For decades, the Chief Product Officer's greatest challenge was simple: too many ideas, not enough engineering capacity. Product leaders spent their careers learning to prioritise ruthlessly, to gate releases carefully, and to manage stakeholder expectations when great ideas sat on roadmaps for months, or disappeared entirely. Engineering capacity was the bottleneck, and the CPO's job was to manage scarcity.

That world is ending. And the implications are more profound than most product leaders realise.

The capacity revolution

AI-powered development tools are fundamentally rewriting the economics of software creation. At Hg, we're seeing portfolio companies achieve up to 8-10x productivity gains in engineering through tools like Devin and Cursor. What previously took months now takes weeks. Engineering capacity, for the first time in the history of software, is approaching unlimited. This should be cause for celebration. Except it isn't that simple. Because when you remove one bottleneck, you inevitably create another. And the new bottleneck is product management itself.

When engineering moves at AI speed

Consider the traditional SaaS rhythm: product management stays a comfortable distance ahead of engineering, defining next quarter's work whilst engineering builds this quarter's features. Now imagine engineering capacity increases tenfold overnight. That comfortable distance evaporates. Engineering teams can build in weeks what previously took months. They're ready for the next project before product management has even finished validating the current one.

This creates an uncomfortable new reality: unlimited AI engineering capacity is only valuable if you can feed it with the right ideas, fast enough, and with sufficient detail to build correctly at speed.

But the challenge extends far beyond keeping engineering busy. Think about what happens when you compress a six-month product development cycle into six weeks. Can your marketing team create compelling messaging in that timeframe? Can sales absorb new products and sell them effectively at that pace? Can customer success onboard and support customers when capabilities expand weekly rather than quarterly?

The answer, in most organisations today, is no. The productivity gains in engineering get trapped. You've moved the bottleneck from engineering to everywhere else.

This is the central challenge facing CPOs in 2025: how do you lead supercharging the entire product lifecycle to match the new velocity of AI-powered engineering?

The CPO at the centre of the universe

This transformation elevates the CPO's role from important to essential. Product management becomes the orchestration point for the entire organisation's AI transformation – not just R&D, but the complete ecosystem required to turn ideas into customer value.

This topic will be at the forefront of our CPO Forum in New York this week. We'll hear from Oji Udezue, who's navigated this at Twitter, Calendly, Atlassian, and Microsoft, on how the CPO role must evolve. Core CPO skills still matter: customer insight, strategic vision, prioritisation discipline. But new capabilities become critical.

You must embed yourself in critical customer workflows. Lindsay Sparks, Hg Serial Chair, will talk about ‘owning the last mile’ – what that means and how to put that into action to set product direction and build defendable positions, all while keeping the commercial model front and centre.

Deep customer intimacy becomes non-negotiable. You must understand what customers actually do, not what you think they do. This means watching them work, mapping workflows in detail, instrumenting platforms to see real behaviour. Carrie Osman from Cruxy will explore why user intimacy is now the true competitive intelligence. You can't reimagine workflows you don't deeply understand.

Cross-functional orchestration must happen from day one. Travis Arthur (Chief Marketing Officer, GTreasury) and Wes Childs (Product & Engineering Director, Access) will talk about their journeys to rethink team structures and bring product, engineering, marketing, sales, and customer teams together early when building and launching new AI products.

Know your competition - leverage your strengths. Litera CEO Avaneesh Marwaha will share how the company is defending its market position against AI-native startups, not by matching them feature-for-feature, but by combining 30 years of legal workflow expertise with AI-first product development.

AI-native thinking means reimagining the art of the possible when AI agents can actively execute tasks rather than just assist. The shift from passive SaaS tools to proactive agentic systems changes everything: pricing models, adoption patterns, competitive advantage.

The transformation playbook

So what does this mean practically? Based on our work across Hg's portfolio, successful CPOs are making several critical shifts:

Prototype everything, ship selectively. With AI tools like Claude and Replit, CPOs can build working prototypes in hours. The best product leaders prototype 10 ideas, but may only ship two, using low-cost AI prototyping to fail fast and validate thoroughly (including with customers) before committing engineering resources.

Expand your orbit early. Product management partnerships must extend beyond those with their Engineering colleagues. Successful CPOs huddle with pricing, marketing, sales, and customer success from the ideation phase. When release cycles compress from months to weeks, sequential handoffs become impossible. Everything must happen in parallel, with effective coordination.

Instrument obsessively. You cannot build AI-powered workflow automation for processes you don't deeply understand. The best teams invest heavily in telemetry, user research, and workflow mapping. They watch screen recordings, analyse click patterns, and walk through customer workflows in forensic detail. This deep intimacy with actual usage (not assumed usage) becomes your differentiating factor.

Rethink team capacity. If engineering consistently delivers 10x more but product management only improves 2-3x, you may need more product capacity. But hiring isn't the only answer. AI tools can accelerate market research, competitive analysis, and PRD creation. Strong CPOs are deploying AI systematically across their orgs to close the gap.

Embrace the “Last Mile”. Generic AI capabilities are commoditised. Everyone has access to the same foundation models. Your advantage lies in understanding the specific, nuanced workflows your customers execute. This domain expertise (how accountants close books, how lawyers conduct discovery, how treasury teams manage cash) becomes your position. The best CPOs are investing in owning and deepening this relentlessly.

The risks of moving fast

There's a critical caveat: unlimited capacity to build the wrong things faster is not a competitive advantage. It's a recipe for waste and distraction.

This is why the CPO's role becomes more demanding, not easier. Yes, you can ship more. But you must ensure you're shipping the right things. That starts with brutal honesty about what you know versus what you assume about customer workflows. It demands the discipline to prototype and validate before committing engineering resources at scale. It requires the courage to kill projects quickly and adapt when prototypes reveal flawed assumptions. And it necessitates organisational alignment so speed in one function doesn't create chaos in others.

The temptation will be to build everything. The discipline required is to build only what genuinely raises the bar for customers.

Why this matters now

We're seeing a clear pattern across Hg's portfolio: companies that transform their product organisations to match AI-enabled engineering velocity are pulling ahead. Those that don't are finding their theoretical 10x productivity gains trapped by organisational friction.

The window of opportunity is narrow. Competitors are moving. Technology capabilities are doubling every few months. The organisations that successfully transform their entire product value chain - not just engineering - will define the next era of software leadership.

The call to action

If you're a Chief Product Officer, the uncomfortable truth is that your role is changing faster than any product cycle you've ever managed. The skills that made you successful remain important but are no longer sufficient.

You must now lead a transformation touching every function connected to product development. You must move from managing scarcity to orchestrating abundance. You must think in weeks, not quarters. And you must do all of this whilst maintaining the strategic vision and customer intimacy that only product leadership can provide.

The CPO has become the centre of the universe. The question is whether you're ready to lead from that position.

At Hg's CPO Forum this week in New York, we're exploring these challenges with product leaders navigating this transformation in real-time. The conversation has never been more urgent. The opportunity has never been greater. And the cost of waiting has never been higher.

Jason Richards is Head of Portfolio Technology at Hg, where he works with Chief Product Officers and technology leaders across more than 50 software businesses navigating the AI transformation.

This article reflects Hg’s current views and opinions regarding broader market trends, based on information available as of the date of publication. It is provided for general informational and discussion purposes only and should not be construed as investment, legal, accounting, or other professional advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any security. Any forward-looking statements, projections, or estimates are based on assumptions that may change without notice and are not guarantees of future events or performance. Past performance is not indicative of future results. There is no guarantee or assurance that the trends described or depicted above will continue and actual results may differ materially. References to companies or technologies are illustrative and do not imply any affiliation with or endorsement by Hg.