• by Joe Mandese  @mp_joemandese, March 7, 2022

Question: When was the last time GM asked you for permission to use your identity to target you with ads? When was the last time you consented to it?

Speaking for myself, the answer is never.

That’s what I was thinking when I watched GM’s Heather Stewart and her agency partner, Carat’s David Lai, open the Association of National Advertisers Media Conference last week with a presentation boasting how they have obtained the identities of, well, everybody.

Their presentation, which was literally titled “Everybody In,” was intended to showcase the “inclusivity” of GM’s consumer targeting system as part of a day focusing on inclusion, diversity, and what the ad industry is doing to support and fund diversity media suppliers.

And while Stewart and Lai certainly demonstrated how GM is doing that -- “whether it’s a Chevrolet Silverado Hispanic audience to a Cadillac Escalade African American audience to a Bolt EV Asian American audience” –-- they left out the part about when, where and how they explicitly asked and got consent for targeting everyone that way.

“We like to call it the CRM database of the U.S.,”  Carat’s Lai said, using the acronym for a customer relationship management (ie. “first-party” ) database to describe Carat’s solution enabling GM to target everyone.

The database -- which is actually Carat’s parent Dentsu’s, by way of its sister agency, Merkle -- is called M1, and Lai noted it has 242 million “personally identifiable” American identities in it. While that’s not everyone, it’s pretty close to the U.S. adult population.

If you’re like me and you don’t remember giving explicit consent to GM to use your identity, it’s probably because you did not.

But like many of the massive consumer identity databases that the ad industry has assembled -- whether its any of the big agency holding companies, like IPG’s Acxiom, Publicis’ Epsilon, etc. -- or the ones created by their media suppliers, their consumer identities were acquired through a long daisy-chain of consumer opt-ins, EULA, warranty sign-ups, as well as actual CRM “loyalty programs” that someone may have given consent to some third-party at some point for, but most likely never gave explicit consent to have their identities used to personally identify them as part of everyone. For perpetuity.

So whenever I see some humongous industry entity boasting that they have accomplished this, I immediately wonder how tone deaf they must be.

Just the other week, during the IAB's annual conference, incoming chair and NBCU honcho Krishan Bhatia touted "NBCUnified," a “first-party” consumer identity database that has already amassed the identities of 150 million people and would “scale to 200 million in short order.” He said the identities were acquired through NBCU properties, but he did not say how or whether people had explicitly consented to having themselves targeted personally based on their own identity.

Now I know most of this is not news, much less surprising, to MediaPost's readers, but there is a reason I'm pointing it out today, and it's because I think the industry has split into two different directions when it comes to personally identifying people: a good one, and a not-necessarily-so-good one.

I'm going to suggest that based on my own personal identity, the ones I described above would fall into the not-necessarily-so-good category, because I'm pretty sure some form of my identity is in every one of those databases, and I guarantee you I did not give my consent to have them personally identify me the way they are or are intending to use it.

"It's like a joke," MediaPost's Wendy Davis told me some time when I asked her how consumer privacy experts looked at the massive "first-party" databases the ad industry has assembled. Davis, who is probably the smartest and best-informed journalist covering the subject, was explaining to me explicitly how they viewed the deprecation of cookies and other digital consumer tracking IDs.

She said the industry has done such an amazing job with all its fancy algorithms, machine learning, identity resolution techniques, lookalike matching, deterministic/probabilistic, and other black-box-ad-tech-black-magic, that it can effectively personally identity anyone without actually identifying them. If you know what I mean.

On the plus side, some great organizations are moving in the other direction. You know, the good one.

It's one of the reasons we named GroupM as our holding company of the year and its Wavemaker unit as our agency of the year, largely because they have not just paid lip service to consumer identity consent, but have actually operationalized it. And they began doing it the right way, by applying ethics, including a sophisticated operating system called an "Ethics Compass" that all parts of their organization are supposed to go through before determining whether to utilize consumer identity data for their clients.

It's a remarkable turnabout for GroupM, because it was only a few years earlier that it was bragging about its ability to personally target everyone via its own massive proprietary consumer identity database, ironically also called M1.

GroupM is not alone. Plenty of other organizations have not just seen the writing on the wall from a regulatory and legal compliance point-of-view, but from a humanitarian one. Because, you know, they're people too. And they've applying an ethical approach to consumer consent to be personally identifiable. And that means more than exploiting a moment of time, showing someone an impossible-to-read verbose legal disclaimer, and then mining their soul for perpetuity.

It means thinking about your relationship with them every time you want to relate to them.

It means having a consensual relationship with them.

This column began with the opening presentation at last week's ANA media conference, so I'm going to conclude it with one made near the end of it. The one updating the industry on the progress of its Cross Media Measurement initiative. Because as far as I understand it, I think it is another good example of the right way to do this.

The initiative, which is part of parallel global and multinational efforts with the World Federation of Advertisers, and other advertising trade entities, is close to piloting a test of a new system that will utilize a new consumer panel -- one that people explicitly opt into -- to enable marketers to effectively model how their own "first-party" data reaches equivalent "cohorts" available through media.

The panel will not be an audience panel in the sense of Nielsen, MRI, Scarborough or other industry media audience ratings, but will only be used to calibrate the relationship between the data a marketer has ostensibly gotten consent to use and the universe it is trying to reach.

As best as I understand it right now, it is more of a Rosetta Stone than an audience panel, and it seems to me to be a very practical, honest and transparent solution to the problem.

It also leans into another important industry trend, which is the resurgence of contextual targeting, which doesn't rely on the personal identification of a media supplier's audience, but on the contextual value of consumers who want to be there in the first place. And of course, how that aligns with what a brand wants to tell them.

I don't think I'm coining this phrase, but I think we should call this "consensual targeting." Because whether it is done with a specific consumer's explicit consent or merely based on the art and science of understanding their cohorts and the context of the media they use, it is the right way to target people.

Especially when they're everyone.