How do you test for ad effectiveness vs annoyance? Especially so for a captive audience where they can’t leave and go elsewhere? It seems like every market leader that gets ads eventually “optimises” towards making them look like not ads. Obviously they will be more effective if people don’t realise what they are, so how do they account for annoyance (and the other negatives a user experiences) while doing these a/b…
In a walled garden like apple? You simply don't, just make the test gradual and long enough until people get used to it.