The Role Is Disappearing. The Responsibility Is Not.

Two reports landed this year that product leaders should read side by side. One predicts the end of the product manager as a job title. The other shows what happens inside organizations when nobody defends the product decision. Read together, they describe the same problem from two ends.

The first is the 2026 CPO Insights Report, published in May by Products That Count and Mighty Capital on the basis of a survey of more than 1,500 Chief Product Officers across Fortune 1000 companies and startups. Its headline claim is deliberately blunt: the product manager role “as we know it will be obsolete by 2030.” The numbers behind it are less rhetorical. The number of traditional product managers has declined by 30 percent across all industries and by as much as 70 percent in SaaS. In the last year, the number of people in roles the report calls “Product Builder”, which it describes as a hybrid of product and engineering work, has increased tenfold. Three out of four product leaders now use AI to accelerate prototyping and product exploration. And CPOs report spending more of their time on strategy (74 percent, up from 69 in 2025) and innovation (31 percent, up from 21).

The second is ProductPlan’s State of Product Management Report 2026, based on 250 product leaders. It contains a finding that deserves more attention than it has received: over 60 percent of prioritization frameworks are overridden by leadership escalations. The report traces roadmap misalignment back to one dominant cause and frames its answer around a single question: whether product strategy travels clearly across leadership, sales, and delivery teams. Nearly three quarters of the product managers surveyed expect their role to blend across multiple disciplines. And while more than half of organizations hesitate to expand AI adoption, the reason, according to the report, has little to do with usefulness.

What the two reports are actually measuring

It is tempting to read the CPO Insights Report as a story about AI eating a job. That is not what the numbers say. If prototyping has become cheap and engineers can build directly from a well-framed problem, then a role whose main output was the translation of requirements into tickets has indeed lost much of its reason to exist. What the report documents is the disappearance of a particular shape of the role: the coordinator between business and development, the person who writes things down so that others can build them.

What it does not document is the disappearance of the decision. Somebody still has to determine what deserves to be built, why, and how the organization will know it was right.

AI is not eliminating the product decision. It is eliminating the role that used to write it down for others. The CPOs in the survey are shifting their own time toward exactly that work. The question the report leaves open is who does it one and two levels below them, in the product areas and teams where the old role is being cut.

That is where the ProductPlan finding becomes uncomfortable. If six out of ten prioritization decisions are overturned by escalation, the organization does not have a prioritization problem. It has a decision-ownership problem.

Six out of ten prioritization decisions overturned is not a framework problem. It is a decision-ownership problem. The framework was applied. Somebody weighed the options and made a call. And then somebody with more authority and less context replaced the call with an instruction. We have described this pattern before as “Management said”: a statement from above is treated as sufficient justification for work, without being translated into a customer problem, a decision criterion, or a measurable outcome. The ProductPlan data suggests this is not an occasional lapse but the majority case.

Why the two findings belong together

Put the reports next to each other and a mechanism appears. Organizations are removing the role that used to sit between strategy and delivery, on the reasonable grounds that AI has made much of its translation work unnecessary. At the same time, the decisions that role was supposed to own are already being overridden in most organizations. A decision that is overturned by escalation six times out of ten was, in practice, never anchored in a strategy clear enough to hold against it.

In other words: the seat is being emptied at precisely the moment when the organization most needs someone in it who can hold a decision.

The seat is being emptied at exactly the moment the organization most needs someone in it who can hold a decision.

We see the consequences of this in our work, and they rarely look dramatic. Development continues. With AI in the loop, it often continues faster than before. What changes is the quality of what enters delivery. Without a clear owner, the backlog starts to follow whoever asks most loudly or most recently. Stakeholders fill the vacuum with their own priorities, each of them legitimate on its own. Escalations increase, not because people are difficult, but because nobody is authorized to say no with reasons. After a few months there is a roadmap that nobody actually stands behind, and a team that is producing more output than ever against it.

This is upstream debt in its purest form. The organization saved the cost of a role and is now paying interest on every decision that was never properly made. And because implementation has become cheaper, the interest compounds faster: it is now possible to build several versions of the wrong thing in the time it used to take to build one.

The organization saved the cost of a role and is now paying interest on every decision that was never properly made.

The uncomfortable implication for leadership

The convenient conclusion would be that organizations should keep their product managers. That misses the point of both reports. The old shape of the role is genuinely losing its function, and defending it would be defending the wrong thing.

The more useful conclusion is that organizations need to become explicit about where product decisions are owned, by whom, and with what authority, before they remove the people who used to own them implicitly. The CPO Insights Report shows leaders reallocating their own time toward strategy. The ProductPlan report shows that the decisions made below them are being overturned more often than they hold. The gap between the two is where product leadership actually happens: in the translation of strategic intent into decision boundaries that a team, or an AI-assisted builder, can work within without asking for permission every week.

That translation is a leadership responsibility, not a documentation task. It requires someone who understands the customer problem well enough to frame it, understands the organization well enough to know which escalation deserves a hearing and which does not, and has enough standing to hold a decision when it is challenged. Whether that person carries the title Product Manager, Product Leader, Head of Product, or something that does not exist yet, matters far less than whether the responsibility is filled at all.

The title matters far less than whether the responsibility is filled at all.

A practical test

For product leaders reading this, the two reports suggest a simple diagnostic. Take the last five decisions in your area that were escalated and overturned. For each one, ask whether the person who made the original call could point to a strategic intent that the escalation contradicted. If they could, the organization has a communication problem, and it is fixable. If they could not, the decision was never really owned in the first place, and removing the role that nominally held it will change less than expected, because the vacuum was already there.

And if the seat is currently empty, whether through a departure, a parental leave, or a hiring process that is taking longer than planned, the question is not whether the team can keep delivering. It can. The question is what it is delivering toward, and who is deciding that while the seat stays empty.

The team can keep delivering. The question is what it is delivering toward, and who decides that while the seat stays empty.

Framing a customer problem so that it holds, mapping the stakeholders who will escalate and deciding in advance what to do when they do, writing specifications an AI-assisted team can build from, and defining the criteria by which the result will be judged: this is the work the two reports point to, and it is precisely what participants practice on their own product case in CPL-1® – Certified Product Leader. The next cohort starts on 8 October.

---

Sources: 2026 CPO Insights Report, Products That Count and Mighty Capital, May 2026; The State of Product Management Report 2026, ProductPlan, 2026.