ProductSense and X5: 97% of product managers already use AI, but metrics are barely improving

ProductSense and X5 published the results of a study on how product managers work in 2025. The main conclusion is simple: AI has already become an everyday tool for almost everyone, but its mass adoption has not yet turned into equally widespread growth in product metrics.

How AI entered the workflow

The survey included 1,220 professionals — from product managers to Head of Product and CPOs. Most of them are mid-level, and a notable share of respondents work in large companies with 5,000+ employees. B2B and B2C lead by specialization, and as seniority increases, the skew toward B2B becomes stronger. This matters because it is precisely in large organizations that it is easier to see how new tools move from an employee’s personal habit to a corporate process.

The key number in the study is that 97% of respondents use AI at work. Among CPOs, that figure reached 100%. At the same time, 72% say AI helps save time: some gain up to an hour a day, some gain from one to two hours, and some specialists estimate the effect at more than two hours daily. AI no longer looks like an experiment for enthusiasts: for product teams, it has become as much a working layer as analytics, documents, and meetings.

AI is used most often for tasks where speed and quickly fleshing out ideas matter:

• information search;
• idea generation;
• data analysis;
• competitive analysis;
• checking and challenging hypotheses.

At the same time, managers themselves are not inclined to overestimate the quality of such assistants. In the study, AI is more often compared not to a strong expert, but to an intern who quickly suggests options but requires constant checking.

Why metrics are not growing

At the personal level, the benefit is noticeable, but at the business level the picture is far more modest. 77% of participants do not see changes in product metrics, and only about 20% report growth in indicators.

The study leads to a fairly sober conclusion: local acceleration for an individual specialist does not yet mean the whole company will start moving faster. If approvals, processes, and team dependencies remain the same, the time gains often dissolve inside the organization.

The adoption figures confirm this. Only 34% of companies have at least basic AI guides or training. For 24%, the technology is already reflected in strategy and pilot projects. And only 10% have embedded AI into processes with specific target metrics.

That means most product managers have already adapted AI to their daily work, but in most cases the company itself has not yet rebuilt its operating model around it.

Hence the main paradox of 2025 for the profession: specialists are already working faster, but businesses have not yet learned how to systematically turn that speed into measurable results.

For the market, this is an important signal: the next stage of competition will not be about access to AI tools, but about the quality of their organizational implementation.

Skills, money, and the role

AI skills ranked first for development in 2025. They also lead plans for 2026: 37% of respondents intend to keep improving them. Analytics, economics, and management skills come next.

At the same time, the responses show a gap between what is considered important in hiring and what people actually put into their personal development plans. Companies still place high value on analytics, project management, soft skills, and team collaboration, but individual priorities are much more strongly dominated by whatever helps speed up current work.

As income rises, the focus of learning changes as well. The upper end of the market is increasingly investing not in operational efficiency, but in the ability to manage complexity. Hence the growing interest in meta-skills that are especially important for senior-level professionals, Head of Product, and CPOs.

At the same time, 78.4% of respondents do not expect any noticeable reduction in the number of product managers at all, and 46% believe the role will become less operational and more conductor-like. Another 40% predict the emergence of AI-oriented roles within the profession.

On salaries, the study records a familiar but demonstrably widening range. Among juniors, the 80–150 thousand rubles band is more common; among mid-level specialists, 220–300 thousand; among seniors, 300–550 thousand; among Head of Product and CPOs, from 400 to 750 thousand and above. At the same time, tension in the labor market is also rising: the share of specialists actively looking for a new job grew from 18% to 25%, and compensation has become as important as interest in the product and the tasks.

What this means

For product management, AI has already become the norm, but so far mostly as a personal accelerator rather than a source of systemic business impact. The next milestone for companies is not just to give teams access to models, but to build them into processes, goals, and accountability so that time savings translate into metric growth and into the growth of the product manager’s own role.