In this guide
Separate what must happen from what you want to test
An incident preventing customers from working, a contractual commitment and a commercial idea have different reasons to enter a roadmap. Making those reasons visible helps you discuss each with suitable criteria.
For each request, state the consequence of delay and the evidence behind it. Understand a deadline described as urgent: who needs it, what happens if it moves and what alternatives exist. A real obligation remains a constraint; an opportunity needs evaluation.
Connect work to an observable outcome
Choose the outcome that matters for the period: for example, helping a segment complete a relevant task or understanding why it does not renew. This is an example to define with your context and data, rather than a universal priority.
A request described only as a feature is hard to compare. Rewrite it as a problem, people affected, available evidence and expected change. State where you are still making an assumption.
Compare alternatives without hiding uncertainty
A framework can support discussion by making reasons visible. A very precise score based on weak estimates can suggest more certainty than the evidence supports.
- Which problem does this address, and for which segment?
- What evidence suggests it deserves attention?
- What dependencies and capacity does it require?
- What is the cost of delay, and how reversible is the choice?
- What check could change the decision before building?
State what leaves the plan
A choice becomes operational when the team knows what to do and what to stop doing. For the selected priority, clarify the owner, first step, constraints and review. For postponed work, record the reason and the condition that could reopen the discussion.
Also define how to handle a new request: who evaluates it and what work it would replace. Otherwise, a shared roadmap may be overturned by the first conversation outside the process.
Revisit decisions for a specific reason
Revisit a priority when new evidence appears, a constraint changes or the intervention does not produce the expected signal. Agree a review point proportionate to the work and distinguish it from daily interruptions.
Start with the next review: bring open initiatives, available capacity and one decision to make. Advisory can help the founder choose; fractional leadership is relevant when someone also needs to take responsibility for coordinating the work.