from AI pressure to AI strategy
When everyone's asking "why aren't we doing anything with AI yet?" the hardest thing is slowing down enough to answer "what would actually be worth doing?"
the situation
It’s a very familiar kind of tension: competitors are announcing AI features, internal stakeholders are asking why the product isn't keeping up and the team is caught between moving fast, moving right and figuring out how to move at all.
The problem the team was facing wasn't a lack of ambition to get started with a new topic. It was that "do something with AI" isn't a brief. Without a clear answer to who it was for and what problem it actually solved, every direction felt equally valid but also equally risky. The result was decision paralysis dressed up as strategic caution.
what I was brought in to do
Over three months I worked with a cross-functional team (designers, engineers, data scientists, product owners) to turn the pressure into a plan. We didn’t want to build a theoretical roadmap, but a clear sequence of where to start, why and how to introduce AI in a way users would actually trust.
how we got there
I started by getting stakeholders aligned on what success would actually look like. Not in the sense of "we launched an AI feature" but what user value, business impact and responsible use would mean in practice for this product. That conversation alone shifted the frame from "how fast can we ship?" to "what's worth shipping?"
I followed this up with user research with 20+ participants. It revealed something important: the user base was split. One group wanted AI tools immediately and would push the pace. The other needed to be brought along slowly, with transparency and proof before trust. What we learned from this was simple but important: a single rollout approach would have failed both.
From there we mapped and prioritised AI opportunities by value, feasibility and risk and designed a phased framework that started with assistive, low-stakes features and built toward more automated workflows as trust grew on both sides.
what changed
The business moved from asking "where can we use AI?" to "where should we?". This sounds like a small shift to being with, but what it really does is change every decision that follows.
Early prototypes validated the approach without over-committing the team. And the framework gave new AI initiatives a consistent way to be evaluated, rather than each one starting the conversation from scratch.
the broader point
In most B2B product teams, AI is still the area where the gap between business pressure and user readiness is widest.
The companies that navigate it well aren't the ones who move fastest. It’s the ones who have a clear enough picture of their users to know what would land and what would backfire. That's a product strategy problem before it's a design problem.