The optimizers were the crowd in this business model. Some of them made good decisions, others made bad. But since the amount of money available was limited, those bad decisions brought down the performance of the campaign and ultimately the advertiser gave up.

An expert would've made better decisions.

I think the wisdom of the crowd can sometimes be effective, but there are a lot of ways it can go wrong. Some things I was thinking of:

1. the crowd wasn't big enough

2. the time frame was too short to be effective

3. the heuristic was too complex or poorly tuned to lead to a good outcome

4. creative endeavors with high rates of failure lack enough signal to properly optimize (a strategy might look terrible till it suddenly works)

5. misaligned incentives undermine the goal

To illustrate the last case, in high school I had a teacher tell us that he would scale the test by the highest score. For example a 95 would give 100 and +5 to every other score in the class. With incentives like that, if everyone in the class answered no questions, everyone would get 100. (but then all it would take would be one individual to screw the whole class)

Crowds are complex, difficult to understand and hard to predict.