Commentary
We Need Nielsen's Help In CTV -- Now
- by Dave Morgan , Featured Contributor, December 17, 2020
This morning, The Wall Street Journal and Digital News Daily reported on what Oracle and its digital analytic division Moat have dubbed “StreamScam,” its detection that 28 million CTV “viewing” households over the past few months were fraudulently “spoofed”: they didn’t exist, despite being recorded by ad servers as having viewed the better part of a billion premium-priced ads over that time period.
As reported, the simulated devices “took advantage of flaws in streaming-TV ad-serving technology and the supply chain to fool marketers into paying for ads that were never actually seen by viewers on real devices and apps.”
Spoofed device scams have been dogging the Internet ad industry for the past 15 years to the tune of tens of billions of dollars over that time. So no one should be surprised this technique has been exported to the TV ad industry now that it's adopting the same vulnerable programmatic ad-serving and supply-chain technologies that have managed digital banner ads since the mid-2000s.
advertisement
advertisement
Where is the recognition and remorse from buyers and sellers wasting so much of their advertiser clients’ money? Ad budgets are finite, which means that this money is not going to legitimate video media companies -- the same ones we read about having to lay off thousands of their employees because of declining revenues.
No one can argue that these “budget shifts” are due to changing consumer behavior. No, this is all about growing criminal behavior, and the failure of an industry to stop it -- probably because half of the players and platforms are able to make some money on the transactions, fraudulent or not. And the other half don’t want to seem like Luddites, railing against technology-born changes in a time when adhering to the old means first up on the layoff list.
No one can fall back on the pat answers that companies like Google used in the past when trying to dismiss bid fraud in the search world, arguing that since advertisers were buying search and contextual ads based on performance and were tracking the click performance, that any fraud in the bidding was baked into the price they were willing to bid and pay.
No, the CTV campaigns here were sold on a cost-per-impression basis, and a premium one at that. Each and every impression carries a cost, and they were valued on a comparative cost to trying to reach those same audiences on linear TV. There was no “fraud fluff” baked into those prices.
If a television company sold ad campaigns with 28 million fraudulent households, there would be investigations, maybe even congressional hearings, and executives would go to jail. Don’t forget the amount of swirl created a number of years ago by a single “wardrobe malfunction.”
But this can’t happen in linear television. Why? Because campaigns are bought, sold and measured against an independent, verified panel of actual human viewers to insure that paid-for campaigns viewership correlates to actual human viewership. As small as that panel is (and it is too small), you can’t drive a 28-million household fraud through it.
From what I understand of the recently unveiled Nielsen One cross-channel video measurement initiative, it should be able to do the same for CTV. I know it will take several years to get everyone on board for using if for all video investment needs. But please, can we have the CTV elements accelerated?
The money that went to fraudsters here should have gone to legitimate ad deliveries on streaming services owned by CNN, Discovery, ViacomCBS and Disney and many others. Those companies have been laying off thousands of employees because of a loss of revenues.
If we don’t move soon, we will continue to see connected TV ad fraud stories juxtaposed with stories about layoffs at those companies.
We owe it to those hard-working TV colleagues to stop the CTV fraud now. Give all CTV streaming providers a level playing field.
What do you think?
But Dave, who is going to pay for such a Nielsen initiative? Certainly not the advertisers and only a bit will be gleaned from the ad agencies/buyers. So it's up to the ad sellers to get the ball rolling in some concerted manner--and they must be willing to sign firm contracts, not just make promises. Even if this type of joint action among the leading sellers were to take place, we are priobably talking years to get going---first with groups of sellers and agencies going over the specs with Nielsen, then with some programming work, next, trial data and a look at the output, then more discussions, followed by more prigram work, etc. etc. When Nielsen started meauring TV the three major networks led the way and they were joined by advertisers like P&G as well as the larger agencies---all trying to develop a fair and objective product. Are there any signs of this ytpe of approach even being discussed in the CTV world?
Great piece Dave - the harsh reality is that buyers must be more strict about only buying from partners that are truly implementing solutions that do something about fraud pre-bid. Perhaps select a partner that offers an MRC approved pre-bid fraud solution as well as other tech to root out these bad actors and avoid buying these impressions pre-bid.
So many buyers and sellers in CTV seem to think that buying from a premum publisher direct or through a PMP helps them avoid fraud. This is simply not true either.
There are unfortunately many companies claming to have fraud solutions in place (some even with self-attested fraud tag certificates) that are doing little or nothing. Having been in this ad market for over 20 years, I like you ,am shocked that so many turn a blind eye or worse just count fraud as a cost of doing business. This insanity must stop!
Bravo Dave.
There is another vector to this slippery slope. Advertisers were (and are) hot on driving down media agency fees. Many good buyers were lost or purged by automation (which I am all for). But what was lost was oversight.
Here in Australia media agency fees were typically 3% of spend and that was driven down to 2.5% or lower (some desperates were at 1.5%). I also note thet the creative fee is, and remains, substantially higher.
As audiences splinter yoiu need to buy more spots across more vendors and more devices and media channels. It's a classic more work less pay scenario, so automation was an obvious solution.
But as predicted, automation of increased spots reduced oversight. One wonders just how much that reduction in service fee has cost advertisers.
Ed, yes. The video ad sellers are going to need to step up here, and push for an acceleration of Nielsen's hybrid panel and connected TV ACR measurement. Maybe eliminating the 20-40% of programamntic CTV ad fraud, which will mean more money for those with verified measurement, will get folks buying into Nielsen One early.
Thanks Gabe. I agree. We need MRC certified pre-bid verification, and also the need for buyers to really care about what they're getting. The vast majority of the top streaming sevices today (6 of top 7) carry no ads, and those apps represent 80+% of all streaming viewing. So, buyers needs to start with a real skepticism of quality and legitimacy when they make their CTV buys. Hopefully, we will see more of them care more than they do now.
Thanks John, you hit a really important point right on the head. Advertisers have been so focused on driving down the cost of media and the cost of procuruing and managing it that their buyers can't give campaigns, vendors and systems the scrutiny and oversight that is needed. Fixing this will start with making advertisers care.
When will the cost of doing the right thing be seen as a way to keep much much more money from going to the fraudsters? Advertisers continue to lose hundreds of millions rather than the much smaller amount it would cost. In the end, adverrtisers pay either way. Maybe the government will step in like it does for other fraudulent trade.
Jack, it's certainly a conundrum. I'm not a big fan of regulation, but I bet if some government agencies got involved through prosecution - there is certainly criminal behavior here - it could mae a real difference. No advertiser or agency or programmatic platform wants to be called before a grand jury. Pretty tough to explain to their bosses or owners, let alone the legal bills they will incur in the process.