Article: 2 OF 3
In the first piece in this series, we described a puzzle. 97% of product managers report that AI makes them more productive, but only 64% say it’s actually improving their product outcomes. A third of the productivity gain evaporates before it reaches the business.
We said we’d gone looking for what explained the gap, and that the answer wasn’t better AI. It was what surrounded it.
Here is what we found.
Five conditions associated with success
When we analysed the responses in our 2026 Survey of Product Management, we looked for the factors that most consistently separated teams with strong outcomes from the rest. Not just AI outcomes — better product outcomes across the board. We looked at dozens of variables. Five organisational conditions kept surfacing, across every outcome we measured. Here we call them foundations — conditions that need to be true for effective product management.
None of them will surprise an experienced product leader. What’s surprising is how rarely they exist together, and how dramatically they compound when they do.
Foundations that correlated with better product outcomes using AI, fewer disruptions, and hitting deadlines on time.
1. Customer time
Do product managers feel they spend enough time engaging with customers and understanding the market? Only 29% said yes. That makes it the rarest of the five foundations — and arguably the most consequential.
Among heads and directors of product, the picture is even starker: 75% say they don’t spend enough time with customers. The more senior and busier you get, the further you drift from the source of insight that can help drive great decisions.
Without genuine customer understanding, AI helps you move faster toward assumptions rather than evidence. You generate better-looking outputs, more polished decks, faster analyses — all built on foundations of sand.
2. Role clarity
Are roles and responsibilities clearly defined? 59% of respondents said they were. The other 41% are navigating ambiguity every day — unsure where product management ends and project management, engineering, or marketing begins.
The impact on execution is concrete. When roles are clearly defined, teams hit deadlines 78% of the time. When they’re not, that drops to 48% — a 30-point gap driven entirely by whether people know what’s expected of them.
It sounds obvious. But the data suggests most organisations still haven’t addressed it. When product management boundaries are fuzzy, PMs fill the vacuum — taking on operational work that should sit elsewhere. AI doesn’t fix this. It just helps them do other people’s work more efficiently.
3. Strategic focus
Is strategic work — e.g. deciding the right problems to solve, the right product to build — made a priority? For most, no. Only 25% said they spend most of their time on strategic activities. 60% of respondents said they are frequently disrupted by unplanned work.
57% said strategic product activities were a number 1 or 2 priority. But wanting to focus on strategy versus doing it are very different things. The gap between intention and reality is filled with urgent messages, escalations, and the steady drip of unplanned work.
4. Leadership recognition
Is product management seen as a leadership role in the company? 67% said yes. But the number shifts dramatically depending on reporting structure. Where PMs report to a Chief Product Officer, 80% say the role is recognised as leadership. Where they report into sales or engineering, it drops to around 50%.
This is not just a matter of titles. When product reports into sales, it risks becoming an order-taking function. When it reports into engineering, a feature-specification function. Neither is the strategic function that product management needs to be.
5. Clear accountability
Does each product manager have a clear primary metric they’re held accountable for? 34% said they don’t have one. A third of product managers are working without a defined measure of success.
The consequences ripple outward. Without a clear metric, everything feels equally urgent. There’s no basis for prioritisation, no way to say “this matters more than that” and have it stick. PMs end up accountable for everything in principle and measured on nothing in practice. Give a PM a clear metric and you give them a decision-making framework. Without one, AI-generated analysis and recommendations have no anchor — no way to be assessed against what matters most.
The compounding effect
Each of these conditions matters individually. But the striking finding is what happens when they come together.
Product managers with all five foundations in place hit their deadlines 86% of the time. Those with zero or one hit them 44% of the time. That’s a 42-point gap.
And on the question that started this series — does AI improve product outcomes? — the difference is even more pronounced. Product managers with all five foundations were over five times more likely to say AI had significantly improved their product outcomes. 44% versus 9%.
Only 8% of respondents had all five foundations in place.
Most organisations sit at two or three, a great starting position. And each foundation you add correlates with measurably better results. Deadline performance climbs progressively from 33% with none to 45% with one, 53% with two, 63%with three, 74% with four, and 86% with all five foundations. This is not all-or-nothing. Every step counts.
McKinsey’s research found a similar pattern from a different direction. Their 6% of high performers share characteristics that map closely to what we’re seeing: workflow redesign, senior leadership ownership, and defined KPIs. Different survey, same small percentage getting everything right.
Bigger than AI
We opened this series with the 33-point AI productivity gap because that’s the question everyone is asking right now. But here is where the story gets wider — and more important.
These five foundations don’t just explain the AI gap. They correlate with dramatically less disruption — 68% were frequently disrupted when zero or one foundation in place, versus 28% with all five in place. They correlate with better professional development — from essentially 0% rating it as good with no foundations, to 63% with all five. And they correlated with economic resilience: teams with all five foundations were five times more likely to be accelerating long-term product initiatives despite current economic headwinds.
The AI question was the door into this room. But the room is much larger. These foundations all correlate with effective product management — the conditions that determine whether a product team delivers great product outcomes despite economic headwinds.
AI works. It works a lot better when the organisation around it does too. And as AI helps teams to deliver faster, the PM role shifts toward spending more time on strategy, customer insight, and judgement — exactly the capabilities these foundations enable.
Why they’re rare
If these foundations sound like common sense, you’re right. Every experienced product leader would recognise them. So why do only 8% have the full set?
Because organisations have gravity. 60% of product managers are frequently disrupted by unplanned work — not because anyone chose that, but because systems and incentives pull in that direction. 49% cite budget as the top barrier to professional development. 64% of organisations are focused primarily on efficiency and cost-saving. When the pressure is on short-term results, the foundations that enable long-term effectiveness are the first to be deprioritised. There can be a structural pull toward firefighting, short-termism, and underinvestment in the conditions that make product management effective. It’s not malice. It’s organisational inertia.
This is not a criticism. It’s a description of what many organisations are navigating. The encouraging finding is that most teams aren’t starting from zero — they’re at two or three foundations, and each one you add makes a measurable difference.
The data points to what matters. The question is what to do about it — and where to start.
Next in this series: Steer First, Then Accelerate.
Based on the 2026 Survey of the Product Management Profession (677 respondents, 40 countries). Read the full survey report
Leave a comment