Bug fixes are the same thing as a feature request. They have an implementation cost, and an expected gain in revenue (= reduced loss in revenue). It's still the

Bug fixes are the same thing as a feature request. They have an implementation cost, and an expected gain in revenue (= reduced loss in revenue). It's still the role of the product manager to prioritize it. A 1:10000 freak accident that is not reproducible and therefor hard to fix probably won't be fixed for quite some time, no matter how much it sucks for the single affected customer. Unless you are running a business where each customer pays a relatively big chunck of money (I'd say enough to employ a single developer), you always have to run the numbers, no matter if it is a feature request or a "holy shit" bug.