My mate worked at a giant ecommerce company, it would send hundreds of millions of e-mails a month... and with virtually no performance evaluation (only high level stats which are by no means actionable). It took a single SQL query to propose cuts to traffic in the order of 20% with ~zero loss in revenue.
I remember attending a talk, where this guy talked about their freemium app used by dozens of millions of users. Their in-app popups were cut by 30% or so without losing revenue, all thanks to a few simple if statements (they tried machine learning as well, but this did it).
One can only wonder how much of this excessive advertising there is, I guess it's mostly driven by absolute revenue numbers without much consideration for costs and efficiency.
I remember attending a talk, where this guy talked about their freemium app used by dozens of millions of users. Their in-app popups were cut by 30% or so without losing revenue, all thanks to a few simple if statements (they tried machine learning as well, but this did it).
One can only wonder how much of this excessive advertising there is, I guess it's mostly driven by absolute revenue numbers without much consideration for costs and efficiency.