Sounds like we need better bots. Ones that don't move from point to point in a straight line, ones that read a description at between 1.6 and 4.8 pages per second, ones that stay on a landing page for 1.8+d100 seconds. Why aren't people using better bots?

Yes, that's all sarcasm. But expect an increase in more human acting bots to be the solution offered by someone and not less bots. Because someone wants to keep making money off fake traffic.

The article only alludes to answering the question "Who is engineering this?"

It's the ad platforms. They're doing it. They're selling ads, then they're also using bots to pump up the impression numbers, so that they get paid more.

This is fraud.

Marketing doesn't work nearly as well as marketers would like you to believe, so much so that they have to fake their effectiveness.

Sounds like another bubble ripe for bursting.

How old is this article? Is it a reprint from years ago? Because he talks about suddenly pivoting his data by “conversion rate” and having an AHA moment about how we measure success.

Except… conversion rate is a bone-standard, absolutely ubiquitous way to measure traffic quality in ecommerce. No one places ads without knowing how many of them lead to conversions. Defining your conversion event is often part of setting up an ad in the first place.

He then goes on to describe his hand-rolled script that analyses mouse movements to differentiate humans from bots.

Except… that’s exactly what the “I am human” checkbox from CloudFlare and Google have been doing for years.

CloudFlare have said that about 30% of Internet traffic is bots. This is well known. It could easily be 70% for some sites.

I would say that there’s nothing to see here, but it’s probably a little worse than that: just adding some really shaky analysis and anecdotal data to an already widely-covered topic. Are we actually going to trust an internet marketer’s hand-rolled mouse movements analysis over CloudFlare?

I’m not.