When you analyze the effects of fraud, viewability and ad blocking on the digital display advertising business, then add the ever-increasing abilities of the traffic launderers to game the system, you reach an inevitable conclusion: ad tech has evolved into a toxic ecosystem that is killing itself, and it is taking digital advertising with it.
How We Got Here
In the golden age of digital, Google SEO ruled the web. Back then, you wanted the best possible placement for your organic search results. This meant following Google’s rules for almost every aspect of your online architecture. That was then.
Today, clickbait and recommendation platforms rule the web. The consequences of this transition are unfortunate. Publishers are economically motivated to use any means to maximize page views. This has spawned a plague of traffic brokers who specialize in traffic laundering at an almost unbelievable scale. It has also caused publishers to devolve user interfaces into the emotionally unsatisfying experiences they are today, motivating users to deploy content blockers, motivating publishers to create countermeasures, elevating the arms race to the next level, resulting in a vortex that is sucking the entire industry into oblivion.
At its core, ad tech’s sorry state is a function of scale and misguided media procurement strategies. Publishers need more traffic than they can realistically generate because their “real” traffic is undervalued or, some would say, indistinguishable from fraudulent traffic. If you need to take an order for 20 million impressions and you only have 2 million to sell, you will use a traffic broker to procure the additional 18 million impressions – everyone is doing it, even if they think they aren’t. Fraud is out of control.
A regulatory revolt is on the horizon. Things are going to change, whether they change because of a government agency or civil lawsuits or simply an economic crackdown. Too much money is being wasted in too blatant a way for this fraudulent behavior to continue unchecked.
I don’t want to speculate as to the method of ad tech’s demise. If I had to guess, I would posit a reinvention of ad tech with regulatory oversight, rather than its total destruction. But instead of worrying about where traditional ad tech may end up, I’d like to explore a possible future for the evolution of consumer messaging via the Internet.
Business Outcomes Are Better than Impressions
How can small and midsized publishers compete with super-scaled mega-sites without resorting to traffic laundering? By turning the business model right-side up again.
No business wants to purchase a CPM or a GRP or an impression. Given a choice, most businesses would prefer to purchase a business outcome. You could argue that the system already does this. It does not. We have developed sophisticated proxies that are optimized to correlate to business outcomes, but we (the industry) sell impressions. What if we actually sold business outcomes?
I’m not talking about direct response advertising. It already exists, and it doesn’t work for brand advertising or nonimmediate calls to action such as upcoming sales events. I am suggesting that through the use of data scientific research and machine-learning tools, we are on the cusp of understanding how to generate leads that will result in acquired customers and directly drive business outcomes. The goal is not new, but awesome new tools are just becoming commercialized.
If you watch a video entitled “Unboxing the Samsung Galaxy S7” four times, there’s a pretty good chance that you are in the market for a new smartphone. This is well understood. Current ad tech tools relentlessly retarget you to the point where you cannot visit another website for days without seeing display ads for an S7 or a competing brand that is willing to pay more money than Samsung to get in front of a person who is “in market” for a new device. There is a better way.
There are new tools emerging that enable publishers to combine common measurements for content adjacency, context, and behavioral targeting with pattern-matching algorithms to rank users into “in market” clusters (as opposed to targeting specific people). There are several programmatic creative tools capable of putting the right message in front of the clusters. The machine learning algorithms can continuously refine the clusters and, in many cases, can identify attributable paths to purchase. The result is the ability to guarantee a business outcome at a price. If publishers believe that advertising works, they can prove it by assuming some (or all) of the risk for a greater reward.
This may look like an iterative step up from bounties, and it may be. But the capabilities of the toolsets are increasing at an exponential rate.
Adtopia might look like this: platforms continue to sell targeted impressions (but would eventually offer business outcomes). Mega-sites get regulated into unprofitability. Quality publishers, realizing that they can never compete with platforms or mega-sites, evolve super-sophisticated data-driven business outcome capabilities and enjoy a renaissance where erudition (the filter set of choice) leads to guaranteed business outcomes for the highest-paying clients.
It won’t happen this way. There is no Adtopia, and ad tech will be with us in its current form until someone goes to jail. That said, things have got to change, and data science can lead the way.