> My impression is (based on what I've seen with music plugins and heard from plugin developers, which tend to be also produced by small shops) that piracy data
does matter a fair bit if you don't do anything about it.
The problem is nobody really knows that because it's nearly impossible to do a controlled experiment for this.
What happens in practice is that a company's first product is unprofitable. So they go back and make significant improvements, spend five times as much on marketing, and add anti-piracy code. Then the new version is more profitable.
The people who want the solution to have been the anti-piracy code credit it with the improvement, but there is no way to know if it was that or the product improvements, or the marketing, or more favorable market conditions at the time of the second launch, or the fact that the same marketing effort now produces 50% more customers because there is now an existing user base (including the pirates) who are easier to convert because they're familiar with the previous version, etc.
To actually know the answer you would need a statistically valid sample of product launches where the determination of whether anti-piracy measures are taken is made at random. But as far as I'm aware no one has ever attempted this, and the choice to use anti-piracy measures has to be randomized or you could trivially be measuring the wrong thing, e.g. larger products with more resources are both more successful and more often include anti-piracy measures but the causation is reversed.
So nobody really knows the answer and it's all just wild speculation.