the question around how you get the consent or legal basis for that. And is it transparent and obvious enough to consumers? I don’t necessarily think that’s for the ID solutions providers to solve themselves, that’s for them to work with the industry to improve the TCF, to make it clearer and more obvious to consumers what is happening.
Rob Webster – 00:15:40: And perhaps to provide options. Right. I mean, I look at the following. So the new solutions, they’re all different, but they do at least have,”more privacy” by design built in than the cookie did. The cookie was basically a free for all that almost anyone could perhaps not quite as dangerous as some think it was because all it was was the idea that this browser saw the user. But anyway, these new ones are encrypted, which provides some extra safety if for the most part. And the dream is that users can then choose to opt in to these systems. And I think opt in is going to be very important, hopefully enough scale, and as you say, the adverts that they get will be better tailored towards them and also, as a result, publishers will earn more money and everyone’s happier. Does that also then leave the world, though, for people who go, you can tell me it’s safe all you like if I just don’t want it, I want to opt out. Do you think that’s going to be a big part of the future?
Stuart Colman – 00:16:19: Absolutely. And it should be. This is about consumer choice, ultimately, and GDPR is bounded on this principle, but I believe it’s a principle we should all adopt, which is consumers should have a choice. Absolutely. There shouldn’t be any world where it’s forced upon them. I think, however, we need to do a much, much better job at informing and explain to consumers why it matters. You talk about advertising, but universal IDs and the ability to understand a consumer journey across the Internet doesn’t just impact on ads. It impacts on the experience you have on a website. It impacts on how products are delivered to you in terms of how you see them, how you understand them. It impacts content that you’re engaged with. It’s a broad spectrum of value that’s created for consumers.
Rob Webster – 00:17:01: And ultimately the free web right. Which relies upon it.
Stuart Colman – 00:17:02: Literally just going to say that should consumers have free choice? Yes, of course they should. We should not impose on anybody anything that they don’t want to do. But it has to come with the understanding that the Internet isn’t free. It’s funded by the revenue generated by sites to power the services that they offer. And that comes from an ability to very bluntly monetize the eyeballs that come to them. Anything you do to limit that monetization opportunity reduces the value those sites can create and therefore the services they can offer that journey. That understanding really needs to be driven home to people. I saw recently somebody, I think it’s Twitter as a rumour that they were going to start offering micropayments for articles, which is a really interesting idea. I have no particular leaning one way or the other as to whether it. Will work, but what it would do. Is it would drive home to consumers that the Internet’s not free.
Rob Webster – 00:17:48: Sure. And it also share that value exchange. Right. I think people probably get annoyed when they think there’s all this money being made off of their data, which they’re not getting a part of. And if you kind of, as you say, educate, but also then share the benefit. Right. The value exchange is so important that people understand what it is. The adage that if you’re not paying for it, then you are the product. I think that is true, but if that’s transparent, then that’s okay, I would have thought.
Stuart Colman – 00:18:08: Absolutely. And I know there’s been I’m sure we’ll talk about some emerging technologies soon. I know there’s been a few things, a few companies launched over the last few years who tried to put the power back in a consumer’s hands and say, hey, monetize your own data. I think there was a browser that was on Dragons Den a while ago that did it. And I know there’s been two or three other companies launched. The fundamental challenge there, I believe, with those kind of approaches, I think superficially is a nice idea, but fundamentally, consumers don’t really care. We’re quite lazy as consumers. We’re quite passive rather than lazy, actually, in terms of expressing our interest in stuff. And the only time we would actively express an interest in stuff that could create value would be when we think there’s something in it.
Rob Webster – 00:18:43: For us, almost is a value point from an advertiser’s point of view. Everyone thinks advertisers want to target them, but actually they don’t. They want to reach millions of people. Right. So that lots of individuals might then go and buy their products. But actually, I think the challenge for all of these new systems I really like the idea, but unless you have enough scale there or enough scale within a certain audience, certain niche to make it worthwhile running against it ultimately, every time we have one of these new ecosystems, if you like, you need to go and pay somebody to work out how to advertise on it. And you’re not going to do that for five man than three dogs, right?
Stuart Colman – 00:19:11: No, absolutely. But that speaks to what I said right at the start about this kind of new portfolio approach. And one of the big changes, the reset changes we need, is this ability to work across different platforms and stitch stuff together. You’re absolutely right. But I think there’s some really exciting opportunities, whether it’s data connectivity driven by clean rooms or whether it’s the role of second and third party data providers working with universal ID solutions. I think there’s huge opportunity for looking at new ways to create scale and create opportunity. There things like modelled audiences, the role of AI in creating and understanding consumer journeys. Digital Twins. There’s a new company called Rayn, Rayn who are launching or trying something called Digital Twins of Consumers, which is all about creating models of consumers and then mapping through journeys to understand what could happen. There’s ways in which we can use these foundations of data that we’re able to create through universal IDs and clean rooms to create other opportunities like campaign modelling, like lookalike audiences, etcetera, to create the scale and the value that we want without necessarily the direct marketing.
Rob Webster – 00:20:13: Well, that’s it. I think we’ve had an amazing run through of identity and the solutions and problems around that the last half an hour or so. Let’s kind of line up a little bit. If I talk about the world of AI and some of the new things that are emerging, I mean, has it changed? Is it the same thing again? What’s the take of the world of AI in the last few months and the hype and push back on AI?
Stuart Colman – 00:20:28: I think AI is incredibly exciting and incredibly scary, I think is probably the way I put it. Our interest is AI in advertising and marketing, and I think it can have a really significant impact on that space. Clearly, AI is much, much broader than that. And you look at some of the things we’ve spoken about today, around model audiences, around campaign modelling, around the use of digital twins, around the use of smaller subsets of data to create seeds and build from there. AI could play a really interesting role there. I think there’s some massive questions around ethics. There’s some massive questions around ownership. There’s some massive questions around privacy. I’m sat here with my computer next to me with a Cyberdyne system sticker stuck on it. Has anybody asked ChatGPT to build a better version of itself yet?
Rob Webster – 00:21:11: Before we get onto the doomsday scenario, let’s have a look at some of the positives. One thing say is that I think that the gene is at the bottle, right? I don’t think we can stop. If it’s going to happen, it’s kind of inevitable. What I would really encourage folks to do is to really kind of lean into some of the positives and let’s get your take on this. One thing that I’ve noticed for a long time, it has been changing the way that we work and the way that the world works. But for a long time, it’s been locked away that only a small group of very counted developers can access. We have less than 30% population. The wonderful thing about the recent AI thing is it’s been much, much more democratic in terms of how anybody could use it. You got your thoughts on that.
Stuart Colman – 00:21:41: You took the words out of my mouth. It’s a democratisation of access to abilities, whether they be technical abilities or data science abilities or creative abilities, is significant. In the early days when ChatGPT first started coming out, I read a story about somebody that spent a weekend getting Chat GPT to write a children’s story for him, then got Chat GPT to or a version of chat GPT to create imagery around it, and they stuck it as a book on Amazon and made $20,000 in the weekend. That democratisation of creativity can apply to website development and data science, etcetera. That’s a huge opportunity. I think the way to harness that opportunity and it’s a typical Hype curve thing that we see with many new technologies, is to very quickly get to a point where it isn’t just about, hey, this is cool and fun and interesting and new, and this is the value it creates and really focus in on the outcomes and the reasons for it. Why would you use AI? What’s the benefit of using AI in this scenario versus not using AI? How does it create a greater opportunity and greater value for you? And that’s where we’ll start to see it really gain traction when you can prove those points.
Rob Webster – 00:22:43: And I think that example you gave about the person who used it to create a book, so we’ll go, well, he didn’t write the book. Is it his? And there’s some merit to that. But that individual did tell the AI, I’m guessing, what chapters he wanted, what sort of characters. It wasn’t just a case of, ChatGPT wrote me a book, please. He will have written many, many thousands of lines of questions to make the AI produce it. Then I presume he would have proofread it.
Stuart Colman – 00:23:04: Absolutely all of that. And I think it’s a really interesting dynamic. My stepson’s at university studying computer science, he loves ChatGPT because it makes his life easy. But we’ve talked a lot about in that particular genre, the world will evolve from physically writing code that can be done very, very effectively by an AI. Two aspects. One, it’s about once it’s written, who does the auditing, who does the process of checking to make sure it’s right and relevant. But more importantly is the input is the curation of the questioning of AI. That’s, I think, where the new skills will come through. And that applies to every genre, whether it’s marketing, it creativity, how you ask the questions generates the value that you’re going to get. And that’s a new discipline.
Rob Webster – 00:23:42: It’s the application, isn’t it? It’s the fact that whatever skill sets that any of us have, we can use ChatGPT to use that to a greater degree because we can automate a lot of the tasks.
Stuart Colman – 00:23:50: Yes, absolutely. And you apply that to putting together two or three strings of what we’ve been saying. If I’m in a world, an advertising world, where I’ve got a good understanding of connectivity of data based on a clean room. And I’m able to model that or convert that through perhaps a digital twin into a much larger data set that is a good representation of my audience, and then feed that into an AI model that can then predict 1000 different outcomes from a campaign requirement or some other requirement that could be hugely valuable and hugely interesting in terms of what it can do to our industry. It is about the inputs, it is about the control of the data and the underpinnings of privacy. But yeah, let AI loose on campaign planning and even measurement. Can I make some good predictions about outcomes from a measurement point of view where I don’t necessarily have to measure it at the end? It could be a really interesting evolution.
Rob Webster – 00:24:38: If you think about the world of also sort of ad creation right at the moment. If you’re a graphic designer, it takes you a long time and you need to produce all these different ads for all these different programmatic worlds, that there are methods, different ads to Amazon needs. But if you’re a graphic designer, you can build one or two test examples to program system and go right, make me more like this, and you maintain and then obviously the designer in question. So that’s right of reply of which ones are good enough, which ones aren’t. And in that way they can use AI to produce much, much more content than ever they could before. So they’re turbocharging their own skill set while still being the architect, if you like. The user, if you like, is the conductor of the orchestra, but he has the whole orchestra to amplify what they’re doing.
Stuart Colman – 00:25:11: If I was to play that back to you, it’s essentially saying the core human skills are still a requirement. You still need a graphic designer, you still need a programmer, you still need a creative person, you still need a data engineer or data scientist. Their skills would evolve into more around orchestration curation, prompting managing outcomes rather than the physical work of actually the creation itself. AI can step in and do that kind of heavy lifting on the physical work and that both enhances the outcomes, but it also perhaps broadens the spectrum of what people can then do and it opens up a bunch of opportunities.
Rob Webster – 00:25:42: The programmer might not write all the code anymore, but they will design the program, they will tell it what to build.
Stuart Colman – 00:25:47: Exactly. And I think a lot of people are scared of AI because they’re unsure of what that means to them. But I don’t think anything to be scared of other than CyberLine systems and Terminator and all that fun stuff. But I think it’s a new discipline that people need to understand and we need to get quickly to a point of value for that understanding to be able to really manifest and develop.
Rob Webster – 00:26:03: And I think those that lean in will enrich their working lives and their lives generally, right? I don’t think you have to lean in, but I think those that do will get a lot of benefit from it. In terms of the negatives there, what should we be worried for?
Stuart Colman – 00:26:13: I think there’s unanswered questions around IP who owns what, and they’re to be resolved. I think there are some unanswered questions around privacy. I know there’s been some big headlines in Europe with ChatGPT being banned in Italy for a short period of time, etcetera. I’m not necessarily sure the privacy issues are as deep as perhaps the press are reporting, but I think there are some questions to answer there. For me, I think the two biggest concerns from an AI point of view are ethics. Who controls the ethics around what can and can’t be done, and is that done at a government level, is that done at a corporate level, etcetera, and how that impacts on what can and can’t be achieved.
Rob Webster – 00:26:48: We’ve seen how governments aren’t great at legislative for the new stuff, right?
Stuart Colman – 00:26:52: Exactly. And that can manifest in a bunch of different ways around what can physically be done through to how you manage things that are done that perhaps wrong. Some of the deep fake stuff that’s done, some of the realistic photos that are created, how do we start to see the real from the imaginary in world where AI is so good at creating that? So I think there’s some fundamental questions to answer there. The thing that would keep me awake at night it doesn’t keep me awake at night because I sleep quite well. But the thing that would keep me awake at night is bad actors is how do you even in the world if you get the ethics right and you get the foundations of it right, you get privacy right, et cetera, you’re not going to stop bad actors. And when we sit here and say, think about how amazing AI could be in supercharging creative industries, or data science or all these other disciplines that can really use AI to its strength and really supercharge their efforts, you could apply that same logic to bad actors and how AI could supercharge that as well. So I think there’s still some fundamental things we have to consider and work with. I just don’t know how necessary we’re going to answer half of them.
Rob Webster – 00:27:45: Yeah, there’s no easy solutions to those challenges, right? I think some of them are around transport. Education are the best two ways I could think of solving it, education, because if you’re aware of what could go wrong, you can kind of mitigate it. One example is that I think people are relying on these systems as being correct. That’s relevant to the fact that people used to log on to Wikipedia and know that it wasn’t always right. If you only use one source, it can be biassed. And guess what? If Wikipedia has X million articles, some proportion of them will be incorrect and some could be malicious. But if you’re aware that’s possible, you can kind of move forward the same as ChatGPT or any of these generative AI programs. So I think part of it’s about education, part of it’s around transparency, about knowing what’s actually happened. The problem with AI is it does things that the programmers don’t always expect and you can’t always get that transparency, it’s the way it works, is that you don’t always know why it did something which can make things more challenging. So I guess it’s just down to education again.
Stuart Colman – 00:28:33: It is. But we face this before. I think AI is particularly spiky with these things because its power is significant. It’s got access to an endless supply of information and knowledge and understanding that we’ve never seen before. The ability to essentially read the entire internet. It’s crazy, but we’ve been here before with new technologies. You and I are just about old enough to remember life pre Internet. The thought of having everything available on a machine in your house was just scary at a time. And people, I’m sure at a time were doomsayers. And this is going to be the end of the world as we know it, etcetera, but it’s been, in the main, a significant improvement in the world as we see it. So I still think AI ultimately will be good thing. I think it’s really exciting right now. I think we’ve got a really big journey to go on in terms of understanding its impact and how we live with that impact.
Rob Webster – 00:29:17: I agree. I think it’s a really exciting time and it’s fast open to be around when these things are changing. Which brings me on to my final question for you today, Stuart. It’s been an amazing run through, but my final question, what advice would you give marketers or the sort of younger professionals that are starting out in the industry or early stages?
Stuart Colman – 00:29:29: Yeah, and I probably tie it back to what we said at the start, about time for a reset. Is the world around us is going to change quite significantly over the next few years. It’s going to be much more of a portfolio answer to things, to the way that we work and it’s going to be a bunch of technologies pinned together. Something that stood me in really good stead throughout my career and I would definitely encourage anybody early on in their career to do, is to really understand the things that you’re using, get into the technology, get into the processes that underpin it. When I was at the FTI I worked in sales, but I used to spend time sat with the Ops team to learn what they did. Not because I was going to sit there and traffic campaigns, but if I had an understanding of how it worked and the impacts of things I did on it. It made me do what I did better. So I would encourage everybody to really get under the skin as far as they can of how the technology works, what it does, how it operates, what are the processes that underpin both the technology and the companies that are using the technology. It will help you, I believe, significantly then, in what you then do with it and help the industry move forward.
Rob Webster – 00:30:25: Yeah. Lead into the future. Right. Lead into change. Absolutely.
Stuart Colman – 00:30:27: Yeah. Don’t be scared of it. You know, we’re always all learning and I’ll be the first to hold my hands up and say, I don’t know how that works, but I’m going to find out. I think that inquisitive nature, I think, particularly in our industry, which is quite fast moving when it comes to technological change and has lots of actors in it, so there’s lots of things to learn. Having that desire to know more is a good thing.
Rob Webster – 00:30:45: Yeah, I think if you do lean in, there’s somebody fuck out there you can talk to about it. Right. I think we’ve been doing this here for a while. Luckily for us, is that most of these new things that are happening, we know folks who are somewhat experts in them, I guess, to a degree.
Stuart Colman – 00:30:55: Absolutely. And you’ll always find people will be generous with their time to ask. And I know people like the IAB and ISBAR and others do a lot of educational courses around this. I know experts like you support them on that process. So there’s content out there and there’s people out there that can help you on this. Just be inquisitive. That’s the key thing.
Rob Webster – 00:31:11: And with that and I just wanted to say, Stuart, thank you so much for your time. It’s been an absolute masterclass of content. So thank you for this. To everyone who’s got this far with it, well done. Thank you for listening. I will reverse my earlier joke. If you do actually have any questions, I would love to hear them and we’ll try and answer them. My future podcast. Stuart, thank you so much. And yeah, chat again soon.
Stuart Colman – 00:31:27: Thank you. Rob, great speaking to you.
Outro – 00:31:31: We hope that you enjoyed this episode of Time for a Reset. Thanks for tuning in. We’ll be back talking to another senior marketer very soon. Make sure to leave a review and we’ll catch you next time.
Time For A reset Podcast. The marketing Podcast Insights from global leaders
