Just so we're all on the same page, ads.txt allows a publisher to list whom it allows to sell its inventory. This can cut down greatly on rogue operators because, a little like a nightclub bouncer, if your name isn't on the list, you're not coming in. The problem is, of course, that if you have a system in place, you never really know how good it is because it's there protecting you from what you're trying to avoid.
That is, of course, unless you're brave enough to run an experiment and take out the ads.txt protection. That is exactly what The Guardian did, so hats off to the site for being brave enough to stare into the abyss.
With ads.txt on, the site reports there were no discrepancies and all budget spent by advertisers resulted in proper placements. However, with the protection not in place, the amount of ad fraud is astonishing.
On the one hand, you have just 1% of display being sold and placed by rogue programmatic operators. That doesn't seem too bad, does it? Well, that statistic climbs up to a massive 72% of video inventory going through fraudsters. That's right -- nearly three in four video ads were targetted by fraudsters.
So why video, and not normal display? The simple answer lies in the old cheeky Londoner's phrase that if you're going to play the system, you might as well be hanged for stealing a sheep than a lamb.
Video is worth much more, so if you're out there acting fraudulently, why wouldn't you target the type of ad that gives you the biggest return?
The scary thing is that only around four in five publishers are believed to have adopted ads.txt so far. That is a huge increase from the early number of adopters because take-up was pretty low soon after launch. However, even now, a fifth of publishers are still leaving themselves wide open to have unscrupulous programmatic operators plying their fraudulent trade.
So the next time ads.txt is discussed -- with the pros and cons of whether it achieves everything it sets out to -- just remember that without it, nearly three in four video ads were targeted by fraudsters in an experiment by The Guardian.
It's not a perfect system, but it's a very simple tool to give sites a lot of protection they wouldn't have otherwise. From today we have the killer stat to back up why it's so important that its rollout is universal.