January 25, 2023
One Of These Things Is Not Like The Other: Vice Calls To End 'Data Discrimination'
- by Joe Mandese @mp_joemandese, May 6, 2021
In what may be the most compelling argument yet to finally end the ad industry’s long-held practice of targeting media audiences based on their demographic composition, Vice Media Group Wednesday made the case that the method is not just ineffective, but outright “discriminatory.”
“Let’s grow up, and let’s stop playing the generation game,” Chief People Officer Daisy Auger-Dominguez said during Vice’s virtual NewFronts presentation, charging, “The time is now to end data discrimination.”
The message was profound, coming at a time when Madison Avenue is reassessing every aspect of diversity and inclusion -- from the composition of its own organizations to the media it plans and buys, as well as the research methods it uses to do that.
advertisement
advertisement
It’s not the first time someone has issued a rallying cry to move beyond demographic targeting, which developed as a stopgap effort to move beyond Nielsen’s household ratings in the late 60s and early 70s.
As far back as the mid-1970s, renowned media researcher and now retired CBS research chief Dave Poltrack was making the case in the Advertising Research Foundation’s Journal of Advertising Research about why “demographics were a poor surrogate for targeting” consumers.
As much of an improvement as demographics were over the “boxcar” numbers of household ratings, indexing and targeting consumers based on age, sex, income and other demographic attributes has long been known to be a weak indicator of consumer intent and propensity.
More than a decade ago, legendary media planning authority Erwin Ephron used new “fusion” data linking the disease states of actual consumers with the Nielsen demographics used to target them, and during an ARF conference showed how the demo for targeting allergy sufferers was about 25% less accurate than targeting actual allergy sufferers.
But old habits die hard, and despite innovations in psychographics, behavioral targeting, cookie-based “deterministic tracking,” and a host of other much more refined methods of identifying and segmenting consumers, demographics have persisted.
But noting how the cookie is finally “crumbling,” Vice’s Dominguez made the case that the ad industry is finally at a “unique inflection point” to move to something far more inclusive of the way consumers actually identify themselves: the contextual attributes embodied by their “passions,” “hobbies” and other lifestyle indicators.
To illustrate this point, Dominguez flashed the image of two well-known people -- Elton John and Donald Trump -- on the screen during her NewFronts presentation and explained how they would both fit into the same demographic cluster: “Both male. Both 74 years old. Both ultra-high net worth. They are what a boardroom would call a ‘tightly defined target’,” she said, adding: “Best of luck trying to create a campaign that is going to cut through this tightly defined target.”
She then went on to assert that the ad industry uses that logic every day to apply “this type of bad strategy at scale” via demographic targeting.
It was a visceral example.
Not surprisingly, Dominguez followed up with a tacit, but hard sell for Vice’s product -- noting that its research team has spent 18 months developing a “community-based contextual targeting solution that allows brands to target these communities with off-the-scale rates of engagement.”
She claimed that Vice’s research found that the contextual targeting approach boosts average returns on “advertising goals by 289%,” and that the method reduces “cost per acquisition across display ads by 2.5 times.”
4 comments about "One Of These Things Is Not Like The Other: Vice Calls To End 'Data Discrimination'".
Read more comments >
Next story loading 
Joe, once again we see the ongoing confusion about media versus advertising "targeting". While demographics---usually in a fairly refined manner, not just adults aged 18-49----is one of the ways that national TV advertisers define the groups they want to influence with their commercials, most ad campaigns go far deeper than that, appealing to varying mindsets as well as product user groups. Just because media buyers who want their fannies protected in case of audience shortfalls make pacts with the sellers to guarantee audience tonnage delivery using broad and virtually meaningless age/sex breaks, this does not mean that the brands are targeting consumers in this ridiculous and simplistic manner. So endlessly beating this dead horse is a waste of time.
As for better ways to target media buys, sure, in an ideal world we would try to match not only the level of product usage in a media vehicle's audience but more important, what percentage of the audience has the type of mindset that our ad campaign is trying to motivate. It's not merely a question of targeting just product users. You need to go deeper to get the most positive results. And there are sources which provide information about both mindsets and product use---the MRI/Simmons studies for example. However, will the sellers guarantee audience tonnage on such refined metrics? And what happens when the sellers note that some of theirTV shows are suddenly in great demand---based on improved targeting methods? Wont they simply jack up their CPM demands for such placements, thereby nullifying the expected benefits? These are the questions I would ask regarding trying to create more "meaningful" media buying metrics.
Your story is interesting, also in the context of Vice's plea last year to end the discriminatory practice of "blocklisting" key words:
https://www.mediapost.com/publications/article/353018/will-media-buyers-heed-vices-plea-about-racial-bl.html
I agree with Ed.
There seems to have been a transmogrification of the term "targeting", or at least a misunderstanding.
Of late, "targeting" seems to be thought of as those, and only those, to see the ad and everyone else is wastage. The way it works is (i) because no-one can afford to target everyone (ii) the marketer defines the "sweet spot" (iii) the media buyer then targets that "sweet spot".
For example, a CPG marketer's researcher profiling users of their brand may find that 70% of the purchases are made by females aged 25-44 with at least one child at home, though they only constitute 25% of the population (all numbers made up purely for explanation". So it makes sense to ensure that they see the ad for the product.
It does NOT mean that ONLY females aged 25-44 with at least one child at home are to see the ad. It's just that in a broadcast medium there are no media properties that can guarantee that only females aged 25-44 with at least one child at home are served the ad. It is a way of ensuring that the ad is in content that will be the most effective and efficient.
Once outside the world of broadcasting you CAN specify what audience should be served the ad digitally. As long as the site/app has the necessary informastion. And that the data they hold is true. But I wonder why I still get ads for Suunto watchess when I bought mine over six years ago. I'm also not sure why I am a target for infant formula.
But beyond that if you target just your "sweet spot" current users you are not getting yur brand out to non-users which is where brand growth is most effective (i.e. new income streams). If in doubt, read Prof. Byron Sharp's "How Brands Grow".
"Transmorgrification" between media and advertising targeting indeed! Apparently practitioners in media and advertising, at least in the US, have a real issue with using and worse understanding the defintions and therfore the meaning of the terms they use. (How many definitions and interpretations of "impressions" are there in the industry???)
Per John's example, part of the value of traditional media versus social digital media is that they will always deliver audiences outside of the "sweet spot" target group selected for the the brand's campaign. At least some of those "outsiders" exposed will be potential category users or probables or even brand convertibles/switchers. Bonus!