Why A Unified ID Will Never Work – And Why It’s Time To Fall In Love With Walled Gardens

Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media. 

Today’s column is written by James Avery, founder and CEO at Kevel.

The digital ad industry seems faced with an existential choice – go Google’s way, and give up your destiny while bearing the consequences.

Or band together and fight.

It’s that second route that is leading to some unexpected enthusiasm. Could the beleaguered publishers of the web, often left to fight for the revenue scraps Google and Facebook don’t get, really join forces and level the playing field? Maybe even put the walled gardens on their heels a bit?

I’m sorry to be a downer – but I think we’re kidding ourselves. The idea of every publisher and ad tech company getting together and agreeing on a new unified ID – whether via LiveRamp, The Trade Desk or dozen other alternatives –is about as likely as Mark Zuckerberg and Tim Cook becoming best friends.

I’m talking of course about the impending end of the third-party cookie, big changes in how advertisers can track mobile ads (thanks to Apple) as well as regulation.

There’s no doubt that digital advertising is staring down a fundamental shift/threat to its business model. This industry has largely built on its precise targeting power, and that power may be set to disappear. Just look at the recent layoffs and restructuring over at Criteo, a company built on retargeting.

It’s certainly understandable to find any way possible to hold onto this deterministic potency. It’s scary to imagine going back to a time when publishers would have to sell ads by using content as a proxy (men visit sports sites, right) or by using audience surveys and the like –especially as titans like Google, Facebook, and Amazon will keep their logged-in user advantages post cookie. 

This is why there are roughly 20 different ID alternative projects in the works all vying to fill this coming void

It’s also understandable that the ad industry is wary of the cookie-replacement solution that Google is spearheading –because funnily enough, Google tends to do things that favor Google (like any company would).

So let’s instead take our hard-earned user data like subscriber info, email addresses and/or wireless numbers, pool it all together and come up with a new consistent way to identify web users –then let the ad targeting begin!

Wait a second here.

All we’ve heard as of late is how every brand and publisher wants to be a direct-to-consumer brand. Not only do these companies want data on their customers in order to market to them, but they also want the trust and long-term relationship this promises.

Now they’ll all of a sudden start sharing this info? The New York Times and Dollar Shave Club and Hubble Contacts are all going to work together so that tons of smaller brands and blogs and whomever can benefit? I don’t see this as likely.

I don’t care how many encryption and non-PII steps this process promises –this is not going to happen. It’s not in these companies’ long-term interests. 

To be sure, there are some signs of big players coming on board. For instance, The Washington Post has committed while exchanges such as Magnite and OpenX have endorsed the project. You can make a case that advertisers –theoretically –would like to be able to continue to target specific audiences online –especially on brand-safe properties.

However, think about the history here. The airlines have been trying to create a unified way of booking flights, called NDC, for the better part of a decade. How much time did the cable TV world try to create a common ad platform in Canoe before giving up? 

Closer to home, how many premium, brand-safe ad sales consortiums have we seen, only to get very little traction (and barely get noticed by the duopoly). From QuadrantONE to Pangea to Brand.net to ShortTail Media to TrustX – inevitably, individual publishers always prioritize their own business and struggle to enact change. Even ads.txt, a seemingly non-competitive effort aimed at stamping out ad fraud, has struggled to get mass adoption.

And If you don’t get mass participation, these UID projects are DOA. Further, Google has come out against the concept of many of these UID projects – which often use email as the identification basis. 

To be sure, the stakes are a lot higher this time around – if your entire business feels suddenly vulnerable, you may be a lot more inclined to prioritize a consortium.

Still, I know this sounds negative, but I think we’re better off. Instead, it’s time to let more walled gardens bloom.

Think about it. As dominant as Facebook, Google and now Amazon are, other ad platforms like Pinterest, Snap and TikTok are all in a position to increasingly steal share.

Meanwhile, we’re already seeing a wealth of retailers follow the Amazon path and dive into advertising, using their robust consumer data – think Walmart, Target, Walgreens, CVS, UberEats, Instacart, etc. We’re likely to see lots more of these mini-gardens pop-up.

We think building a custom ad platform can be a win for publishers and brands because they’ll have more options and better control of their own revenue destinies. Adopting a new independent standard – with Google or otherwise – is liable to change again in the future, and we’ll be having this same debate again.

The reason we are in this spot  is because so many of us relied on Google and cookies. Publishers who can pivot strongly to direct-sold ads and proprietary data will be much less impacted by arbitrary changes driven by tech giants or lawmakers.

It won’t work for everyone, and maybe won’t be a fair fight, but this is a competitive business. The publishers with high-revenue, successful ad platforms will have more money to grow than ones that don’t.

Better to win on your own merits than end up riding someone else’s wave and hoping you don’t crash. 

Follow James Avery (@averyj) and AdExchanger (@adexchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. Amelia

    This is an astonishingly bad take. Advertisers need to be able to measure their digital media dollars holistically… which they can’t do when everything is its own walled garden. End of story.

  2. Terrible take, and clearly in his best interest to promote “more walled gardens” as Kevel “builds custom ad platforms for sponsored listings, internal promotions, native ads, and more,” effectively helping brands create their own walled gardens. Ad Exchanger could do a better job with editing out clearly biased opinion pieces with few facts and a huge heaping of self-promotion.

  3. But Amelia, isn’t that the case with walled gardens now where most of the spend is going? They seem to drive ad revenue just fine inspite of it.

  4. Sir Larry Wildman

    Another vote for incredibly bad take.

    You could have used the exact same argument to argue why header bidding would never work. Adtech companies working together on a common goal to prevent one company from gaming the auction process, that would never work. Or what about the RTB protocol itself, Adtech companies working together to build an ecosystem that is not controlled by a single company, that would never work.

    I’ll take the other side on this one, Adtech will find a way, the ecosystem is diverse and dynamic by design. Synthetic ID, user privacy first design, contextual analysis, FLOC and SKAds integrations, Adtech companies will support all of it and whatever new format, features or challenge that emerges.

  5. When I read this article and consider the recent agreement entered into between Walmart & The Trade Desk, I am more convinced than ever that no matter what way the road ahead looks, TTD will be a solid front runner. J Green will guide the industry to higher heights.

  6. Agree — bad take, brands (and consumers) as it turns out need and want identity driven personalization and attribution/measurement, that requires a consented universally accepted ID. The examples cited with Airlines and Cable Cos. are hardly worthy examples of poor collaboration and standards tying an industry together. SQL in the 1990s, HTTP, VAST, VPAID, OAuth, SSL are examples of hundreds of standard developed by competing companies for the greater good of the industry. In the end brands will decide where they spend their dollars and it will likely be wherever there is more transparency, accountability and attribution.

  7. Perhaps we need to consider disciplines other than engineering to come together and find a solution. After all, if you’re only tool is a hammer every problem looks like a nail.

    A solution based on economic, legal and engineering disciplines. A solution where people, not corporations, own their identifiers and identity. A solution where operators steward data rather than fight over ownership.

    Whilst fear and confusion reign it’s easy for people to accept or advocate for what appear to be satisfactory rather than optimal solutions. This is know as “bounded rationality”.

    Just look up at the sky and see an airplane to see how a common resource, the sky, can be used by multiple competing organizations and keep people safe. We don’t ask people to understand how air traffic control works before booking a flight. Why should the digital market be any different? After all in 2021 digital advertising is forecast to be worth $100bn more than aviation.

  8. Patrick

    Okay, I’m glad that I wasn’t the only one who read this and thought this was a bad take on things. This makes me think that the author is in favor of giving the oligopoly of Google, Facebook and Amazon more power, rather than grab it back for a healthy ad tech (and consumer) ecosystem. We need less from/for the walled gardens, not more.

    I work in a cooperative data environment, and it can be done (and I’m not the only one either).

  9. I think this is a realistic take. In the meetings I’ve listened to with the prebid team now in charge Trade Desk’s UID 2.0 (UID opensource), the sentiment is that we will be lucky to be able to get a 1:1 match for 20% of programmatic auctions in 2022, and realistically this match rate will be 10% or lower as it requires the user to login with a publisher sharing the encrypted email. I believe that unless publishers are incentivized to share their login data that UID will struggle to get the critical mass necessary to grow beyond a 10% match rate. Publishers really have 3 choices: Subscription Model, Build your own ad products leveraging your unique assets or hope that Chrome Floc targeting are going to fill the loss of revenue from non-matched users. While I don’t like to rely/hope on Google in many ways floc targeting may eventually prove to be more effective than cookies because it based on your chronological browser history meaning Google could give weight to your most recent online activity allowing marketers to target you based on what you were just doing 1 minute ago. Publishers/Developers should be hedging their bets, and expect that the programmatic channel for the open web will lose >50% of earnings when cookies deprecate for the short term.

  10. Jerome Taillard

    What about solutions offered by CRTO? They have 22,000 publishers working with them, a map of 1st party data on 2.5Bn users. They basically stitched together all the individual data into one big database. They are working with TTD on UID 2.0 but they have been very clear (like others in the comments) that they are preparing for many different solutions (e.g. contextual, etc.) and the marketplace will determine what the winning strategy will be. It does seem that their approach is quite robust. Could it be that the firm that was the most threatened by a cookie-less world (they were #1 in retargeting) has spent the last three years spending half a billion to become the ultimate winner (outside the major walled gardens)?!?