AMA: Optimizely Chief Product Officer, Rupali Jain on AI Product Management

We're hiring for judgment, not just execution. AI compresses the execution layer significantly, so the premium skill is knowing what to build and why, not just how to build it. Specifically, here's what I look for now (always needed these, but alongside core execution and synthesis skills which were equal weight, but the former just became non negotiable and the latter a lot less important): Problem framing over problem solving. The ability to define the right problem is more valuable than churn ...Read More

This is probably the most important discipline question in product right now. The honest answer is that maintaining focus requires saying no to things that are genuinely cool but don't move a structural lever. AI allows us to build and ship features faster than ever, that does NOT mean that we build everything anyone comes up with, there is much more judgement needed now on what to actually build that solves real customer problems.  The filter I use isn't "can AI do this?" It's "does AI create l ...Read More

The landscape has split into three camps: point-solution AI tools that are reimagining how work gets done but often shallow given the lack of customer learning, and platform players embedding AI deeply into reengineered workflows designed ground up for agents. We're deliberately playing the second game taking advantage of our deep penetration with customers, transitioning to operate like AI natives and reworking workflows to be agent first. There is a 3rd camp of SaaS players adding AI capabilit ...Read More

Few patterns stand out from what I've seen across our customer base and our own product development: AI in the flow. The highest adoption happens when AI is embedded in the existing workflow. The moment you make someone leave their workflow to use AI, adoption craters. Meet people where they already work. You can absolutely change how they work, but to do that, they need to still find / "discover" what you have to offer Transparent explanations. Users adopt AI faster when they can see why it's r ...Read More

Great question, and I think it gets at something a lot of PMs are quietly anxious about. AI is exceptionally good at generating possible solutions. Give it a problem statement and it will give you ten plausible approaches in seconds. But here's what it can't do: decide which solution to build given constraints the model can't see. Organizational capacity, strategic direction, business model implications and what are unique advantages that your team/company bring to bear. The PM's job in solution ...Read More