AMA: Google Group Product Manager, Android, Shahid Hussain on Product Development Process

In large orgs this happens all the time -- it's just part of the job. The first important this is to accept that it happens -- don't hang on to sunk cost. Secondly be really clear about what that scope change is. If you're getting it from a leadership discussion, be crystal clear on the change and rationale, because even if you're operating with a willingness to manage sunk cost, not everyone on the team may share your enlightened view. Third, have a good sense of what the scope change means for ...Read More

Either out-perform for a smaller segment of the market, or wait for the competitor to fail. If the established competitor is serving a broad set of users or customers, it’s likely that the product isn’t a perfect fit for everyone. There’s lots of complexities here — can the product be customized by users? Is it so dominant that users are just used to it, or it defines the standards in this space? But, if there’s a gap that looks tricky for them, identify it and do a better job serving them. If y ...Read More

If first mover advantage matters, bias toward buy/partner — but only if the org and product integration is realistic, and the cost math works. (Side note — partnerships cost too. It might be opportunity cost, it might be something you’re trading, like sales distribution, to get access to margin or a product integration. But it could also be an opportunity to see up close whether an acquisition could make sense.) If there is a big first mover advantage here, that tips towards buy or partner, assu ...Read More

The fundamentals of building AI products are the same as ever. There's a large cache of discussion on what AI PMs do and how it's different, and there's some value to it. I'd argue that the means of production have meaningfully shifted into high gear -- it's way faster than it's ever been to build and ship product. Whether PMs are vibe coding or not, this has shifted the time and cost of rapid iteration to a point where code is often no longer the constraint -- it's now the testing environment i ...Read More

Less of a framework, more of a question -- are the product needs aligned with what AI is fundamentally good at? LLMs are phenomenal at language, phenomenal at coding, excellent (although costly) at generating images & video, and getting much better at new areas like understanding the physical world and operating as an OS harness. There's a tendency to shove AI in places it actually isn't great at -- understandably, it draws investors and attention. But, I read the backlash against "AI slop" ...Read More