To prioritise work reliably, cap active commitments to your team's real capacity and rank tasks by what breaks if delayed by two weeks. Make every new request replace an existing item on the queue.
Most priority lists fail because leadership adds work faster than teams finish it. Teams facing that bottleneck often fix the intake rule first when working with a fractional CPO on product strategy.
How do you prioritise work when everything feels important?
Separate genuine deadlines from loud requests, tie each task to one measurable number, cap active bets, and publish one ranked queue. When every stakeholder sees the trade-off in writing, urgency stops turning every week into firefighting.
- Separate deadlines from loud requests. A legal deadline, a live customer failure, or a signed launch date has a clock. Everything else is important work waiting its turn, as covered in how to prioritise when everything feels urgent.
- Name the number each task moves. Every piece of work should point to leads, activation, retention, revenue, or hours saved. If nobody can name the number, drop it down the queue.
- Cap work in progress. Half-finished work consumes team capacity without returning value. On a mobile language app with a fixed App Store deadline, cutting scope by 40% shipped the launch on time with Day 1 activation tracking.
- Publish one ranked queue. One ordered list removes side deals. Adding item number three means pushing item number four out of the cycle.
Which prioritisation method fits your stage?
Match your prioritisation method to the data you actually hold rather than copying a scoring sheet from a larger company. Early-stage teams lack reliable reach metrics, while mature teams waste time arguing without numbers.
| Method | Best stage | Data needed | Where it fails |
|---|---|---|---|
| Cost of Delay | Any stage with competing deadlines | Consequence of waiting 2 weeks | Teams overestimate commercial loss without checking churn or pipeline data |
| MoSCoW | Fixed launch dates or early MVPs | Shared agreement on launch blockers | Stakeholders label every feature "Must-have" and break the filter |
| RICE | Growth teams with steady analytics | Measured reach, conversion baseline, effort | Early teams guess numbers and produce false precision |
| Kano | Established products losing retention | User interviews and usage signals | Skipped when teams rely on internal opinions instead of customer evidence |
For a full breakdown of scoring failures and when to override a formula, read feature prioritisation frameworks that hold up.
What is the cost-of-delay test?
The cost-of-delay test measures what breaks, degrades, or goes uncaptured if a task waits two weeks. Tasks with a flat cost of delay wait behind tasks whose value drops each week they sit unfinished.
Run every candidate through three questions:
- What breaks if we wait two weeks? If an enterprise buyer delays signing or a regulatory check expires, the cost is concrete. If nothing breaks in 14 days, the task belongs behind time-sensitive items.
- What measurable number degrades while we wait? Count churned accounts, stalled onboarding steps, or weekly hours lost to manual rekeying. Concrete weekly drag beats vague concern about falling behind competitors.
- Does finishing this buy future team capacity? Work that removes a recurring bug class, automates manual reporting, or clarifies a product operating model frees engineering and product hours for every cycle after it.
How do you say no to stakeholders without losing trust?
Replace a flat "no" with a visible trade-off against committed work. Requesters push back against closed doors, whereas visible capacity lets them help decide which bet moves.
Use one sentence when an urgent request lands mid-cycle: "We can ship X this week, which moves Y back by two weeks. Which do you want?" That shifts the stakeholder from lobbying to choosing.
When an executive asks for a maximalist feature, offer the smallest version that solves the underlying trigger first. Most requests shrink by half once the team asks which decision or customer problem the first version needs to unlock.
A weekly 30-minute checklist to prioritise work
Run this 30-minute triage every Monday before sprint planning or weekly stand-ups:
- List active work in flight. Verify whether any item is stuck waiting on a decision or missing context.
- Cut anything unlinked to an outcome. Archive backlog items older than 90 days that have no named user problem or commercial metric.
- Rank the next 3-5 items by cost of delay. Put time-sensitive revenue blockers and capacity builders at the top.
- Confirm what is explicitly off the table. Tell stakeholders what is waiting until next cycle instead of letting requests sit silent.
- Check last month's launches. Verify whether shipped items moved their target number before starting similar work.
Book a 20-minute call
If your roadmap is full, every stakeholder calls their request urgent, and delivery keeps slipping, bring your current queue to a call. You will leave with a clear structure for ranking bets and getting leadership aligned around one set of priorities. Book a call or review how I work on product leadership.