Once upon a time I was responsible for a high traffic site with ads as a primary revenue source. The constant problem was that competitors would automate click

Once upon a time I was responsible for a high traffic site with ads as a primary revenue source. The constant problem was that competitors would automate clickbomb attacks -- fake traffic with click throughs that would ultimately get the site banned from ad networks (or severely devalue CPM rates) for fraudulent traffic. The irony was that we never used the same tactics ourselves, assuming that the risks outweighed the rewards. The malicious bot traffic ultimately spelled the end of the site. I would imagine that an Amazon product/seller would suffer from high quantities of any fake reviews, whether positive or negative: Too positive? Amazon or Fakespot & similar services penalize them. Too negative? Consumers penalize them. Seems like with a modest budget and a Fiverr-like service, smaller retailers could be harmed with relative ease. Then, faced with these attacks, I'd imagine many companies with means might try countermeasures, triggering a sort of review arms race. How do you separate the signal from the noise from deliberate fake reviews (of any kind) from third parties with malicious intent?