Busy, Articulate, and Commercially Irrelevant

There’s a particular kind of board meeting that product leaders dread. Not the one where the product is failing, those are uncomfortable but at least legible. It’s clear when you’re getting hit directly in the face! The one I mean is worse: the product is objectively good, the team is shipping consistently, and the CEO or leadership still can’t explain why NRR is sliding. The product leader arrives with a roadmap, a velocity chart, and a discovery backlog. The board wants to know when product will start moving the NRR in a positive way. Neither side is using the wrong words. They’re speaking different languages entirely.

I’ve now sat at that table enough times to stop treating it as a communication problem. It’s structural. And unfortunately for some businesses it’s getting wider and with no plan to close the gap.

What the CPO role was built to do

From roughly 2015 to 2022, B2B SaaS operated on a relatively forgiving commercial logic. Capital was available, growth covered a lot of sins, and the primary product question was: can we build it faster? Discovery frameworks, agile practice, roadmap prioritisation, and stakeholder alignment were the skills that created value. A CPO who could run a tight planning cycle and communicate strategy in a board deck was genuinely useful.

The current framing of CPO responsibilities still centres on these capabilities: vision, team leadership, organisational navigation. There’s nothing wrong with that as a portrait of what the role became. The problem is that it describes a role built for a different commercial context. One that doesn’t include ownership of pricing power, accountability for NRR, or the ability to translate product decisions directly into revenue.

This division of labour felt reasonable when growth was fast enough to obscure the cost of it. When you are adding new logos quickly, the fact that product and revenue are running on separate tracks is not obviously painful. The gap only becomes visible when growth slows, churn becomes structural, or AI commoditises the features you just spent two quarters building.

The market has moved. The predominant archetype hasn’t and maybe ironcically what is actually needed, is what has always been needed in product organsiations but very few had.

Why this is active harm, not just a gap

Here’s the part that took me personally probably too long to see clearly. The described skillset doesn’t simply fail to move revenue. It actively buys time for the decline which is much harder to see, but you feel it.

A product function that is excellent at discovery, prioritisation, and delivery produces a constant, convincing stream of evidence that things are healthy. Velocity is up. The backlog is groomed. Discovery is rich. NPS is respectable. Every artefact a competent product team generates says we are doing our job well, and on the terms it was trained on, it is. That evidence is exactly what delays the intervention the company actually needs. The board sees motion and reads it as progress. By the time NRR has declined far enough to override all that reassuring signal, you’ve lost two or three quarters you couldn’t afford.

Competence at that game doesn’t leave revenue untouched. It unfortunately just masks the problem long enough to make it expensive.

Six engagements, one pattern

Across a series of fractional CPO engagements covering companies at different stages and in different categories, the same six failure patterns appear. They do not all show up in every company, but most companies show at least some of them.

First: capacity is consumed by maintenance, not value creation. Before any intervention, the typical pattern is around 80% of engineering and product capacity going to keeping the existing product running. Bug queues, integration debt, compliance work, customer-specific configurations. The roadmap exists on a slide. In practice, there is almost no room to build the things the roadmap describes. The team is busy. Nothing strategically important is moving.

Second: there is no functional definition of a successful release. Features ship. There is no prior agreement on what adoption looks like at 30, 60, or 90 days, no commercial metric attached to the release, and no one accountable for the number. Discovery is thorough. Post-ship measurement is absent. The team cannot tell you whether anything they built this year changed a commercial outcome.

Third: pricing exists but nobody owns it.

Most B2B SaaS companies in the £2m to £20m ARR range have a pricing model that was set at Series A and updated once, reactively, when a large customer pushed back. Product does not own pricing decisions. Sales owns discounting. Finance owns packaging. Nobody owns the connection between what the product does, what customers pay, and whether that relationship is structurally sound. Pricing power decays quietly for years before it appears on a board dashboard.

Fourth: churn is treated as a retention problem, not a product problem.

CS gets the churn number and the mandate to fix it. Product gets the feature requests that come out of QBRs with at-risk accounts. The structural question — whether the product actually delivers the outcome the customer bought it to achieve — does not get asked in a way that changes roadmap priorities. The feedback loop between churn signal and product decision is slow, indirect, and usually mediated by whoever is loudest in the customer success team.

Fifth: the CPO’s board communication is strategy-forward and outcome-backward. The board deck contains a product vision, a roadmap, a set of capabilities in progress, and a customer story or two. It does not contain a clear answer to: what did product ship last quarter, what commercial outcome did it produce, and what is the causal argument for the next quarter’s priorities? Boards tolerate this longer than they should, then lose confidence in product leadership all at once.

Sixth: the team has added AI features but has not changed how it measures success. This is the newest pattern and it is accelerating. Companies ship AI features because competitors are shipping them and because it is now relatively easy to do so. Adoption is low. The commercial case for the AI investment has not been constructed. The team is now maintaining AI features alongside everything else, with no clearer line to revenue than before. Shipping AI into a product that lacks commercial instrumentation does not produce an AI-native operating system. It produces a more expensive version of the same problem.

In one of these companies, I watched a genuinely talented CPO walk the board through their discovery framework in immaculate detail, then go quiet when a director asked which three features were retention-critical and which were noise. It wasn’t a knowledge gap. He’d simply never been asked to hold the question, so he’d never built the answer. The team was doing precisely what it had been trained to do. That’s the whole point.

The commercial skills that weren’t in the manual

Product management frameworks do include commercial thinking in principle” competitive pricing, differentiation, positioning. But there’s a meaningful distance between knowing how the competition prices and owning the outcome of your own pricing decisions. The first is market awareness. The second is accountability, and most product leaders, in my view incorrectly and to the detriment of the entire business, were never asked to carry it.

Their incentives were tied to delivery, shipping, and satisfaction. Maybe adoption. NRR was someone else’s column in the board pack. That was fine while the product-to-revenue connection was loose enough to manage separately. It isn’t fine now. AI has made features cheap to replicate and commoditised whole capability categories that used to hold pricing power. Churn has and really always was a product strategy problem. Expansion revenue has to be designed, not hoped for. None of that work sits cleanly inside discovery, roadmapping, or agile process.

The counter-argument worth taking seriously

The reasonable objection here is that product leaders are not supposed to own revenue. That is the CRO’s job. Holding a CPO accountable for NRR conflates roles and sets up product teams for unfair blame when commercial underperformance has multiple causes.

There is something to this. A CPO cannot control sales execution or market conditions or the quality of customer success delivery. Assigning revenue ownership to product without giving them the commercial levers to pull would be a different kind of dysfunction.

But the argument dissolves when you look at what actually causes NRR decline in B2B SaaS at scale. Customers churn because the product does not deliver the outcome they expected, or because a competitor now delivers it more completely, or because the category has changed and the product has not kept pace. Those are product problems. Expansion stalls because there is no natural usage path that creates a commercial event. That is a product design problem. Pricing power collapses because features are commoditised and the product has not built the composite moats that make switching costly. That is a product strategy problem.

The CPO cannot own the revenue number in isolation. But if a product leader cannot articulate a direct causal argument from roadmap priorities to commercial outcomes, they are not doing the job the company needs done.

Clarity about what to build is not the same as accountability for what building it produces. Both matter. Most product leaders have been trained only in the former.

Where to start, if you’ve never done this

Diagnosis is the easy part. None of what follows needs a reorg, a new framework, or budget. It needs a product leader willing to go and look at things they were never asked to look at. Roughly in order:

Count the capacity split, actually count it: Take last quarter’s shipped work and tag every item: keeping the lights on, or tied to a defensible commercial thesis (retention, expansion, pricing). Don’t estimate. The real number is almost always lower than anyone in the room would guess, and it’s the most persuasive artefact you’ll produce all quarter.
Read your last twenty churned accounts yourself: Not the CS summary, the accounts themselves. Usage curves, support history, the cancellation reason in the customer’s own words. You’re hunting for the product signal that gets filed under “price” or “champion left”. It’s usually there and if not, seriously consider some proper loss analysis. It’ll be time well spent.
Build a one-page retention map. For each significant or key feature, answer one question: if this disappeared tomorrow, who actually leaves? Pair usage data with what you found in the churn read. Most teams discover that a handful of features carry retention and a long tail carries none, while the roadmap is busy funding the tail or polish.
Put one question in front of every roadmap candidate. Which revenue motion does this protect or create, and how would we know it worked? Things that can’t answer don’t get cut on the spot, but they stop being free. That single filter realigns a backlog faster than a quarter of reprioritisation, because it changes what counts as a good idea.
Get yourself into the next pricing conversation. You don’t need to own pricing, you just need to be in the room, with product usage data, when packaging is decided. If pricing currently happens without Product, that’s the fastest place you’ll add commercial value this year. Where have you learnt that the value actually sits in the product?
Find the expansion path that already existsm or prove there isn’t one. Where does growing usage naturally create a reason to pay more? If the honest answer is “nowhere,” you’ve found a potential roadmap input, not an account-management problem.
Change one line in the board pack. Put product’s commercial contribution next to the velocity chart, even roughly. Having to fill that line in every month quietly reorganises what the team works on, because now someone has to defend it.

None of this is a beautifully structured strategy deck. It’s an operator starting to run the product function as a commercial system, and most of it you can start this week.

What boards should actually be asking in 2026

The question “is our product leader executing well?” is the wrong question. The right question is whether product leadership is operating with commercial logic at its centre, and whether that logic is visible and testable.

A product leader who is commercially connected can answer four questions clearly, without a long setup. What are the top three drivers of NRR in our product right now?

Where is pricing power eroding and what is the product response?

What did we ship in the last 90 days, what adoption have we seen, and what is the commercial implication?

And: what would we need to stop doing to create the capacity to actually move on those priorities?

If the answers to those questions require a 20-minute context-setting session before they can be given, something is structurally wrong. Not with the individual. With the operating model the individual has been given, and with the assumptions the company has made about what the role is supposed to produce.

The companies that navigate the next five years well will be the ones where product strategy converts measurably into commercial outcomes, where the product leader holds the causal logic between roadmap and revenue, and where that logic is rebuilt continuously rather than documented once and left to decay.

The CPO archetype the market currently selects for was built for a different set of conditions. Those conditions are gone. The role needs to catch up.

Frequently asked questions

What does it mean for a CPO to own commercial outcomes?

It does not mean the CPO controls the revenue number. It means the product leader can construct a clear causal argument from specific roadmap decisions to specific commercial outcomes, including NRR, churn, and expansion. They hold that logic accountable over 90-day cycles rather than delegating the measurement to other functions.

How is this different from what a CPO already does?

Most CPOs are trained in discovery, roadmap prioritisation, and strategy communication. Those skills are not the same as commercial ownership. The gap shows up when a team ships prolifically with no measurable linkage between what was built and what happened to revenue or retention. Being busy and articulate is not the same as being commercially connected.

Why has AI made this problem worse?

AI has made it easy to ship features quickly. That means product teams can now be even more active without being any more commercially effective. Shipping an AI feature into a product that lacks commercial instrumentation produces a more expensive version of the original problem. The question was never whether the team could build; it was always whether building would move a commercial outcome.

What should a board look for when evaluating product leadership in 2026?

Ask whether the CPO can answer four questions without a long setup: what drives NRR in the product right now, where is pricing power at risk, what did recent releases produce commercially, and what capacity constraints are blocking the priorities that would actually move the business. The quality of those answers is more diagnostic than any roadmap or product vision document.

What is the difference between a product operator and a product advisor?

An advisor produces clarity: a strategy document, a workshop output, an aligned set of priorities. An operator holds the commercial logic and keeps it running. The test is what happens six months after the engagement. Clarity without operating infrastructure decays. A product operator builds the system that runs after they leave.