Programmatic advertising

Ralf Ollig: Google announced in 2020 that it wants to phase out the cookie. It’s still here. What will happen going forward? What has happened since 2020?

Barry Adams: Just for context, the cookie was crumbling long before all this with Google. The largest browser in the world is the Google Chrome browser; it accounts for 65% of traffic globally and it varies by market a little bit. That was where the third-party cookies were really alive. If you look at places like Safari and Firefox, they were phasing out cookies a long time ago. So, the remaining bastion of cookies has been in the Chrome browser. Google announced in 2020 that they were going to deprecate that, which was supposed to end in 2022. Then in 2022, they said they needed more time and now they are going to do it in 2023. They kicked the can once again and the latest is they have a proposal that is going ahead of the CMA, the governing body in the UK. They are live-testing with this body in Q3 of this year. If those tests go as expected, they will start to phase out and deprecate the cookie in January 2024 and the full phase-out of the cookie will be through next year. So, by January 2025 there will be no third-party cookies in Chrome. If the CMA plays ball. So that’s roughly where we are today.

RO: Could that date be postponed further?

BA: I think that they have committed to these dates now. Not just the embarrassment of having already postponed this, but they have put a tonne of resources into the Chrome team. Since 2011, the e-privacy initiative that the European Commission put in place; didn’t go that well. The implementation was a bit sketchy, to say the least. This whole thing has been cooking for a long time. It’s just been the US that has been a bit of a holdout. Europe has been on this for a long time. But I think Google is going to do it and the ad team is a little more aligned. Especially given recent earnings calls and pressures from the market. Google has a lot to gain if it can have a decent environment that can support addressable advertising, not at the level of individual users, but close to that, they could probably gain share through Chrome.

RO: What do we do with the cookie today and why is it so important to us? How do we navigate without the cookie and what are the alternatives for us?

BA: The strange thing about this is that it is kind of like a frog boiling in water at this point. The frog is happy in the water until it’s dead. This is real this time. This is going to happen. My advice to everyone would be if you are not yet testing alternatives to the third-party cookie, you should be. We have around maybe five to seven tactics you can deploy instead. I want to be clear that even if you deploy all these tactics relatively successfully, you will not have the coverage you get today from cookies.
  It would be nearly impossible. Partly the reason is the whole industry, driven by privacy concerns, is moving away from individual user level targeting and user identification, towards segments and cohorts etc.
  Some of the tactics you can and should be testing will vary from industry to industry. If you can leverage and collect first-party data, that is the gold standard. You want to have a first-party data strategy which involves the ability to collect data directly from consumers. If you’re not on the supply side, like a publisher or media owner, you want to be working with others who can do that to some extent. A lot of stuff will be modelled on your own audience data, emails, phone numbers, the personal information you can get directly from consumers. That is super valuable.

RO: Looking at the timelines, here are some solutions people should be considering. One of many options that we are looking into is contextual, a better understanding of how to apply the context of websites and where to show users to reach the audiences. Using a service that helps us understand the website. What is your understanding of this one?

BA: You should be testing contextual for sure. Contextual can be relatively powerful depending on your product and service. If I think about sports betting, that context is still available. I would absolutely be looking at contextual as an alternative. Again, it won’t completely cover your butt. But it will do a slice of the pie for sure.

RO: What you won’t have is the measurement. So you can do the targeting, but measurement might be tricky for the contextual part when conversions are happening.

BA: The hardest thing for everyone, including our company, is that fundamentally you have to reconceptualise how you think about measurement and attribution. Moving forward it’s not going to be possible to have user-level attribution in these kinds of environments. This is also true in the app ecosystem. Apple for a long time now have been removing the ability to track at a user level. But there are solutions to these things.

RO: We looked up a few numbers from our platforms. Depending on country and region, on the attribution side, our cookie attribution is anywhere between 40-60%. So depending on the penetration of Chrome we have more cookies. But even today the complaints you are having and running are being attributed to non-cookie-based conversions. When we apply contextual targeting, you can still track them using fingerprint methodology. Right now, we are using a combination of user agenting and IP addresses. Hearing what the browsers are doing moving forward, such as removing the IP address, do you think that is a solution that would be sustainable?

BA: It’s going to be like whack-a-mole for a while. The second-largest advertising platform out there right now, now considers IP address to be PII and they will not use it. In anything. And they’re huge. So it’s a signal, I wouldn’t bet the farm on IP address. I’d say you hedge your bets and do what you can there but don’t be a happy frog on IP address.

RO: For the clients that are running mobile app install companies, there is still a lot to figure out, such as how to track in-app events, how do we use different apps between sportsbook and casino, how do we mix it up and how do we see traction there? There is still a lot to be figured out, especially on IOS but also on android. But cookies won’t go away completely. You can still opt-in? You can try to find ways how you utilise first-party cookies and pass on data through a first-party cookie to a third party.

BA: First-party cookies will still be around.

RO: The key takeaways are that there are multiple solutions coming from contextual, Google Sandbox approach, PIDs and other things people should be talking about in preparation when the cookie does crumble away. Anything you would like to add to this?

BA: We wanted this to be practical. At the end of the day, I would not be the frog in the pot of water anymore. It was reasonable to take time because Google was kicking the can and it was going on forever and there was a real debate; will they get rid of the cookie at all? I really don’t believe that it is going to be around, so my prediction is that by 2025, it’s gone! How long does it take a big company to get a plan together? In our case, it’s a couple of quarters debating and a couple of quarters to see what we can implement. There goes that year. So I would get on it now, I would create some urgency and discuss the optimal ways to deploy the tactics we’ve talked about. All this cookie matching and syncing a huge amount of carbon that is going up into the environment. So the earth will heave a sigh of relief when the cookies are gone. Even if we don’t.