Short answer
In most founder-led SaaS companies, product decisions are formally the CEO’s, practically the CTO’s, and actually nobody’s, because neither runs a decision process with evidence behind it. The fix is not a personality contest. It is naming the product decision role, giving it an evidence engine, and turning conviction and feasibility into inputs.
The CEO commits features in customer calls. The CTO quietly re-sequences the backlog by what is architecturally sensible or urgent. The product managers navigate between them like the children of an amicable divorce, careful not to quote either parent to the other.
Decisions take weeks, get relitigated after they are made, and the actual roadmap is whatever survived the last argument. Nobody is behaving badly. The structure is producing this, and it will keep producing it until somebody changes the structure.
Why the roadmap ends up as conviction arbitrage
Engineering must build something, so whoever schedules engineering ends up deciding the product. That is the mechanism, and it operates regardless of anyone’s intentions. Meanwhile the CEO’s commitments arrive from outside the system entirely, made in rooms where nobody could check feasibility or priority. What is missing between them is not seniority, since both are senior, and not judgement, since both have it. What is missing is a seat whose whole job is holding customer evidence against both positions and making the call.
The tell in your product managers
Founders in this situation often say their product managers lack initiative. Watch what happens to a PM who makes a firm call: it gets overturned by one of two founders within a fortnight, in front of the team. Two or three rounds of that and any sensible person stops making calls and starts writing tickets instead. The passivity is a rational response to an authority structure, not a hiring mistake.
“I have never once found the CEO and the CTO to be the problem in these companies. The problem is that both of them are being asked to do a third job, part-time, that neither of them signed up for and neither has the time to do properly.”
The test: three questions
Answer these honestly and out loud. Each one is designed to be answerable with a name, a mechanism or a date, so hedging is itself the result. If two founders answer the first question differently, you have found the whole problem in under a minute.
- Name the single person who can kill a feature that both founders like.
- When the CEO and CTO last disagreed, what settled it: evidence, or stamina?
- Do commitments made in sales calls pass through anyone before they reach engineering?
What good looks like in 90 days
The target state is modest and specific. It is not a reorganisation and it does not require anyone to give up territory permanently. It requires a named seat, a weekly forum, a written record, and one visible case where the evidence won. That last item is the one that actually changes behaviour, and it cannot be scheduled, only allowed when the opportunity arrives. See our guide to what the engagement terms should say.
- A named product decision owner: a founder wearing the hat deliberately, a promoted internal leader, or an outside product leader.
- A weekly decision forum with a written log, so decisions stop being relitigated from memory.
- Sales commitments routed through the seat before they reach engineering.
- One case, in public, where customer evidence overturned a founder preference without drama. That is the culture moving.
Your options
The third row below is where the fractional shape genuinely fits better than the alternatives, and the reason is political rather than commercial: an outside product leader is not competing for either founder’s territory, is not angling for a promotion, and will leave. That neutrality is what lets a decision stick between two founders who each have a legitimate claim to the call.
| Option | What it takes | What changes | Fails when |
|---|---|---|---|
| Do nothing | Nothing | Decision latency rises with headcount and the arguments get more expensive as the stakes grow | Continuously |
| A decision-rights one-pager | An afternoon, then discipline | Who decides, who is consulted, what evidence standard applies | It works exactly as well as founder discipline does |
| A neutral product leader | A monthly retainer | Someone senior owns the tradeoff, with authority drawn from customer evidence rather than from either founder | The founders will not actually delegate the call |
For context on the third row, our own engagement runs at $8,000 per month for 20–25 hours, with a 3-month minimum and 30 days’ notice in either direction. The 60–80 customer interviews in the first 60 days are what supply the neutrality: a decision backed by forty conversations is not one founder’s opinion beating another’s.
Product decisions bouncing between two founders?
Thirty minutes, no pitch deck. You will leave knowing which seat is actually empty.
Book a Product Strategy SessionFrequently asked questions
Should the CEO just decide? It is their company.
The CEO owns the strategy, and that is not in question. Running the weekly product tradeoffs is a different job that needs continuous attention, and doing it from the CEO seat means doing it in fragments between fundraising, hiring and sales. That fragmentation is how most companies arrive here, and our piece on the founder bottleneck covers what it costs. Owning the strategy and running the decision process are separable, and separating them is usually the unlock.
Is this what a CPO is for?
Yes. The CPO role is the institutionalisation of exactly this decision seat. Whether it is filled full-time or fractionally is a budget and stage question, and our guide to how many hours the seat actually needs covers it. That the seat exists at all stops being optional somewhere around $3M ARR, when the cost of a slow decision starts exceeding the cost of the person making it.
Our CTO does not want to give up product. Now what?
Frame it as returning their real job rather than removing part of it, because for most CTOs that is literally what it is. Doing product badly, with no evidence pipeline and no time, is exhausting for someone hired to do engineering brilliantly, and the criticism for stalled product lands on them unfairly. Our piece on the technical founder running product covers the same conversation with a founder-CTO.
Can two founders share the decision seat?
In practice, no, and the attempt produces the situation this page describes. Shared authority means every decision can be reopened by the other holder, which converts decisions into positions and meetings into negotiations. One name, with the other founder as a consulted voice and a clear escalation path, works. Two names does not.
The question is never which founder is right about the roadmap. It is whether anyone in the company is accountable for weighing them both against what customers actually said, and if the answer is nobody, the loudest conviction will keep winning by default.