Amid the scramble about what to do when Google does away with third party (3P) cookies in Chrome (the browser with 2/3 market share) and Apple does away with access to device identifiers on iOS, many of the same questions keep popping up from marketers — does it impair the ability to target individuals, does it increase ad fraud, does it make reporting on campaign outcomes impossible, etc. In response, there’s much misinformation swirling about work-arounds like fingerprinting, targeting FLoC’s of consumers (cohorts grouped by blackbox mechanisms), and the rise of ad fraud. But let’s take a closer look and dispel some of the disinformation being spread by ad tech companies to protect their current income streams.
The moves by Google and Apple increase privacy for consumers. That’s a good thing, after years of wanton data collection by ad tech companies, violating consumers’ privacy without their knowledge, consent or recourse. Ad tech companies are all crying foul — that Google and Apple are abusing their monopolistic powers. That is true. But no other entities could force the changes necessary to actually increase the privacy of consumers. This increase of privacy is relative, not absolute, because Google still has all the data about its users, who are constantly logged into free services like Gmail, YouTube, Android devices, etc. and Apple also knows everything about its users.
Identity and Ad Targeting
Ad tech companies are arguing that the loss of 3P cookies is a bad thing because that means the loss of the ability to target ads down to the level of the individual. What they are really saying is that it impairs their ability to make revenue selling services based on 3P cookies, like audience segments, behavioral targeting, etc. Marketers should consider whether targeting down to the level of the individual is actually worth it or actually works in the first place, given how poor the data quality is. (See: How Accurate Is Programmatic Ad Targeting?)
Too many marketers are paying extra to “chase people around the internet” with retargeted ads – ads that are shown to users after they visited your site. And most consumers know how crappy that is — e.g. they already bought the item and then they see tons of ads for the exact same item. Too many marketers are paying extra for hyper targeting — using many targeting parameters, thinking that it makes the ads more relevant for users. Research has shown that it doesn’t. (See: Despite Claims That More Targeting Means More Relevant Ads, Nope. Here’s Proof.) In fact, consumers are seeing far too many ads that are not relevant that they have taken action to protect themselves – by installing ad blockers.
Over-Targeting and Campaign Effectiveness
Advertisers have gone too far down the rabbit hole of targeting in digital advertising. They have been misled by ad tech companies because the more targeting parameters they buy, the more revenue the ad tech companies make. Advertisers are over-targeting individuals and under-spending in basic brand building — exposing new potential customers to their brand and product. By doing so they are “leaving large sums of money on the table.” To illustrate with a more specific example — pharmaceutical companies have been desperately trying to target the 1% “high-prescribers” — doctors who already prescribe a lot of their particular drug. These advertisers are neglecting showing ads to the other 99% of doctors who are “low- to no-prescribers.” This much larger group may simply not be aware of the pharma company or their drug. If the pharma advertisers rebalanced their digital ad spending to include more brand-building and less over-targeting, they could drive a lot more sales. Small increases from a group that is 99X larger than the “high-prescribers” group far outweighs eking out a few more prescriptions from doctors who already prescribe a lot.
Furthermore, most marketers don’t need to know which specific individual bought something from them; they just need to know that more people bought something as a result of being exposed to the ads. This means that 3P cookies, which represent the individual are not necessary in the determination of whether campaigns worked. As long as advertisers can see “exposed” versus “not-exposed” and that exposed users exhibited higher rates of purchases or business outcomes, they know their digital campaigns were effective. Even performance marketers don’t need to know WHO installed their apps, they just need to know THAT users exposed to ads actually ended up installing their app. This means doing away with 3P cookies or Apple’s device identifiers won’t make campaigns less effective and won’t make them less measurable either.
Doing away with 3P cookies will also NOT increase ad fraud. That’s because ad fraud is already very rampant and severely under-reported by the fraud detection companies. (See: What Your Fraud Detection Vendor Misses). Furthermore, the fraud detection tech companies that rely on 3P cookies to do their job have already lost before they begin. These fraud detection vendors rely on cookies to mark the bots that they have already detected; this way, when they see the same cookie again, they know it’s a bot. The problem with this approach is that bots dump cookies and get new ones, just like humans clear cookies from their browsers. If the bot detection tech vendor hasn’t seen a specific cookie before, they let it through — i.e. they let the bots through. These fraud detection tech vendors are utterly useless in doing what they purport to do – IVT detection. Advertisers who paid millions of dollars for these services didn’t get what they paid for and should sue these vendors for selling them services that they didn’t deliver on.
Further, with 3P cookies, bots can pretend to be any audience segment that advertisers want to buy for ad targeting purposes. For example, bots will deliberately visit a collection of medical sites to make themselves appear to be doctors; this way, they earn higher CPMs from pharma companies that pay more to target specialized audiences. Bots can also type search keywords or add items to cart and abandon it in order to trick search and cart abandonment retargeting algorithms to earn more money. Doing away with 3P cookies reduces these specific fraud vectors, so less money will be lost to fraud. Counter-intuitive, right?
Brand Safety and Disinformation
Because advertisers had been led to believe they can target ads to individuals even when they show up on long-tail sites that no one has ever heard of, they allocated more and more of their budgets to programmatic channels. In doing so, more ad budgets went to sites that no one had even heard of, let alone visited. This allowed the proliferation of fake sites designed solely for ad fraud, and disinformation sites, designed to spread false information. These types of sites had no way of making money before. Now, they have ad tech helping them make money via ads so they can thrive and continue to expand. This enables them to create more fake news and disinformation and spread it further. (See: research studies by the Global Disinformation Index)
Big brands are inadvertently funding fake news, coronavirus misinformation, and the disinformation campaigns of foreign state actors. Even though they pay for “brand safety detection” those technologies have actually made the situation even worse. They are blocking ads on mainstream news sites, thus defunding legitimate news sources, while not catching and blocking ad revenue from going to fake news and disinfo sites. (See: We’ve Known Brand Safety Tech Was Bad—Here’s How Badly It Defunds The News). Advertisers who paid for brand safety detection didn’t get what they paid for; they should sue these vendors for selling them snake oil.
But, all of these problems, and the fees paid to brand safety and fraud detection vendors, could have been avoided entirely in the first place if advertisers didn’t spend so much on cheap ads on long tail sites through programmatic channels. If advertisers bought smaller numbers of ads from real publishers with real human audiences, they would have done better digital marketing for the last decade.
Cost Savings and Better Outcomes
In addition to the cost savings from doing away with fraud and brand safety detection vendors (that don’t work), doing away with 3P cookies also means advertisers have a unique opportunity to undo the bad digital marketing habits of the last decade, paying for the snake oil that was sold to them by ad tech vendors.
With 3P cookies, advertisers were misled to believe that more targeting means more relevant ads and more better business outcomes. But despite all of the hype fomented by ad tech companies about how magical targeting would be, when big brands turned off or paused their digital ad spend, nothing happened. Showing ads on 400,000 websites yielded no better outcomes for Chase than showing ads on 5,000 websites (99% less). Perhaps there were not many humans on those 395,000 other sites; perhaps the ad targeting was not all that it was cracked up to be. Perhaps the vast majority of the “users” on the vast majority of long-tail sites were bots; those sites had little to no humans that visited; so they purchased all their traffic. Perhaps the targeting based on inferred audience segments were worse than no targeting at all.
The “Cookieless Future” is Bright and Grim
When all of the useless and harmful things built on 3P cookies go away, the associated costs go down and the effective business outcomes from digital marketing go back up. More specifically, when advertisers buy ads from real publishers with real human audiences, they will get better outcomes than hyper targeted ads shown on millions of long tail sites to bots, pretending to be various audiences. This will also reduce the need for privacy-invasive data collection and the extra costs of targeting parameters, audience segments, fraud detection, brand safety detection, etc.
The “cookieless future” is indeed bright — for advertisers, good publishers, and consumers. But it will be grim indeed for ad tech vendors that had been profiting off of privacy-invasive data collection from users, diverting ad revenue away from good publishers to themselves, and selling snake oil to advertisers.