We chat to Dr. Augustine Fou who is an expert in Digital Fraud, having spent the last 25 years studying and documenting the nexus of cybercrime and ad fraud. We speak to him about his Forbes article that looks at what happened when giants like P&G, Chase, Uber and eBay turned off their spending on digital ads... and saw ZERO impact.
We also ask him the question, should brands STOP spending money on digital ads?
Hi Augustine, to kick us off you could intro to people listening about what you do and what you found when you conducted your study about what happened when big brands stopped spending on digital ads?
Dr. Augustine Fou: I’m Augustine Fou, I’m an Ad Fraud researcher and a digital marketer of more than 25 years. I started out in digital marketing in 1995/96 in Silicon Alley - the name for it in New York - and I’ve been the entire arc of the evolution of digital marketing, from the early days when we had to actually convince clients they needed a website. So we’ve come a long way in the last 25 years to now, where pretty much everything, most dollars, are spent through digital compared to traditional channels.
Recently I’ve focused in on the problem of Ad Fraud, because when we started automating the buying and selling of the ads - programmatic exchanges - it also allowed the bad guys to automate the fraud and to scale the fraud. So previously, when marketers bought ads from real publishers there was someone you could negotiate with sitting across the table from you, but now the you’re buying through programmatic exchanges you basically give them a chunk of money and tell the to spend it all, right? And they look for all the different websites and apps where the ads can go, so because of that we’ve seen a dramatic rise in ad fraud in the last ten years. So as a digital marketer we have to actually solve that because when we have a lot of bot activity it’s messing up all of our analytics. Because the bots are generating more impressions, more traffic, more clicks, the analytics that we see are no longer reliable, so then we’re not making good business decisions based on faked analytics.
Do you have an idea of the rough percentage the total fraud is?
It’s a number that no one wants to hear. So, you’ll see that the industry trade associations have put out estimates over the years, but I always find fault with those estimates because they’re projecting from a very small sample set. They definitely don’t see the entire universe and they’re also relying on the technology of specific fraud detection tech companies and because of that some of those companies don’t properly disclose the amount of impressions that cannot be measured. But for the impressions that can’t be measured you can’t just assume they’ll be fine.
So based on my own data, I’m seeing very large amounts of fraud compared to the 1-3% that the industry typically represents. But even that figure - 1-3% - is multiple billons of dollars, because in the US for example, we’re spending $150 billion in digital marketing every year, and worldwide it’s approaching $400 billion. So even a small percentage of that represents billions of dollars that are going into the pockets of bad guys and not towards showing ads to real humans.
Marketers that are spending this money are not going to get any impact from it.
But if the brands aren’t seeing any return, why do they keep spending money on it?
Well they see returns, but I usually use the correlation versus causation; they happen to see sales happening while they are running digital marketing. It doesn’t mean that the digital marketing drove those sales.
So for some of the larges brands, and I would categorise these as the CPG companies that are selling soup and soda in grocery stores, the way they use digital marketing is for reach and frequency. Similar to the way they use TV, right? Billboard advertising, things like that.
When they do that they say they don’t need performance metrics because they don’t sell anything online. They simply use it for branding and to make people more aware. So they go about buying as many ad impressions as possible, and they think they’re doing better by buying cheaper impressions. They think they’re saving money by buying lower CPM inventory. But those low cost ads coming through programmatic exchanges, a large portion of those are coming from fake websites. The reason these fake websites can sell at such a low cost is because they don’t have writers, they don’t have journalists, they don’t have editors. Their job is not to create real content for real humans to read, their job is solely to load as many ad impressions as possible.
So the marketers are essentially using a reach and frequency model, and they think it’s working because they still see sales, but very few of them have done the correlation. Did the digital marketing actually drive the sales or would the sales have occurred anyway in the absence of digital marketing?
In the last couple of years there has been only a few documented examples where the marketers actually turned off their digital spend. So, when P&G turned off $200 million of their digital spend, they saw no change in business outcomes. When Chase reduced their programmatic reach from 400,000 websites showing their ads to just 5,00 websites showing their ads - that’s a 99% decrease in reach - they saw no change in business outcomes.
But very few of these advertisers are willing to admit that their digital marketing wasn’t working, so there are very few of these examples to cite in the recent couple of years.
If more marketers did their homework and really looked into whether their digital marketing was actually driving any incremental business outcomes, they would think otherwise. Right now not enough of them are doing this, so they’re continuing to spend as that’s the thing that they’ve done before.
There are parallels with data in the sense of lead generation where people concentrate on the number of leads they get when they should be concentrating on the number of sales made from those leads. But they can’t get off that hamster wheel.
That’s exactly right. That’s a great analogy. It’s a hamster wheel that once they’re on… they don’t wanna be the first to jump off.
The bottom line is, they’ve been overly focused on quantity metrics or vanity metrics because those are easy to gather and easy to report on.
Tactics and tips served straight to your inbox. Sign up to the Databowl newsletter and receive weekly custom content for unlocking growth.