[akiselev]

> What this article is about, and what most people mean when they say you should be more opinionated is that you should not be to generic, that you should focus on your core use-case and from a companies POV that is always a good idea IMHO. At least as long as you core use-case is the use-case people by your software for.

Nowadays I think the problem isn't a lack of opinions but people's opinions chasing messy (it not outright useless) data and feedback without a vision for what the product is. They become so obsessed about whether they could [implement this feature/expand to more markets/get more big clients/earn more revenue] according to X data ("because SCIENCE!") that they never stop to think whether they should.

IMO that's how opinionated people help build great products: by stopping cargo culting, scope creep, and desperate measures of all kinds that are backed by bad data. That doesn't mean that they know exactly what their team should be working on next sprint, but they do care enough to shut down attempts from other departments that would degrade the product, even if that means passing up short term gains that look good on paper due to customer feedback or usage data.