Podcast Wait, We Don’t Eat These Cookies? Season 1, Episode 2

Isha Chhatwal 

What I usually hear from our customers, right from [the] beginning – we started discovery, say last year to now – is a lot of unknowns, which is mostly happening because of this privacy revolution. And then having, say, that impact of [the] frame that the open web is on the crossroads of a change, because of certain reasons that we’ll of course get into. Mostly because [of] the impact that not just mobile or CTV is going through, but also desktop. But then also talking about how the scrutiny is increasing and can they continue to do business with walled gardens as against in open, etc., etc. And then also thinking about that they enjoyed the free fast ways, where there was no limitations, as against what we have now, in terms of these speed limits, which are put by privacy laws either be GDPR, or CCPA, or any upcoming ones in Virginia, Brazil, Canada, and also alternatively talking about what browsers are trying to do, which is that you just can’t go beyond that particular speed at all. So I think it’s a combination of so many things that make them nervous about being able to do their business in the future the way they’ve been allowed to do so far in the programmatic journey.

Somer Simpson 

Is there a general sense of fear or confusion? Are there groups of people who are like, “Oh, that’s cool – the industry will solve it; we’ll be all good?” What’s the general feeling?

Isha Chhatwal 

Since the point we started talking to the customer last year, as I said, it was like ‘wait and see’ approach – waiting on that gold standard solution. That just never happened because of more scrutiny of certain walled garden solutions. And then now, even getting a breather for a year or two, is now the urgency to really find out the solution that fits their needs. Because we know there’s not going to be one solution that’s going to perfectly work for them in the future, as it did today.

Somer Simpson 

Peter, you’re out there talking to other CEOs, CTOs, industry leaders at different companies and agencies. What’s your view on how everybody’s thinking about things?

Dr. Peter Day 

I think it’s a very mixed bag. I think there’s been a lot of, I say, it has felt like a bit like a Gartner hype cycle, the last couple of years, but most people I speak to now are very much in that ‘test and learn’ mode. I don’t think anyone believes that there’s going to be a single drop-in solution that makes everything work as it works today. And so we’re increasingly finding that people, in the conversations I’m having, are just desperate for information and data at the moment, so they can understand what’s really going on, what’s really going to work, what’s not going to work.

Somer Simpson 

Cool. So let’s talk about how we got here, right? How did we get to the place that Google said, “Chrome is going to get rid of third-party cookies by X date?” And then, of course, they changed it to Y date. But what’s led us to that announcement?

Dr. Peter Day 

I think, in this whole conversation, we’ve got to recognize that this usage of cookies was a complete hack, based on how HTTP protocols work– this small hack that allows you two things. One was fetch images from someone else’s web service; I didn’t have to host all my content myself, and then set these cookies so that I could remember who somebody was, like a better user experience on my site. And I didn’t have to send all the information back and forth every time. We’re there to enable the internet as we know it today to function. And then, we in adtech worked out how to use that mechanism to do things like measure and see how many people who saw an ad then went ahead and bought a product, and to maximize the value of an impression by saying, “Hey, look, someone who was on my site, looking at something they forgot to check out, I can remind them, right?” Cookies were not invented for anything related to advertising, measurement, activation, and so on, and so forth. So there was this kind of hack and it worked really, really well for solving a few different challenges. And cookies have been used in adtech for three very distinct purposes. The first one is measuring. Measuring, ideally at an impression level, is my advertising working? So there’s a measurement thing that cookies have been there to solve. And then, off the back of that, we built these insane, multi-touch attribution systems, and so on and so forth. But it was all based on cookies. The second thing that the cookies have enabled is, as I mentioned a second ago, maximizing the value of an impression, making sure it’s a little bit more relevant, therefore a little bit more valuable. So when I measure it, I can see that it’s worth more. And so I can get this advertising campaign, which is essentially just a portfolio of impressions to have the characteristics that I want it to have, and having this data enables us to do that. And the third thing cookies have allowed us to do, and these horrible kind of stapling processes and match tables, has been to enable this kind of adtech lumiscape, these various pieces of technology to talk to each other. It’s become the kind of glue that’s held together, things like third-party data, and DMPs and DSPs and exchanges. The kind of identity matches third-party cookies have enabled have made that ecosystem work. I think that there’s going to be no drop-in replacement that can solve for all three of those. And I think it’s important that we separate out those challenges and say: how are we going to measure if our advertising is working or not? And what can make sense there? And without deterministic data, some of this stuff is unknowable. So we need to use things like statistics. And then how can we maximize the value of attention based on information that is available such as contextual information or geographic information? And then how do we glue systems together? I don’t necessarily think that’s a problem worth solving. Others would disagree with me, but certainly some people are stepping into that space. But I think we got here because everything was based on a hack that was not really there designed to support adtech, and we built solutions to specific problems all backed by this hack, and people became uncomfortable with that, because there weren’t necessarily the right controls in place. And increasingly, the tech giants want more and more control over everything. So they’re trying to pull things into their walled gardens. And this is another mechanism by which they can do that. So I think there’s a bunch there.

Somer Simpson 

I liked the idea that the web today is built on a hack. It’s like the web is a startup. And we’re all living with the spaghetti code. And maybe now it’s time for like web 2.O, although I think that’s already been used. So we’re probably on like web 4.5 by now.

Dr. Peter Day 

I think that’s all NFTs or something. I think it’s always worth remembering that what we rely on today on the internet  was basic infrastructure – these protocols like basic infrastructure – that everything else has been built on. And it’s been phenomenal, the foresight of some of the innovators early on to put some of these things in place. And because these foundations are consistently applied everywhere, and most browsers interpret the things in the right kind of ways, we’ve been able to build entirely new economies, which is phenomenal. Obviously, changing that underlying infrastructure is really hard. It’d be like telling people where I’m from to drive on the other side of the road, all of a sudden; it’s really, really difficult. And I think this is what we’re seeing right now; this is why it’s a little bit longer than people are anticipating because this is basic infrastructure, which people have innovated on top of in ways that this event has never even dreamed of. Now we’re trying to redo some of the underlying infrastructure, and that’s going to be hard.

Somer Simpson 

Isha, do you have anything to add to that?

Isha Chhatwal 

Sure. I’ll just try to put it in a different way in that I’m not old enough in this industry, or overall, to know how programmatic has kind of evolved over time. But it seems that third-party cookies specifically has been like one dominant currency for a long, long time, which was earlier leverage so that you can possibly offer a certain service or offering to the right audience. But then I think it also became industry obsession with personalization, and kind of disregarding common sense of how online advertising has to be. Of course, overall as an advertising industry, I think we need to own that we have not done a good job of communicating the value exchange of the open web, which is that you get to read free content in exchange of seeing these advertisements. That’s one reason why consumers get really confused about seeing those consent pop ups and what this means and how their data is being leveraged for good or bad or worse. And that’s why I think there is a need to set rules in bringing in more transparency, choice, and control in the hands of our consumers. So even if we see third-party cookies going away, which has always been the case for some of the other browser companies, but specifically now with Chrome with 70% of the accessible points, it’s just that we can get back to fixing a bigger problem now, which has always been a problem for our consumer. And getting back to our first principles, bringing a lot of creativity to the things right away.

Somer Simpson 

Yeah. In a previous episode, I was talking to Scott Messer from Leaf Group, and we talked about how for the longest time we’ve thought that this ecosystem that we’re in is a two-sided ecosystem: you’ve got buyers, and you’ve got sellers. And it’s really not: it’s a three-sided ecosystem, because you’ve got the consumers, right, as a part of that barter – which brings us back to what you were talking about earlier with consumer data, privacy laws, and GDPR and stuff like that. What is it – GDPR, and before that ePrivacy Directive – that has led us to the place that we are now, where adtech is having an identity crisis?

Dr. Peter Day 

I think you make a really good point, when we talk about this three-sided ecosystem. You’ve probably been bored to tears hearing me talk about this a lot. But I really think that fundamentally, we’re talking about an economic system. And you’ve got these three different actors there. You’ve got the end users who want frictionless access to content. They want to see less ads; they want fewer pop ups; they want fewer paywalls and registration rules, right? They just want frictionless access to high quality content. You’ve got content producers who want to serve that need of producing this content, but they want sustainable businesses. They need to be able to fund future content production and build valuable businesses. And you’ve got advertisers who are looking to reach and influence people eventually to buy stuff. The nice thing about this economic system is all of those actors are extracting something slightly different from this economic system. They’re not in direct competition, therefore prosperity can emerge. But it very much is an economic system. All technology, adtech in this space – this ad-funded economic system – all adtech is there to make that economic system work more effectively, more efficiently for the participants, that’s our only job. And so we can add value there by, say, taking a random bunch of impressions and making them a little bit more relevant – so less annoying to consumers, more valuable to advertisers – and value kind of comes out of that. We can help publishers monetize more effectively by taking their portfolio impressions and connecting it with the right brands, for example, so they can get a higher return on that. But the job of all technology here has to be framed in the economic system and making it more effective, more efficient for each of those three participants. And if it’s not doing that, it’s just a distraction. And it’s not very good technology. So that has to drive a lot of our behavior and a lot of our decisions – us as an ad tech company, every other ad tech company out there – is just to recognize what our responsibility is, and therefore what we should prioritize, what we should build, and what we should not build.

Somer Simpson 

What is it about this healthy economic model that we have that governments feel like they need to step up and tell players in that ecosystem how to protect consumer data?

Dr. Peter Day 

Well, I think an economic model by itself, every economic system, needs some sort of constraints, right? This is why we have rules around creating monopolies and things in economic markets. It’s why financial markets are regulated because you need rules. Because people will optimize. All of these markets are about how you optimize your use of resources, and completely unconstrained that leads to some really bad scenarios for the thing that it’s trying to  support. I think simple, clear rules are required by any economic system; I think a lot of people still don’t really understand how ads get shown to them or wired to get to them – and where there’s a lack of understanding is often missing information there. The thing that surprises people when you actually walk them through it is: cookies aren’t nearly as long lived as people think they are. They’re not nearly the kind of panacea that the adtech world thinks they are today. They don’t hang around very long; they’re not very sticky; we can only see about half of internet traffic with any cookies on them today, right? So they’re not that powerful. But when I say where there’s a lack of understanding, because we haven’t done a very good job at making it easy for people to understand, there’s a little bit of fear. And I think government regulators around the world have got really nervous on consumers’ behalf, like “what the hell’s happening here with all of my data,” and when you speak to people directly, as we’ve done – we’ve had panels and so on and so forth – we’ve got people from all walks of life to come and explain how they think it works. If it weren’t that way, it is really scary. You see it all the time, is Amazon listening to everything I do, because I suddenly saw a certain brand of washing liquid or something in my top lists and stuff. I’ve gone off topic again – sorry, Somer.

Somer Simpson 

No, that’s all good. We don’t have to stay on topic. We can keep chatting. Yeah, it’s just a chat.

Isha Chhatwal 

There should be red wine. So maybe that’s missing.

Somer Simpson 

I know, right? I know, I know. But then I would just start talking about natural wine. Anytime I read an article in one of the trade magazines, the majority of the time when people talk about cookies going away or identity, they’re talking about the demand side, right? How are we going to target the audiences we want to target? How are we going to do attribution? How are we going to measure things? How are we going to do retargeting, right? And there’s very little talked about the supply side. I know, Isha, you’re in there talking to our publisher partners, like all the time. What is the view from the supply, from the publisher perspective, on cookies going away?

Isha Chhatwal 

Just one thing: will they lose out on revenue? What will happen to their revenue? That’s it. Full stop. And, again, our solutions show that impact on the revenues, if not now, given CPMs are way too low, but in the future, we’ll start looking into alternatives. From the time we always knew that first-party data is like gold and they just want to continue to keep it to themselves and not share it, which largely impacts a number of identity solutions given how they are built, because we know that it is not one solution fits all – about how much deterministic data we can possibly have. But largely, if we don’t solve a problem for publishers, which is to have a sustainable revenue, that’s going to be a bigger problem.

Dr. Peter Day 

Sorry, I’m taking over interviewing now, but Isha, are you hearing from particularly publishers, on their perspective on some of these tech giants? A lot of these publishers, to me, kind of played the game with Facebook for a while and worked out they couldn’t make any money. They played the game with Google and worked out, “Oh, well, wait a minute. We’re making less money, and someone else is taking a larger cut.” And they tried to play the game with Apple, where Apple just wants them to put everything in environments they own and operate like Apple News, and they don’t seem to be monetizing well there, either. So what are you hearing from publishers in terms of their perspective on the tech giants?

Isha Chhatwal 

I think I’ll just repeat myself, in a way – that they want to see the benefit of the solutions that DSPs have to offer in a new world without cookies. And if they don’t get the same kind of returns, the way they’ve been having either, now, do they want to continue to go to the next level of adding more walls or subscriptions? Or finding out the alternatives and try understanding their own first-party data themselves to know their audiences? And then plan their audiences better than other DSPs, or other solutions, can.

Somer Simpson 

They’re looking for ways to adapt and change the way that they’re able to monetize their inventory? Or I guess I hear a lot from publishers about holding on to that – like you said, that first-party data is gold, right? How do we prevent data leakage? How do we remain in control in terms of choosing who gets to use that data, and how much they monetize it for, and increasing the value of the inventory?

Isha Chhatwal 

Yeah, that’s why you hear the keywords ‘privacy enhancing technologies.’ The clean rooms, you just don’t get to access the data, if there is just no use case as against having every data set available to you. A lot of tech changes as well – do you use GraphQL? Peter, correct me if I’m wrong, or as against running the whole set of queries.

Somer Simpson 

If that first-party data is gold, it’s gold; we’ll just call it gold. What percentage of a publisher’s daily site visitors actually sign up and provide deterministic data to publishers that turns into that first-party data that they then want to turn around and activate?

Isha Chhatwal 

I would say 10%, maybe way less than that, maybe actually 2%, who log in, but overall, publishers together say 10% of the data overall in the market, but usually hardly 2% data probably that keeps logging in. I don’t have the stats and the actual places that I pick data from, but I think this is from long-term analysis that we may have done in past one year.

Dr. Peter Day 

I hear a lot about, again, to our earlier conversation, a lot of hype around first-party data. I think one of the problems with data, in this data world we live in, is the amount of data you have – it has this kind of multiplicative effect, right? A small amount of data is worth almost always nothing. But then as you add more and more data together, you can find more patterns in it. And the greater the diversity of that data, the better. I think no individual publisher has that, right? Yeah, Google clearly has it because everyone goes there to search. And they track everything that you do, because they got the biggest browser out there. Apple’s got it now because a large portion of the internet users have Apple devices and so on, so forth. But no individual content producer has anything like the scale of data that’s really required to compete with the walled gardens. The entire thesis this company was built on was to level the playing field, was to try and build some sort of data cooperative, where everyone could put something in and patterns can be found in that so they could then extract value. I personally think publishers are going to struggle to turn their first-party data into something of real value, particularly the advertising ecosystem, in isolation. I think it’s going to be very, very tough because the types of data that we’re really talking about – this kind of world of big data and machine learning, which requires huge swaths of data, technology to make sense of it and organizing it, diverse data sets, going to pull together. So I think first-party data is absolutely, really, really valuable. But unfortunately, the smaller companies have to compete with the biggest companies out there. And so I think it’s going to require some sort of cooperative effort.

Isha Chhatwal 

But why would we even worry about [a] deterministic set of data, which directly confirms that this is the user, because we’re not going to do the same mistake again. That’s why we’re trying to get away from directly tying up that advertisement to a particular user, but rather opening it in [a] more contextual way or in trust’s way and putting them in some groups. I think just because we want to avoid the same mistake to very specifically target or activate a particular advertisement for a user, it’s mostly going to come as like subgroup and trust of which will be the evolving trend, given the lack of this deterministic set of first-party data.

Somer Simpson 

It’s actually a good point. I mean, if governments are upset about cookies being passed around, how upset are they going to be about hashed emails? Even if it goes through a clean room?

Dr. Peter Day 

Yeah, but I mean, the good news is: we know how to solve some of these challenges already. Right? Like, you know, I do expect there to be a subset of users logged in. And it’s been very small; I don’t think most people are willing to log in and create accounts all over the place. We’ll likely see some disruption in that space – and maybe that friction will go down a little bit – that’s going to have to be married with the ability for people to control who has access to that data. So I think some interesting bets in that space; no one’s cracked it yet. But, you know, let’s just assume that that login rates go up a little bit; it’s still going to be a small subset of the overall consumption of content, and we’ve got a name for that kind of audience: that’s a panel. And when you’ve got a panel, you don’t use that for things like targeting; you use that to validate and calibrate what you do elsewhere. And we’ve got techniques; we’ve got statistics to deal with that, and then taking care of bias in panels, and so on and so forth. So I think the good news is, we’ve already got the mathematical approaches to deal with this. One of the things that I think we’re going to have to move away from is this obsession with precision in digital. And we mistake precision for accuracy too often in digital. And what I mean by that is, when you kind of spend time with some teams, they’re obsessed with a CPA to three decimal places, not recognizing that that cost per acquisition number is wrong; direction is probably correct, but getting it right to two decimal places kind of misses the point of advertising, which is to reach and influence people at scale – groups of people changing their behavior. And so I think we’re going to have to give up some of this precision that we’ve got used to in digital and, you know, lean back into some of these core marketing-led approaches of audience understanding, testing and learning, using panels to calibrate and validate what you do elsewhere. Thinking about how can you have messages that work with larger sets of people rather than just an individual? And I think it’s gonna be really interesting, seeing how that evolves over the coming years.

Somer Simpson 

You kind of set it up nicely for the next question, which is: what are the sort of classifications of solutions that we’re seeing in market right now? What are the different ways that people are approaching this problem that we’re seeing?

Isha Chhatwal 

Surely, as we said, like there’s not going to be one-solution-fits all. It’s companies who have either direct access to deterministic data – emails, phone numbers, addresses, etc. – like solely deterministic solutions, don’t need to name them. Then we have probabilistic, which forms mostly like how you want to evaluate the rest at 90% of the market, given that deterministic data is anyhow too small to make an impact. So you have to leverage probabilistic that involves using the required data signals, of course with full consent. And then we have more of contextual, which again are more around how you put similar interests together and make a kind of wide group. And then, of course, something called as cohorts can be put in the same category or a different one. So we just really see these different set of solutions so far, but who knows, where we can lead to an evolving the solution, which is not just focused on privacy or consumer-first solution, but also about having increased performance, which we are kind of not too tied to, at least in the immediate future.

Somer Simpson 

We hear contextual thrown around a lot. And that means it could mean a lot of different things, right? When I first somebody say, “Oh, we’ll just solve this contextually,” I had flashbacks to, like, you know, 1999. It was like, “Oh God, not that.” Right? So, Peter, you and I’ve talked about this a little bit. When we say contextual, when you’re talking math and statistics and stuff, what is it that we’re really saying?

Dr. Peter Day 

Yeah, I think we are now able to take the benefit of [a] stellar leap forwards, there’s been in machine learning over recent decades, the sort of stuff we can do. I used to work as an academic in machine learning, and I can’t believe some of the advances we’ve made in specific niche fields around language and around images, and so on and so forth. What we mean is, we get computers to kind of understand the Internet of Things, like we do by reading it, and literally scanning the text of the pages, and kind of putting URLs in some sort of position in space. Computers are mathematical beasts. So they like to associate numbers with things. And when you associate numbers with things, you can do things like measure distances and compare things, and so when we talk about – the bedrock for any contextual solution is, first of all, based on going and scanning texts, like reading the pages, or extracting stuff, classes, videos, and so on and so forth, using techniques like natural language processing, to kind of make sense of that text and really turn it into numbers that computers can understand. If you think about it, it’s like building a map of the entire internet, where the position on that map is dictated by the contents of the page. So really similar pages will appear really, really close together. So you know, two pages about vegan cookery are going to appear really close together; a little bit further away might be a more general cookie cookery site; and further away might be a bunch of baking sites, and they’re all in the same kind of constellation, if you like. And then as you move further away, you get into car review sites, and so on and so forth. And this is stuff that we do all day, every day: the human brain does a phenomenal job at this; we can kind of scan a page and see broadly what it’s about. And we’ve now got the kind of statistical techniques to allow computers to do the same. So when we talk about contextual, what we really mean is: organize the entire internet in a way that’s based on the richest source of information we have, which is the actual content of the page. And we can make that understandable to computers, and then ultimately to people like marketers, who can suddenly answer questions like, “What types of, what parts of topic space are my potential customers hanging out in? What’s resonating with them really high right now? Oh, I’ve just got a whole bunch of new customers – what’s different about them in terms of the type of media they’re consuming, very, very specifically, compared to everyone else?” So it’s not the contextual approach of a lot of sites telling you trust us, this page is about SUVs, because it turns out as a lot of people spend money on SUV advertising; it’s taking advantage of the fact there’s a huge amount of data out there, which we can turn into information through modern technologies that were simply not possible a decade ago.

Somer Simpson 

Is that kind of the thinking behind some of Google’s solutions? Or proposals? I won’t call them solutions yet. 

Dr. Peter Day 

Yeah, I think a lot of the Google solutions have been somewhat subdued. You know, they had the FLoC announcement of some really clever ideas in that federated learning of cohorts and everything else. However, the first version of it didn’t really have any federated learning in it. So it was really, really good, just cohorts. And so yeah, they’re trying to do similar stuff to do this. They no doubt have a lot of the technical capabilities to do that. I think I have not seen anything that’s made me really excited about that specific technology. And it seems to be continually getting less and less interesting and useful. As I’ve made my trip, tried to make it more adoptable by a broader set of people and accepted by a broader set of people. So I’m not convinced that it’s going to mean an awful lot other than automatic classification of pages, which arguably all the exchanges already do. I think you need to have a deep connection between your data and how you’re using it. And so kind of giving it to a third party, I think, is always very, very difficult.

Somer Simpson 

Yeah. I read a lot about individual companies, coming up with ideas and solutions. And there’s tons of identity providers at this point; it’s not nearly as many as, you know, CMPs in 2019. But lots of people, trying lots of different things. I’m wondering, Isha, when you’re looking at the various identity solutions, what are examples of the industry coming together and working together to create something that we can all like, some sort of framework, like a set of rules?

Isha Chhatwal 

Sure. So to be honest, I’ve been in that industry for only two years. But the amount of news I’ve seen, because of the mergers and integrations that have happened, is a whole lot, I guess. And there are one set of companies who want to be risk averse and have [a] continuous stream of data made in the form of first-party data – so ID solutions really building, so that they have a stream of first-party data coming in, that they can continue to leverage. There are other stream of data, which just want to get away with the dependency on the third party ad server so that the advertisers can just track what they are offering, and doing the measurement on their own through either integrations with CRMs – so I hear a lot about Omniture, which is now Adobe Analytics, and other CRMs of that sort. So at least advertisers are coming out to share some additional stream of data so that they just get out of the dependency they previously had with other third-party ad servers. So these are kind of prominent, at least, in this point of where I am in my product lifecycle of the solution, wherein I am trying to see how I want to best measure not just myself and grade my work, but let our clients, the advertisers side, also do the same, like the way they want to. But these are coming out very frequently that they want to use extra stream of data, either through their offline streams of data, or something that they get from their CDPs, etc.

Somer Simpson 

And in terms of like, IAB has working groups, and I know Prebid is talking about this W3C – there’s lots of industry groups. Where are they lining up in terms of ideas and solutions or, if anything, just a set of rules and framework? Peter, what are you hearing from, like, IBW and W3C?

Dr. Peter Day 

Oh, it’s tough. I mean, doing anything by committee is hard. You know, so it’s tough to move some stuff forwards. People are working really hard. Don’t get me wrong. There’s a lot of tenacity in those groups. But there’s inevitably different perspectives. And so I think it’s going slower than almost all of the participants would like. But part of that is just, you know, it’s harder to do things in a committee. But ultimately, we will have to get this working across these different actors in this economic system. And one of the reasons that the tech giants have taken a larger part of the kind of advertising wallet every year is because they’ve been able to move fast because they haven’t relied on doing a few things by committees. I don’t think there is a big solution to that. But I do think W3C is having a bit of a constitutional challenge at the moment in terms of how they operate. And working through that at the same time, as you know, very different perspectives on what the role of W3C is, fundamentally. And then you’ve got people on the IAB, who do a huge amount of good in terms of technical standards, and so on and so forth, and find themselves in this massively political arena that I think was not anticipated even 5, 10 years ago, [by] the IAB. So I think it’s hard; I think that there is progress; there’s a lot of conversation happening. I think everyone wishes it was getting a little bit faster and being a little bit more decisive. But it’s really hard in a committee to generate alignment between different individuals and different companies.

Somer Simpson 

Alright, so before we wrap up, what – so this could be anything related to identity and the cookiepocalypse – what are your big predictions? What’s going to happen next –in the next, say, 6 to 12 months? I know that’s a long time in adtech.

Isha Chhatwal 

It is! As I said, every day is new news. But I believe there’s a lot in the oven baking. So there’s going to be a lot, many changes – for example, I won’t be surprised if Apple comes with [the] next set of privacy changes. And we probably see the Apple ID, iAd impact continue churning out a lot of revenue for big tech. But also, all in all, though in the past Google has given some breather, I think there is urgency looming, wherein at least the advertisers really want to test, optimize, and be able to measure right in time so that they can have that seamless transition in the world without cookies. But it’s largely going to be testing and testing a lot.

Somer Simpson 

Do you think Google will move the date again?

Isha Chhatwal 

I’ll be surprised, but I don’t know if I [would] cry about it or not. Because I spent enough time building the solution right on time even for the first deadline. But let’s see.

Somer Simpson 

Yeah, all right, Peter, what about you? What’s the next bomb to drop on that ecosystem?

Dr. Peter Day 

I think there is going to be increased scrutiny of the tech giants by governments around the world. We’ve already seen the beginnings of it [in] places like Texas and Europe. I think some of the practices of these very large companies who control so much is going to be deemed not in the consumers’ best interest. And so I think we’re gonna hear more and more about that; I do not expect that to resolve quickly. But I think if these tech giants are at a point where governments have to step in and do something, to control their ability to influence and put everything – because they skew the odds always in their favor. So I think it’ll be a continuing kind of trend that doesn’t come to any big conclusion anytime soon. I think, when it comes to adtech, I think we’re gonna get hopefully – I’m an optimist – we’re gonna get back to basics in terms of recognizing this as an economic system. All technology is there to support that, to make great technology; make simple things simple, complex things possible. Unfortunately, some adtech has made simple things really, really hard. Like, “Hey, I want to show ads to people a bit like this – can you do it for me?” “Oh, I’d need to spend three hours setting it up.” I think that’s just bad tech. So I think we’re gonna get back to basics a little bit. I think it’s going to start with measurement; I think we’re going to find, hopefully, somewhere between the precision of digital and the statistical rigor of the wall, the approach like economic modeling, I think there’s something in the middle, whereby you can take advantage of some of the precision you get from digital. But you can apply the statistical rigor to answer the question – is my advertising working? – that we’ve done traditionally, in marketing, where effectively I think measurement is ripe for maturing in a digital space. So I think I’m pretty optimistic about that. I think we – regardless of what happens with cookies, and everything else – think we’re going to see all advertising conditional and all digital slowly become programmatic. I don’t think programmatic means necessarily open auction and things, but I think just those pipes are assembled, much more efficient for people in the UK ecosystem, that I think everything’s going to continue to go that way. I think there are going to be a lot of companies pop up and try and create a silver bullet for this, a lot of this. One thing that surprised me most is how everyone’s talking about which method people are going to be using for hashing email addresses. No one’s really answering the question: how are we going to convince people to log in and agree to share their data, which is a much harder problem than how do you hash or encrypt an email address? But this seems to be some kind of standards wars there, which are more like VHS versus Betamax. I don’t think anyone’s ever worked out how to solve that more pressing challenge. What else…  I think it’s gonna continue to be a bit of a fun ride. It’s gonna be bumpy. We’re going to have to move fast; we’re going to have to always remind ourselves what we’re here to do.

Isha Chhatwal 

So, Peter, if Google actually moved the date all over again…

Dr. Peter Day 

My prediction is that cookies are gonna become less and less useful. If I’m Google right now, I’ve got this 800 pound gorilla problem of this cookiepocalypse, and it’s just causing me so much strife. They’re probably gonna move the date again. But if I’m Google right now, I’m just gonna, whenever we release a Chrome [update], make cookies a little bit less sticky, and a little bit less prevalent. This is what Apple has been doing; at no point are they announcing big dates, cause it’s slowly been adding this kind of stuff. And, over time, it turns out, you can’t really use cookies. There, that’s my prediction: if I’m Google, that’s what I’m going to be doing, I’m going to be, with every release of Chrome, cookies are going to be going a little bit stricter, and the default options for users are gonna be a little bit different. And so it’s going to be a bit like COVID going away. Everyone was expecting a big end date at the end of 2020. It turns out, it’s taken about a year and a half to kind of slowly go away. If I’m Google right now, that’s what I’m gonna be doing. Already today, we’re seeing Chrome traffic: half of it has no set cookies on it already; that’s probably going to become 60%, 70%. And that’s a much easier thing to handle. And, you know, 600 adtech companies constantly screaming at you to move the date out, or what’s it gonna look like, or needing a drop-in replacement with things like FLoc and other things.

Somer Simpson 

So, date moves; date doesn’t move; it doesn’t matter. The smart people are keeping their heads down and figuring out how do we get to web 4.5. 

Dr. Peter Day

There is a huge opportunity right now for marketers who start experimenting in this underutilized part of the internet. 

Somer Simpson

Cool, all right. So Dr. Peter Day, Isha, thank you for joining me. I appreciate you taking the time out of your insanely busy schedules, figuring out how to do something other than eat cookies.

Dr. Peter Day 

It’s been a pleasure. Thank you for having me, Somer.

Isha Chhatwal 

Thank you so much for having me. It was an exciting chat.

Somer Simpson 

Cool. Talk to you soon.

Dr. Peter Day 

Bye, bye.

Somer Simpson 

This podcast is brought to you by Quantcast. Our mission is to radically simplify advertising on the open internet. We are the creators of a new and innovative intelligent audience platform that empowers brands, agencies, and publishers to know and grow their audiences online. The Quantcast Platform, powered by our patented AI and machine learning engine, delivers automated insights, marketing performance, and results at scale to drive business growth responsibly. Our solutions are leveling the playing field for our customers when it comes to effectively reaching audiences online and helping them power a thriving free and open internet for everyone. Connect with us today at quantcast.com.