The Ad-Tech landscape saw a flurry of consternation, across all of it surrounding cookies’ and the view that Google had single-handedly decided to end digital advertising as we know it. The truth is, of course, more nuanced and I, for one, think the deprecation of cookies will result in advertising being more effective, future-proof & consumer-friendly.
To understand what happened, it’s helpful to have some perspective on how we got here.
Part 1: How we got here
What’s a cookie? The idea of addressability has fascinated marketers since the first digital ad was served in 1994. Addressability, in theory, is the ability to serve the right person with the right message at the right time, every time. To do this, you need some type of persistence so that we can distinguish me from you. So if you’d like to find out what “cookies” mean, go back to the basics.
What’s a cookie in its most primitive sense? It is a web standard that’s neither good nor bad. All it does is make sure that I can identify “you” the next time I see “you”. The original intent of the cookie was to allow publishers to store information about your preferences to personalise their website in an era where the Internet bandwidth was limited. By removing calls to the server, this sped up the page load time.
How do cookies work for advertising? As more and more of our lives moved online (shopping, news, etc.), ad-tech discovered that cookies could also be used to store information about your interests as you browsed across multiple websites. Further, these cookies could then be shared and matched with other ad-tech vendors who would append what they knew about that cookie targeting. And so a robust marketplace developed where companies would profile users with what they and others knew about online & offline behaviors, and make that information available so that marketers could achieve their goal — Right Person, Right Message, Right Time, All the Time.
What were cookies from the governments’ point of view? At some point, governments thought that this collection of data was not a good idea. We wound up with a patchwork of laws around the world to attempt to limit this profiling. None of these laws did anything to curb the practice. What changed, though, was when browsers — Safari, Firefox & Chrome — also decided that cross-website tracking was bad and should not exist. Safari & Firefox were the first to move, declaring that 3rd party cookies (those used to track across multiple websites) were being deprecated. Last January, Chrome declared the same thing.
Ad Tech was up in arms. And why not? They had built businesses and fortunes transacting on consumer data, and now that system was being taken away. They decried that Google, owner of Chrome, was being monopolistic. Or that this will herald the end of the ‘open web’.
But there are several problems with their arguments. To begin with, cookies and advertising are terribly inefficient ways of sharing data. Each domain has its own cookie and so to share data, you wound up with hundreds of redirects and pixels flying everywhere. So it’s a really bad technological solution to achieve a marketer’s goal. It’s also overly complex. This complexity has made it possible for thousands of companies to build nice solutions that ultimately acted as a tax on spend.
Then there is the issue of the consumer. No way do they have an idea who possesses their data, whether it’s correct or how it’s being used. Finally, cookies are not actually people. A single person could have several cookies and several people could share a single cookie. In a very real way, that promise of the Right Message, Right Person was never properly met. In reality, because of all the inefficiencies in marketing, we’re probably closer to the old adage: Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.
The ad tech ecosystem absolved ourselves of blame by repeating the mantra that it is keeping the Internet free. But there is a better way to achieve our goals. A way that preserves privacy. A way to meet advertiser’s needs. A way that provides for an ad-supported Internet. That path was laid out in January of 2020 when Chrome said that they weren’t going to just get rid of cookies (as Safari & Firefox did). They were going to work through the Internet standards body World Wide Web Consortium (W3C) to find a suitable replacement which would place the consumer squarely in charge of their data.
So what does that look like?
Part 2: Focused on the What & Why of cookie ad decline. The How will come later.
Over the last year, I’ve been a part of the W3C, listening to the proposals, the counter-proposals & the invective, and I’m more convinced than ever that what Google is proposing is the right step. Why? Because cookies were never meant to be used like this. Marketers are realizing that 1:1 user-level addressability is likely unachievable and perhaps counterproductive. It seems that the individual doesn’t matter in cookie advertising. The consumer needs to stay aware of what data is known about them and where that data resides. Because the marketplaces that built up around the exchange of data almost never delivered what they promised.
There’s been a lot of arguing that the process of developing these standards was opaque or didn’t take feedback — so let’s look at the process. The W3C (specifically the Improving Web Advertising Business Group) is something of an anachronism. It uses IRC (an old-school group “chat”), has rigid guidelines for doing things, is open to paying members only, and the initial conversations were confidential. But the W3C is also the bedrock of why the Internet has flourished, even with all its oddities. The standards allow entrepreneurs to build the new tech and innovate the old tech.
So every Tuesday morning, anywhere from 50-200 people join a WebEx and IRC channel and talk about whatever anyone has proposed. The early conversations were privileged, but it’s important to understand that what was being discussed was all theoretical. These were proposals that identified the What and Why and largely left the How for later. The primary speakers during these initial phases were the Chrome team and for them, the What and the Why were:
- What: We need a privacy-preserving browser standard that enables digital advertising.
- Why: Cross-site tracking is a violation of end users’ privacy.
As a starting point, the Chrome team delivered a few proposals that have evolved but have mostly intact principles. At the heart of the proposals is Turtledove which introduced the concept of the Privacy Sandbox. “What the hell is that?” you may ask. If you’re familiar with the omnipresent Lumascapes, with an advertiser on one side & a publisher or retailer on the other side and a gazillion vendors in the middle, you know it. Those vendors in the middle today are building profiles of users, doing attribution, checking for fraud, etc., and are selling these things to either the buy side or the sell side. Imagine that each of these vendors, today, has a sandbox where they are collecting cookies, mobile IDs and other identifiers to provide these services.
Turtledove flips this on its head. It says that 3rd parties (all those ad tech vendors) can no longer collect cookie data for profiling. Instead, the only sandbox that exists is in the consumer’s browser. It does this by limiting the data that leaves the browser and how code on websites can access that data. Instead of ad-tech vendors building proprietary audiences, the audiences would be organised and held on the browser in so-called “interest groups”.
It also originally stated that the auction would happen on device rather than at an SSP or an exchange. The rationale being any information that could de-anonymise the user should be eliminated. This stance has softened a bit over the last year, and the idea of secure multi-party computation, or MPC, has been embraced by many members of the community. We won’t go into it in detail but suffice it to say it’s a method of making data available in a cryptographic way that ensures anonymity.
For those familiar with Ad Tech, this is a monumental shift. This basically obviates all those vendors between the advertiser & publisher/retailer who were profiling users. It also threatens DSPs which applied all this data on a cookie-by-cookie basis.
Part 3: Google’s motivations and frustrations regarding cookie advertising
The new standards of cookie-based ads now being developed will create turmoil. As I read some POVs from companies who are clearly conflicted, I feel sorry for them. There is going to be the ability to build a better advertising landscape, yet they want to hold on to what is fundamentally broken — the cookie ads. I mean, Safari & Firefox blocked cookies 3 years ago! Meaning anywhere from 30-50% of Internet traffic does not have a single сookie. Advertising has changed dramatically.
Many people have also tried to divine Google’s motivations in doing this. Here’s my take. Google is fundamentally a product company — YouTube, Chrome, Ads, Gmail, Cloud, etc. These products work entirely independently and do not coordinate roadmaps or agendas. I firmly believe that Chrome made these changes because they believed that it was the best thing for the Chrome browser. And why not? All the other browsers had already done this!
Google Ads team, who have been contributing more & more to the W3C process, said that they would follow Chome’s lead and not leverage any cross-site tracking or support third-party IDs. I do not believe that this was the plan from the start because I don’t think there was any initial plan at all. Chrome did what Chrome thought was best in terms of targeting and advertising cookies.
Of course, not everyone believes this. Jeff Green at The Trade Desk doubted that they would get rid of cookies at all. Oracle published a blog stating that Google is doing this to gain access to the data held in the Chrome browser. That’s cuckoo. And wrong.
Meanwhile, there is the continued push to use PII/E-Mail. What for? Getting down to an individual? Do marketers want this? Do consumers want this? The idea of micro-targeting has come and gone. It’s time to move forward from cookies in advertising.
Part 4: The impact of advertising cookies abolishment on Ad Tech
It’s important to understand that the proposals being debated are the first step. Meaning, the marketing use cases that they solve are just a subset of the current capabilities in the marketplace. That’s a problem for everyone — companies that have built a business based on providing those services and marketers who have relied on those techniques.
Let’s take a look at some of these impacts.
- Lower-Funnel Tactics (Retargeting). Unlike countdown ads, retargeting is probably one of the least liked aspects of digital advertising by consumers (looking at you, Criteo), and it is one of the most thought-out aspects of the proposals. However, instead of the segmentation of what you’re interested in happening on a company’s server, all the segmentation happens within the browser via a javascript API. This API, however, is very limited in its scope. There is no concept of product recommendation that would leverage server-side AI & ML. But the ability to retarget does exist.
- Mid & Upper-Funnel Tactics. There have been a number of proposals that address typical upper-funnel KPIs, such as Brand Lift studies and surveys. The data that fuels these campaigns — usually 3rd party, cross-site or offline based — is squarely not permitted. Instead, this data now moves to being browser-generated and held. How this practically happens is still unknown, and there is growing momentum for the idea of running ML on the browsers to produce look-alike audiences via proposals called FLoC.
- Advanced Measurement. MTA, Test-and-Control & Omni-Channel measurement proposals have been made and are generally well-received. However, much like the mid- and upper-funnel tactics, these are “down-the-road”. They understand the importance of measurement more so than targeting.
- Emerging Media Channels (Audio, DOOH, CTV). I view this as a positive for these channels because the standards that are being considered can easily be applied to these types of media where data is held on the device, aggregated, and reported centrally. This means that these channels can ultimately become mainstream and be folded into a media plan.
- Advertisers. Over the last 10+ years, ad tech has pitched advertisers with the promise that they couldn’t deliver on. “You can have a 1:1 conversation with a consumer and not be creepy”. “DMPs will pay for themselves”. “Collect all the data you want, no one will object”. Each of these created unfulfilled promises. But now the tables have turned. Advertisers and publishers, not 3rd parties, will be the keepers of data. This will require more effort from advertisers to curate this data and make smarter decisions with the data.
- Agencies. I spent 18 years in the world’s largest holding company (WPP). There was a concerted effort to lean into digital. These changes are going to challenge them in ways that they weren’t foreseeing. The changes will lead to more silos (think FB, Pinterest, Amazon, CitrusAd, Etsy, YT, Walmart, Amazon, etc). All those “walled gardens” mean more planning for smart agencies. Tools will need to be built and human force will be required to operate them. I believe agencies will be net winners in these changes.
- Big Tech companies that gobbled up Ad Tech companies. Think Adobe (Omniture/Demdex), Oracle (BlueKai/Datalogix), Salesforce (Krux). All these companies traffic in collecting user-level data for analytics & activation. Look for them to have huge write-offs of these investments in the near future.
- Large publishers. Large publishers will do much better than small publishers as they will begin to (finally) build audiences that advertisers & agencies want. Small publishers will be challenged since they won’t be allowed to launch any single cookie ad to generate leads.
- Small publishers. Everyone loves the underdog, and the small publisher has always held a special place in the Internet lore. To the extent that they are advertiser-driven with a small user base, they will be challenged, at least, initially. They will not have the scale to build PMPs with advertisers & they may not be able to convince consumers to sign-in. This will cause CPMs to drop and will challenge them. I don’t think there’s good news here.
- Retailers. Retailers hold a lot of keys. To the extent that they are able to bring together online and offline conversion data, this will drive brands and agencies to spend with them. COVID-19 changed everything with regard to retail media, and retailers now have an opportunity to capitalise on the changes in consumer behavior and the changes in technology. These new technologies embrace the 1st party data, and retailers are now able to capitalise and monetise their 1st party data.
Part 5: The music’s about to stop, so what are we to do about it?
There’s a great scene in ‘Margin Call’ where Jeremy Iron’s character says something along the line of “There are 3 ways to make a living in this business: Be first, be smarter or cheat…it sure is a hell lot easier to just be first.”
My message to everyone I speak with is to lean into these changes. Learn as much as you can about them, ask me. Consider “workarounds” with caution. If one of your partners tells you they have a solution for this, your spidey sense should tingle. Because new technology doesn’t exist. We are on the path to a brave new world. It will contain bumps, but it will build a more sustainable, efficient marketplace.
