Why Traditional Product Management Is Dying—and What’s Replacing It

As we write this, the discipline of product management is being transformed before our eyes, and the tidal wave of AI-driven disruptive consolidation among Product Managers (PDMs) is already headed towards shore.

We’ve been doing a ton of reading, researching and talking to our peers about all this, especially in recent months. And after hearing the conversations at the HSE AI-Led Growth Summit this week, we are convinced that the lion’s share of the PDM job — those activities that PDMs today spend their time doing — will be going extinct within the next few years at most.

That said, for those PDMs who survive this revolution, we’re also convinced that the role is going to be much more rewarding — and fun.

Here’s a summary of our thinking.

Agentic AI is already mainstream

Anyone who watched Super Bowl LIX this February 9th and managed to stick around until the fourth quarter (despite the shellacking that Philadelphia was delivering upon Kansas City) would have seen a fantastic AI agent commercial from Ramp Corporation, featuring Philadelphia Eagles star running back Saquon Barkley.

Saquon was shown being pulled off the field mid-game to slog through his overdue corporate expense report (which looked really painful), while Ramp asserted that “EXPENSES SHOULD DO THEMSELVES”. As people who despise doing expenses, we heartily agree.

Clearly, the age of Gen AI Agents is not only upon us — it has hit the mainstream.

And, given VC funding trends, many similarly annoying and mundane activities that also “should do themselves” will actually be doing so very soon, as will many less mundane and “sexier” jobs to be done. Gen AI personalization-at-scale also has immense possibilities for new products, and many companies are already developing market-creating solutions, across industries as varied as healthcare and retail.

It’s also no big secret that gen AI Chatbots, Copilots and Agents are changing — have already changed — the ways in which a significant fraction of white-collar professionals go about their daily jobs. A January 2025 McKinsey report indicates most employees (70%) believe that within 2 years, gen Al will change 30% or more of their work. A significant minority (13%) say they are already there.

Product management professionals are no exception — in fact, they are among the groups most impacted. According to the January 2024 IBM Global AI Adoption Index survey of 2300+ organizations across the world, fully 21% of product managers used AI every day even a year ago, making product folks one of AI’s top user groups in businesses.

The Disruption of the Product Manager

The reality, from everything we’ve read and are seeing, is that a large portion of the work Product people spend their time doing today is quickly commoditizing. Much of this work is boring, remarkably inefficient, and ripe for disruption. Product Manager Agents are coming, and people are already building them.

"Destroy Your Business" (DYB) was a strategy exercise that General Electric (GE) used under Jack Welch’s leadership way back in the 1990s, to prepare for potential disruption from the dotcom bubble. The exercise was based on the idea that a company should identify and address its own weaknesses before competitors do.

Some of the best advice we’ve heard lately (from Amanda Kahlow of 1Mind) for professionals in general, and, we think, product professionals in particular, sounds remarkably similar: “Replace yourself.  Build tools for the tasks of your job, so you can do the higher-level things that AI can’t do.” 

We’ve found that many of our product peers are already doing this. They’re adopting and utilizing copilot and agentic capabilities — not only learning prompt engineering within the major general-purpose LLM apps themselves (e.g., ChatGPT, Gemini, Claude, Llama, Mistral, DeepSeek), but also adopting many purpose-specific AI-first solutions — for virtually every part of the product role. These tools are dramatically speeding up their processes and enhancing the quality of their work.

By “product role”, we refer the following broadly-defined set of business jobs to be done (understanding that there are lots of variations out there on what product people are responsible for): (1) discover product opportunities, (2) determine product strategy, (3) develop product plans, (4) flesh out product solutions, (5) run product operations, (6) oversee product development, (7) manage product launches, and (8) guide the product lifecycle. (We’ll be taking a deeper dive into each of these jobs in separate posts, looking at the product jobs, likely disruptive effects and AI tools available today.)

Given that leading LLMs have humanity’s entire past business knowledge at their fingertips to guide predictions, they can already assist in every one of these product roles to a greater or lesser extent — at least in terms of “copiloting” and giving the product professional a very strong first cut of ideas and deliverables on which to review, edit and iterate.

This means that many product activities that used to take weeks and months now take just hours or days.

Moreover, with the rise of AI Agents (with the ability to not only reply intelligently to questions, but carry out actions of their own), AI is going to keep getting better and better at all of them.

We agree with leading AI experts (see our resources page) that machines aren’t — or at least, shouldn’t be — the thought leaders in all of the above, and that humans still need to oversee each of these activities. At least for the foreseeable future, the functions of product professionals will continue exist within companies, these roles are valuable, and somebody has to own them. As Yahoo CEO Jim Lanzone put it this week, “at internet companies, product people are the hub, and everyone else – marketing, sales, tech, finance, operations – is a spoke.” In technology businesses, strategic product leadership really should be THE mission critical function.

In other words, someone must own the vision and plan of how the company solves its customers’ important unmet needs. Product leadership of strategy, of “what are we doing” is becoming far more important than the “how do we get it done", which gen AI is crushing and is becoming a commodity. Even with the presence of AI Product Manager Agents, the technology’s assistance with “What” questions will be only as good as the humans that ultimately use it.

Another human-only role for product folks is having the ability to align the company around this product strategy. Communication management among cross-functional humans is going to be at a premium.

That said, we’re convinced that far fewer product people will be needed, and we believe that in many cases today’s product functions might end up spread among other teams. As professionals it’s disingenuous, and unfair to ourselves, to think any differently about the coming shakeout. To quote Oasis, “now is not the time to cry, now’s the time to find out why”.

Product Management’s future: the big three questions

As we see it, there are three main questions product managers face in this brave new age of AI:

(1) What to build?

• How and where is AI changing the market for our products? Across industries and geographies?

• How do we make money? What is the right business model for this market? For example, how will managers handle the coming transition from SaaS models (which will likely go extinct) to Agent models?

• What new customer needs will emerge, and which will become obsolete — especially given the forthcoming progression from today’s “reactive” LLM solutions (e.g., chatbots and copilots) to proactive AI agents?

• How can teams foresee what their customers will be wanting, once agent-driven possibilities unfold?

(2) How to work?

• How are the roles and processes associated with product management changing with AI, and what does this mean for our team?

• Given such crazy efficiencies taking hold in everything product people do over the next few years, will big product teams soon become a thing of the past?

• Structurally, will many companies decide that having a distinct team of product managers no longer makes sense?

(3) Where to re-skill?

• As a product professional or general manager, what do I do, how do I change, what must I learn, in order to survive and thrive in this new reality that is changing so fast?

• Given the pace of change lately, it’s becoming clear that some skills and knowledge about AI, data, algorithms and models are now “table stakes” for Product People to have, in order to stay in the game. But which, exactly? Technical skills? Design skills?

• How will this vary by industry and size of company?

These are heady questions. We’re looking forward to getting the perspectives of product and GM leaders, futurists and AI thought leaders, and our audience in the coming weeks and months.

Why all this change is ultimately a good thing

How on earth is all this coming disruption a good thing? Especially if a lot of people in product are probably going to lose their jobs over the next few years?

First and foremost because, speaking frankly, across many organizations the entire product management role has fallen into a rather dark place. Agile development in particular, while highly effective for iterative software delivery, has in many cases reduced Product Managers (PDMs) to mere note-takers. This has happened when development frameworks have been applied in ways that minimize strategic thinking and vision-setting.

The adoption of AI tools and agents will likely help PDMs break free from the tactical grind and reclaim their role as strategic leaders. Gen AI’s ability to automate low-value tasks and provide deeper insights more quickly will empowering PMs to focus on vision, customer needs, and high-impact decision-making. Here’s how:

(1) Reducing Overemphasis on execution over strategy

In many companies, PDMs have found themselves spending excessive time managing the backlog, breaking down tasks, and refining user stories rather than driving vision and strategy. Meanwhile, PDMs have found themselves mainly reactive in terms of prioritization: Instead of shaping the product roadmap, PMs get stuck in responding to the most urgent engineering concerns or stakeholder requests.

AI Agents will increasingly help PDMs to automate low-value backlog, task management and user-story work to free up strategic thinking. AI-powered backlog management will help PDMs auto-prioritize tasks based on business impact, user feedback, and engineering effort. Smart assistants will auto-generate and refine user stories, reducing time spent on writing specs. And automated sprint planning tools will balance effort-versus-impact without manual intervention.

The impact will be that PDMs will spend less time in project management tools like Jira, doing backlog grooming, and writing status updates — and more time on product vision and discovery.

(2) Overcoming the limitations of current Agile practices

In software development, the developers often dictate what’s feasible. Engineers often set the pace and scope of work based on technical constraints rather than user needs or business strategy. Meanwhile, under Agile, Scrum Masters typically control the process — in many teams, to the extent that they become the de facto project manager, reducing the PM’s role to taking notes and facilitating meetings rather than influencing product direction.

With AI, PDMs will find that tools for meeting summarization and action tracking will transcribe and summarize standups, sprint reviews, and retrospectives, allowing PMs to focus on key insights instead of note-taking. AI copilots will track decisions, follow up on blockers, and nudge teams to stay aligned with the roadmap.

With more time to do strategic thinking (and faster/better tools to use in doing this thinking, see below), PDMs should find it much easier to weigh in with valuable insights for teams that will help guide teams on product direction.

(3) Reclaiming customer-centricity

Unfortunately, development today is often sprint- and deadline-driven, rather than user-driven. PDMs may find themselves focusing more on delivering incremental features within sprints rather than deeply understanding customer problems and needs. There’s also a relative lack of discovery time: Agile’s emphasis on rapid delivery can leave little time for research, experimentation, and high-level product thinking.

Using AI-driven sprint planning and work distribution copilots to suggest realistic sprint scopes, PDMs will be able to minimize endless discussions on task breakdowns. Automated tools will also flag technical dependencies & bottlenecks, reducing PDMs’ involvement in micromanaging workflows.

Meanwhile with AI-driven customer research tools, PDMs will gain the ability to analyze thousands of NPS surveys, support tickets, reviews, and interviews to quickly identify trends and pain points. Automated competitor tracking tools will help them monitor market shifts, pricing changes, and feature rollouts in real time, while AI personas can simulate user behavior to predict feature adoption and uncover unmet needs.

The impact: PMs will reclaim time for high-level strategy, and time for talking to real flesh-and-blood customers with real unmet needs, gathering better information without manual data crunching.

(4) Overcoming “feature factory” mindset and reclaiming product planning

Agile sometimes pushes teams to focus on shipping over impact, churning out features rather than solving meaningful problems, and making the PDM a feature-order taker instead of a strategic leader. Agile also tends to have a very short-term focus: the obsession with velocity and sprint goals can prevent PDMs from thinking long-term about differentiation and innovation.

In many companies, PDMs likewise find themselves having the roadmap dictated by stakeholders, executing demands from leadership, sales, or customers without having the authority to push back or shape the vision. Often too there is a lack of big-picture thinking: when Agile is misapplied, teams may become so focused on iterative progress that no one steps back to define where the product should go in the long run.

AI tools will make PDMs superhuman in gaining insights to guide their roadmaps. Using AI-generated strategic insights that scan investor reports, internal memos, and market trends, PDMs will be better able to quickly align product decisions with company-wide goals. AI tools will auto-generate product vision drafts and scenarios based on competitor moves, emerging technologies and more.

With these greater abilities, PDMs will become proactive visionaries instead of overseeing feature factories.

(6) Reducing decision-making by committee

Development ceremonies (daily standups, retrospectives, backlog refinements) can trap PMs in a cycle of coordination rather than decision-making. Business pressures also tend to encourage a consensus-driven culture, where PDMs may feel pressured to accommodate every stakeholder request rather than making hard trade-offs based on strategic priorities.

Using predictive analytics for product roadmaps, PDMs will forecast the impact of feature decisions on revenue, engagement, and retention, while AI copilots will suggest data-driven roadmap trade-offs based on past product performance. They will also gain speed in AI-powered experimentation, using AI to run continuous experiments and simulations, testing different feature variations with minimal engineering effort. ML models can predict likelihood to buy, feature adoption, and churn risk before features are built.

All of this will enable PDMs to move beyond gut instincts and stakeholder pressure, using AI-backed recommendations to drive more confident product decisions.

(7) Enabling the merger of product, design and dev

New and forthcoming AI tools will enable convergence of dev, design and product work, and this will enable the combined team to create solutions the way they need to be created. Tools can already take inputs of little more than text prompts to not just research the market and frame PRDs, but also rapidly assist with design, iteration and testing with customers — and can generate code.

This becomes a really interesting job, and pros will have the ability to take a true vision and carry it all the way through to completion more quickly, easily and successfully, focusing not just on features and functionality, but outcomes and customer impact.

The future of Product Management: a more strategic role

The disruption of product management as we know it isn’t just inevitable—it’s necessary. The rapid advancements in AI are automating the tedious, low-value aspects of the role, forcing product professionals to evolve. While this transformation will reduce the number of traditional PDM roles, it will also create a more strategic, high-impact function for those who adapt.

Rather than being stuck in the weeds of backlog grooming, stakeholder management, and execution oversight, the next generation of product leaders will focus on the big questions: What should we build? How do we create value? Where is the market going?

AI won’t replace great product thinkers—it will empower them. With AI handling execution, data synthesis, and operational tasks, PDMs can reclaim their role as visionaries, strategists, and problem solvers. The future of product isn’t about managing tasks—it’s about owning the vision and driving real innovation.

For those who embrace this shift, product management is about to become more exciting, rewarding, and impactful than ever before.

And that is a good thing.