This is why I don't especially worry about fake influences, bogus ad clicks, etc. Any advertiser worth their salt should be able to see the results (or lack of results) very quickly. Sure there'll be dumb brands throwing money away on fake clicks/likes/re-tweets/etc and not measuring results.
And while that is regrettable & wouldn't happen in an ideal world it's their money to waste foolishly.
Attribution is tricky at the best of times and worse when you're dealing with many, small campaigns rather than a few discrete large campaigns.
Let's say you make surfboards that are primarily bought offline at surf shops. You run a campaign with 100 instagram influencers in the San Diego area and you see a 15% greater uplift in sales compared to nationwide. Which influencers contributed meaningfully and which ones were fake accounts where you wasted your money? How much of that 15% uplift is because of the influencer campaign vs a longer summer in San Diego this year vs some totally organic consumer spreading it via WOM?
Even if you want to design an experiment, summer's already over and you'll need to wait until next summer to test out your hypothesis. It's not hard to see how a scammer could cobble together a dozen or so of these different advertisers and make a substantial sum out of fraud consistently without ever getting caught.
It's one thing when entering a new market. You expect your sales in Italy to go some zero to something, and if it doesn't you know your ads aren't working. It's harder to tell the difference if you already have large sales, unless you can connect an individual sale to the ad that drove it.
