The 90-Day Roadmap Template
Prerequisite: a completed One-Page Strategy. Every initiative below must trace to one of its priorities. Rules: 5–8 initiatives maximum, one named owner each, sequenced by dependency. Cut the roadmap to fit your people — never the reverse.
The initiative stack
| Initiative (verb + definition of done) | Owner | Parent priority | Metric & target | Depends on | Month | Kill / re-scope trigger |
|---|---|---|---|---|---|---|
Sequencing check
- Foundations before amplifiers (fix conversion before buying traffic; reprice before scaling sales).
- The riskiest assumption from the one-pager gets tested in month one, not month three.
- No owner appears on more than two initiatives.
- Month 1 is fuller than month 3 — roadmaps should front-load learning.
The weekly cadence (45 minutes, same slot every week)
| Segment | Time | Content |
|---|---|---|
| Metrics | 10 min | The metric tree, pre-read before the meeting. Discuss exceptions only. |
| Initiative round | 20 min | Each owner, 2 minutes: on track / blocked / decision needed. No status theater. |
| Decisions | 10 min | Resolve the "decision needed" items. Log what was decided and why. |
| Re-commit | 5 min | Next week's single most important outcome, per owner, out loud. |
Two standing rules: an initiative may be killed or re-scoped, but never silently dropped; and the cadence never skips two weeks in a row — that's how strategies die.
Checkpoints
| Day | Question | Written answer due from |
|---|---|---|
| 30 | Is the riskiest assumption confirmed, killed, or still untested? (Untested = a planning failure.) | |
| 60 | Which initiative is furthest behind, and is the cause capacity, dependency, or commitment? | |
| 90 | What did we learn that changes the one-pager? Update it, then build the next 90 days. |
Three columns most roadmaps omit, and why they are the important ones.
Initiative, owner and date are on every roadmap ever written. The next three are what determine whether it survives the quarter.
Capacity check. What did this owner stop doing to take this on? Nobody starts a quarter at 0% utilisation. A roadmap that adds work without naming what it displaces is a wish list with dates, and the blank rows are the items that will slip.
Depends on. The item or decision that must land first. Undrawn dependencies are why plans that look parallel execute serially, and why a 90-day plan quietly becomes a 150-day plan without anyone having decided it should.
Kill criterion. The condition and the date at which this stops. Written now, in week one, while you have nothing invested and no public position to defend. By week eight your judgment about when to quit is contaminated by sunk cost and by having told people it would work.
One more rule: five to seven initiatives for a team of twenty. If the list has fifteen, the quarter has already been decided — the two with the loudest champions will get done and the rest will slip.
Using the template
What makes a good kill criterion?
Specific enough that two people would agree on whether it has been met. "If we have not booked six qualified demos from this channel by day 45, we stop" works. "If it is not working" does not — everything looks partly like it is working to the person who proposed it.
How do we handle work that arrives mid-quarter?
Reserve 15–20% of capacity explicitly as a buffer, so genuine urgency has somewhere to go that is not the roadmap. Anything exceeding the buffer requires an explicit trade, and the person requesting the change picks what comes off.
Should this replace our OKRs?
It sits underneath them. OKRs state the ambition; this states who is doing it, what they stopped doing to make room, what has to happen first, and when you quit. Most OKR implementations fail on exactly those four questions, which they were never designed to answer.