The rise of digital media offered the promise of paying for advertising based on definable, measurable impressions.
Finally, decades of basing TV and radio advertising on limited survey models were coming to an end.
While surveys left advertisers with no insight into who was actually seeing their ads, the new digital advertising environment allowed advertisers to track impressions using signals like third-party cookies. Naturally, the pricing mechanism for this new model would be calculated on an impression basis, à la the CPM.
It seemed like a great idea at the time. What could go wrong?
Research shows we’re seeing the worst possible outcomes with the CPM-based approach. Major brands are monetizing CSAM content. Meanwhile, they are also sending ads to known bots or missing their target audience altogether.
At the root of this catastrophe is a combination of worthless media, bad cookie-based audience data and the flawed currency of CPMs.
And with Google’s recent decision to hang on to cookies indefinitely, we’re in danger of perpetuating the worst parts of the digital ad business.
How we got here
I remember sitting in a tiny office with Brian O’Kelley as he whiteboarded his plans for AppNexus. He explained how AppNexus would become the New York Stock Exchange for digital advertising.
However, there was one fundamental difference: Companies on the NYSE are highly regulated, with extreme scrutiny on shares. They can’t just create and sell fictitious shares to drive revenue. Ad tech, on the other hand, had no problem creating fictitious audiences and impressions.
Everyone in the advertising ecosystem quickly learned there were only two meaningful ways to grow revenue within the industry’s existing model:
- Increase CPMs
- Increase impressions
The promise of a data-driven utopia quickly dissolved into the Wild West.
AdExchanger Daily
Get our editors’ roundup delivered to your inbox every weekday.
Daily Roundup
The worst part was that advertisers were the last ones to realize what was happening. Many still don’t understand the risks.
Publishers and tech providers have tried in vain to increase CPMs by focusing on the quality of the audience. This was naive. Agencies didn’t have the time, interest, skills or patience to figure out how to measure quality and make that measurement work for their clients.
Plus, agencies didn’t have to figure it out. They had a built-in way to easily grow their revenue:
They just needed to convince the advertiser to buy more impressions, without explaining why such massive impression counts were important.
And so began the battle of scale versus quality.
A losing battle
As agencies convinced advertisers to focus on scale, DSPs built algorithms to reward the lowest CPMs. Data providers responded in kind, offering up low-quality audience data at scale.
With the focus on scale and zero accountability at the impression level, a multibillion-dollar industry became hooked on the delivery of meaningless impressions and outright fraud.
Here’s but one example. The “auto intender” audience segment was originally created with only Autotrader data, and it worked like a charm. Why? A consumer on Autotrader.com is probably thinking about buying a car. This new segment started selling like hotcakes because it worked.
However, agencies needed more scale because they didn’t want to explain to auto companies how limited this high-quality audience was, much less the fact that brands would have to pay higher CPMs to reach it.
Creating auto-intender audience scale was the easier path to revenue. Agencies demanded more, and data and audience companies were happy to comply. Any visitor to any inventory that had anything remotely to do with motor vehicles was classified as an “auto intender.”
There are roughly seven to nine million people in the US who are in the market for a car at any given moment. Yet, if you go into any DSP and look at “auto intenders,” you’ll find exponentially more consumer profiles available. The math doesn’t work, yet no one questions it.
This thinking supports worthless content with ad dollars. Meanwhile, legitimate publishers are hurting for revenue.
Time for a “flight to quality”
The ad industry has spent much of the past five years wondering what happens when third-party cookies are no longer the primary mechanism for anonymously recognizing and targeting consumers. Despite overwhelming evidence that ad targeting has fallen short of its initial promise, many are breathing a sigh of relief following Google’s recent cookie reprieve.
The fact that so many advertisers are still hooked on cookies showcases the larger issue within ad tech. The process of buying and selling media and data by the ton simply doesn’t work. It’s a massive waste of money wrought with disingenuous players and fraud. And it’s destroying the legitimate publishing industry.
Yet now we have an industry seemingly celebrating its commitment to this paradigm.
To remain viable, the industry must reset its expectations around scale and quality. It is time for agencies and their brand partners to start demanding accountability from ad tech and reward the publishers and data providers that have authenticated audiences consuming quality content.
Scale for the sake of scale has never been the answer. It won’t be easy, but let’s finally build the industry we were promised.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow B2B Tech Group and AdExchanger on LinkedIn.
For more articles featuring Mark Dye, click here.
