debate on IDFAs going away
This weekend saw a vigorous debate on Twitter about the effects of Apple doing away with unfettered access to device-level identifiers - IDFA (“Identifier for Advertising”), and forcing all iOS app makers to get permission to collect and use IDFAs from users. Privacy advocates cheered the move because it improves privacy for consumers. It does. Ad tech advocates see it as bad for ad tech companies selling data and targeting services (it is); but claim instead that it’s bad for consumers because they will see less relevant content and ads (classic misdirection).
Over the years, the same ad tech companies convinced advertisers to buy data and microtargeting services in order to create “better ad experiences” for consumers; lately, they’ve been trying to scare marketers from stopping said “surveillance marketing” by citing “deprecation of content and ad relevancy.” But are ads really going to be less relevant once device identifiers and third-party cookies go away; or is this the last ditch effort of ad tech companies and their funding sources to remain relevant themselves? Let’s unpack this.
Digital ads are not relevant now, despite all the ad tech
Many of the responses in the Twitter thread point out what most consumers already know — digital ads are not relevant right now, despite all of the wanton data collection and microtargeting. If anything, some ads are downright creepy — like the ones that appear to follow you around the internet, after you looked at a specific item on Amazon, or the ones that keep following you even after you already purchased the item. To say that these two examples are due to the limits of the technology would be the kind thing to do. More accurately, it’s because the technology doesn’t work as well as what the marketing materials and press releases represent. Ads are so not relevant and so annoying that more and more consumers have started using ad blockers to protect themselves. “The Advertising Industry Has a Problem: People Hate Ads,” writes the New York Times. Ad blockers also block the trackers that ad tech firms use to collect data for ad targeting purposes.
Do you know who doesn’t block ad and ad trackers? Right, bots. It is their job to cause ads to load, and to make more money by pretending to be attractive audience segments that marketers want to pay more to target. So bots don’t block ads or trackers; in fact, they collect third-party cookies in order to pretend to be in specific audience segments. So advertisers spending money in programmatic channels are showing ads disproportionately to bots; humans are not in these audience segments because they blocked the ads, trackers, and cookies. This is an inconvenient truth the ad tech companies don’t want marketers to find out.
Why is programmatic targeting so bad?
Even if the ads are shown to humans who didn’t block ads, they are not as relevant as ad tech companies represent, or marketers are led to believe. Why? It’s because the targeting is based on inferences — insights derived from website visitation patterns. This is called “behavioral targeting.” By collecting data on what sites users visit, ad tech companies purport that they can tell who they are and what they like — so they can better target ads at them. But consider the following: while it is possible to guesstimate that users are male when they visit sites like ESPN, Sports Illustrated, Maxim, Playboy, etc. of that users are female when they visit sites like Victoria’s Secret or feminine hygiene products, what do you derive about the user when they visit sites like CNN, New York Times, Weather.com, etc.?
As you can surmise, inferring “who they are and what they like” based on what websites they visit is rudimentary at best, and not very accurate in reality. This is not just an opinion. The studies that have been done to investigate this shows just how bad the derived parameters are. (See “How Effective Is Third-Party Consumer Profiling and Audience Delivery?”) With just one parameter - gender - the data is only 42% accurate. That is less accurate than if you just did “spray and pray” with no targeting at all — i.e. you would have still hit the right gender 50% of the time. With two parameters - gender plus age - the accuracy is down to an average of 24%. Some data brokers were far worse, with single digit percent accuracy. Third party profiling of audiences is so inaccurate, it’s better to save your money and do “spray and pray” instead. Jason Kint, CEO of DCN, a publishers’ trade group, continued “3 things I am certain: 1) you don’t need IDFA to know gender 2) “relevance” due simply to tracking is a myth, and 3) brand desire and demand isn’t built by microtargeting individuals. Quite the opposite.” This is another inconvenient truth the ad tech companies don’t want marketers to find out.
Is it all that bad? Yes.
Ad tech companies believe ad targeting works; and they have sold it to marketers that way. Marketers need to be sure to ignore the industry funded studies, run by measurement companies that support ad tech, that ask survey questions like “do you prefer more relevant ads or less relevant ads?” Of course any rational consumer would select “I prefer more relevant ads.” Marketers need to turn to truly independent research like that from Adalytics. One of their studies on ad relevance showed that not only was there an overabundance of ads (too many ad impressions) but that virtually all of them were not relevant. “Over the course of one month in 2020, I saw a total 4,975 ads across the 35 different websites ... [from] 397 unique advertisement sponsors. I decided 32 of them were relevant... [but these same ads were shown] a total of 666 times.” Much larger data sets and studies are ongoing, but the observations are consistent with what most consumers are experiencing. Ad tech companies think their ads are targeted, but consumers’ experiences are the exact opposite.
Even though most consumers don’t recall a single ad, some consumers do note the “creepy ads” that follow them around the internet. Those happen because of retargeting - targeting ads based on the site you visited or the product you looked at. This is done by setting a third-party cookie in the browser to uniquely identify that event (visit or product view). When that cookie appears in the ad exchange again, they “re-target” the user with an ad from that site or for that product. This happens even after the consumer bought the product. That’s because despite all the surveillance and data collection, the underlying technology didn’t get the data/feedback loop that the consumer already made the purchase. This is an inconvenient truth the ad tech companies don’t want marketers to find out; but consumers know anyway due to everyday experience.
What about ad measurement? It doesn’t need unique device identifiers
The next line of argument ad tech firms use to scare marketers is that they will lose the ability to measure the effectiveness of their digital marketing when 3P cookies and device identifiers go away. But literally give that a moment’s thought and you will realize that is not true. Marketers want to know if their digital marketing campaigns worked. But marketers don’t need to know which unique device or user made the purchase to know whether their digital campaigns worked. Marketers just need to know that, collectively, those users/devices that were exposed to their ads, made purchases at a higher rate than those who were not exposed to ads — i.e. incrementality. Marketers need to know that people bought more after seeing ads, compared to those who didn’t see their ads. Unique device identifiers are not needed to do ad measurement. This is an inconvenient truth the ad tech companies don’t want marketers to find out.
So What?
To summarize, the enforcement of privacy regulations is happening now. Big companies like Google and Apple are making their own moves to comply with privacy regulations. Google is doing away with third-party cookies in Chrome, the browser with the largest share of the market at 2/3. Apple is requiring app makers to gather explicit, user level consent in order to get the device identifiers.
Marketers need to take more urgent and aggressive action to shore up their own compliance; and refuse to be misled by ad tech companies selling data, micro targeting, and new workarounds to privacy. Don’t go down with the sinking ship of ad tech, still desperately trying to preserve the “surveillance marketing state” they created over the last 10 years to extract profit.
Smart marketers avoid microtargeting and the associated data, for all the reasons above. Smart marketers think of targeting as the avoidance of waste — a beard trimmer ad doesn’t need to be shown to women and kids. Just a handful of simple parameters will help avoid most of the waste; microtargeting at scale is a fiction. Using a few targeting parameters, and kicking microtargeting to the curb, is a far simpler, and more productive, way to do digital marketing. You won’t need any of the snake oil sold by ad tech and the investment bankers who bankroll them.
P.S. “relevancy” is not a word. It’s “relevance.”
