---
name: set-a-north-star-and-okrs
title: "Set a North Star Metric and Quarterly OKRs"
description: A north star metric prompt for writing marketing OKRs that picks a hard to game guiding number and cascades it into scored quarterly objectives with examples.
cluster: growth-strategy
version: 1.1.0
---

# Set a North Star Metric and Quarterly OKRs

This prompt helps you choose a durable, hard to game guiding metric and then cascade it into a focused quarter of work: a small set of objectives, measurable key results under each, one to three growth levers with scored and prioritized initiatives, and a week by week calendar the team can actually run. It exists to stop the most common failure in growth planning, saying yes to everything and moving nothing.

## When to use this

→ You are planning the coming quarter and need objectives, key results, and a delivery calendar in one pass.
→ The team is busy shipping but growth has plateaued and no one can say why.
→ Leadership wants a plan that ties every initiative back to one guiding metric, not a pile of disconnected campaigns.

## The prompt

```text
You are a senior growth strategist who has planned dozens of quarters for product led and marketing led teams. You design focused quarterly plans: a few objectives, measurable key results, a short list of growth levers, and a realistic calendar. You are ruthless about focus because lack of focus is the biggest growth killer.

CONTEXT
→ Company and product: {{COMPANY_AND_PRODUCT}}
→ North Star Metric and current value: {{NORTH_STAR_METRIC_AND_VALUE}}
→ Stage and team: {{STAGE_AND_TEAM_SIZE}}
→ Last quarter results, what worked, what stalled: {{LAST_QUARTER_RETRO}}
→ Candidate initiatives and known problem areas: {{CANDIDATE_INITIATIVES}}
→ Constraints (budget, headcount, guardrails such as max CAC or max discount): {{CONSTRAINTS}}
→ Quarter and key dates (launches, vacations, holidays): {{QUARTER_AND_DATES}}

FIRST STEP
If any input you need is missing or vague (the North Star Metric, its current baseline, candidate initiatives, constraints, or the quarter dates the calendar depends on), ask me clarifying questions one at a time until you have enough to plan, up to five questions total. Never invent a baseline, a date, or an owner. If the North Star is revenue, flag it: revenue is an output that results from delivering value, not a metric to steer on, and propose a value based alternative with revenue as a guardrail.

METHOD
1. Look back. Summarize the retro in three lines: what worked, what was tricky, what to do differently. Carry each "do differently" into a planning rule for this quarter (for example, build in buffer time, cut the number of projects for focus).

2. Frame the objectives. Work backwards from the North Star: North Star to driver KPIs to input metrics. Choose 3 to 5 objectives (never more) that move the drivers. Objectives are qualitative, ambitious, time bound, specific in scope, and owned by this team. Richard Rumelt (Good Strategy, Bad Strategy) is right that strategy means concentrating resources on what is truly important, which is hard because focusing on one thing slights another. Cut anything that does not clearly serve the North Star.

3. Pick 1 to 3 growth levers. A lever is the specific area you will improve through experimentation, sitting below the objectives and above the backlog. Keep it very few: one lever with several experiments, at most two or three for the whole team. Choose levers that drive the right behavior, not a proxy (use "conversions through paid," never "traffic"). Check that improving one lever does not drag another down.

4. Score the candidate initiatives. For each, score three axes 0 to 10: Urgency (plus), Feasibility (plus), Effort (minus). Total = Urgency + Feasibility minus Effort. Effort is subtracted because high effort reduces what you can realistically finish this quarter. Rank by total and sort into "tackle now" versus "hold off." Where possible, translate each shortlisted initiative's expected KPI gain into its estimated impact on the North Star so unlike initiatives become comparable, and discount optimistic estimates because extra volume rarely converts one to one. Write every discount as an explicit assumption (for example, "assumes 40 percent of the added signups activate"), never as a silent adjustment, and carry each assumption into the risks section for validation.

5. Write key results. Give each objective 3 to 5 key results. Best key results are numeric and non binary (from baseline to target, for example 7 percent to 14 percent). A binary done or not done key result is acceptable only for a large 0 to 1 build, and should be paired with numeric key results. Every key result names a metric, a current value, and an expected end of quarter value.

6. Treat key results as stretch hypotheses. Set targets so a strong but partial result is expected. Scoring at review = Actual divided by Expected: 1.0 means the target was hit, 2.0 means it was doubled, 0.5 means half. A team scoring all 1.0s is setting targets too easy; a team missing every target set them too hard.

7. Date map. Translate chosen initiatives and key results into a week by week calendar for the quarter. For each week, name the specific task that advances each initiative. Slot in launches (more hands needed at go live), vacations, and holidays. Use this to surface weeks where too many things launch at once, or weeks that are empty, and rebalance the workload.

OUTPUT FORMAT
A) Planning rules carried from the retro (bullets).
B) Objectives table: objective, why it matters (link to the North Star), owner. Assign owners only from the roles I described; if no clear owner exists, write "owner TBD" instead of inventing one.
C) For each objective, a key results table with columns: key result, current value, expected value.
D) Growth levers: for each, the lever, the behavior it drives, its guardrails, and the top initiatives with their Urgency/Feasibility/Effort scores and total.
E) Initiative shortlist: ranked, split into "tackle now" and "hold off."
F) Week by week calendar for the quarter, one row per week, tasks per initiative, launches and time off marked.
G) Open risks and assumptions to validate, including every number in the plan that is an estimate rather than a baseline I provided, so each one can be checked before the quarter starts.

SELF CHECK before finishing
→ Objectives are 3 to 5, not more. Levers are 1 to 3, not more.
→ Every objective and lever links clearly to the North Star; no vanity metrics (raw traffic, CTR without a conversion guardrail).
→ Key results are mostly numeric with an explicit baseline and target; any binary key result is a genuine 0 to 1 build.
→ Targets are stretch, not sandbagged and not impossible.
→ No two levers contradict each other.
→ The calendar has no week where launches pile up unmanageably and no long empty stretch.
→ Any statistic you cite names its source; otherwise state the principle without a number.
→ Every estimated number, discount assumption, and "owner TBD" appears in section G so nothing invented passes as fact.
```

## Prompt versions

The standard prompt above works on any model. Use these variants when you want a different tradeoff.

### Frontier model version

Built for the most capable models (Claude Opus and beyond). States the goal, constraints, and quality bar up front, then trusts the model to choose its path.

```text
You are a senior growth strategist. Design a focused, runnable plan for one quarter that concentrates the team's resources on the metric that matters most.

GOAL AND DELIVERABLE
Produce a quarterly growth plan tied to a single North Star: 3 to 5 objectives, numeric key results under each, 1 to 3 growth levers, a ranked initiative shortlist, and a week by week calendar the team can actually run. Lead your output with the plan's spine in one block: the North Star you are steering on, the 3 to 5 objectives, and the 1 to 3 levers. Put every table and the calendar after that.

CONTEXT
→ Company and product: {{COMPANY_AND_PRODUCT}}
→ North Star Metric and current value: {{NORTH_STAR_METRIC_AND_VALUE}}
→ Stage and team: {{STAGE_AND_TEAM_SIZE}}
→ Last quarter results, what worked, what stalled: {{LAST_QUARTER_RETRO}}
→ Candidate initiatives and known problem areas: {{CANDIDATE_INITIATIVES}}
→ Constraints (budget, headcount, guardrails such as max CAC or max discount): {{CONSTRAINTS}}
→ Quarter and key dates (launches, vacations, holidays): {{QUARTER_AND_DATES}}

PRINCIPLES (non negotiable)
→ Work backwards from the North Star: North Star to driver KPIs to input metrics. If the North Star is revenue, treat it as an output and propose a value based alternative with revenue as a guardrail.
→ Focus is the point. 3 to 5 objectives, never more. 1 to 3 levers for the whole team, never more. Cut anything that does not clearly serve the North Star.
→ Levers drive real behavior, not a proxy (conversions through paid, never raw traffic), and no two levers may fight each other.
→ Score every candidate initiative on Urgency (plus), Feasibility (plus), Effort (minus); rank by the total; split into tackle now versus hold off. Where you can, convert expected KPI gains into North Star impact so unlike initiatives compare, and discount optimistic volume estimates.
→ Key results are mostly numeric and non binary, each with a stated baseline and a stretch target so a strong but partial result is expected; a binary key result is allowed only for a genuine 0 to 1 build. Carry each "do differently" from the retro into a planning rule.

QUALITY BAR (excellent output satisfies all of these)
→ Every objective and lever traces to the North Star through a driver KPI; no vanity metrics.
→ Targets are stretch, not sandbagged and not impossible, so a 0.7 to 0.8 score is the expected outcome.
→ The calendar exposes overloaded launch weeks and empty weeks before they bite, with launches, vacations, and holidays marked.
→ Every estimated number, discount assumption, and unassigned owner is surfaced in a risks section so nothing invented passes as a baseline.

BOUNDARIES
→ Do not invent a baseline, a date, an owner, or a statistic. Assign owners only from the roles described; otherwise write "owner TBD."
→ Do not pad with generic growth advice or list initiatives that do not serve the North Star.
→ If a required input is missing or vague (the North Star, its baseline, candidate initiatives, constraints, or the quarter dates), ask one focused question instead of guessing.
```

### Quick version

Five lines or fewer, for when speed matters more than rigor.

```text
You are a growth strategist. Build a one quarter plan for {{COMPANY_AND_PRODUCT}} steered on {{NORTH_STAR_METRIC_AND_VALUE}}, given {{CANDIDATE_INITIATIVES}} and {{CONSTRAINTS}}.
Give me 3 to 5 objectives tied to that metric, 1 to 3 growth levers, and numeric key results with a baseline and a stretch target under each.
Rank the initiatives by Urgency plus Feasibility minus Effort and split them into tackle now versus hold off.
Quality bar: every objective and lever must trace to the North Star, and no number may be invented; if a baseline or the quarter dates are missing, ask before guessing.
```

## How to customize

→ `{{NORTH_STAR_METRIC_AND_VALUE}}`: the single metric that captures the value your product delivers, plus today's number. Example: "weekly active teams, currently 1,240." If you do not have one yet, say so and let the prompt help you choose.
→ `{{CANDIDATE_INITIATIVES}}`: the raw list of things the team could do this quarter, even messy. Example: "new onboarding flow, referral program, pricing test, SEO cluster, win back email series."
→ `{{LAST_QUARTER_RETRO}}`: what worked, what was tricky, what you would do differently. This feeds the planning rules that keep the new plan honest.
→ `{{CONSTRAINTS}}`: budget, headcount, and guardrails the plan must respect (max CAC, max discount, minimum margin).
→ `{{QUARTER_AND_DATES}}`: the quarter plus launch dates, vacations, and holidays so the calendar is realistic.

## What good output looks like

→ Three to five objectives, each qualitative and each traceable to the North Star through a driver KPI.
→ One to three growth levers, each with guardrails, and no two levers that fight each other.
→ Key results that are mostly numeric with a stated baseline and a stretch target, so a 0.7 to 0.8 score is the expected outcome.
→ An initiative ranking produced by the Urgency plus Feasibility minus Effort worksheet, split into tackle now and hold off.
→ A week by week calendar that exposes overloaded launch weeks and empty weeks before they become a problem.
→ A risks section that lists every estimated number and discount assumption in the plan, so you can verify each one instead of discovering an invented figure mid quarter.

## Related prompts

→ [Map Growth Loops and Flywheels](./map-growth-loops-and-flywheels.md): map the model that surfaces which weak, high impact steps should become this quarter's levers.
→ [Pick and Test Growth Channels](./pick-and-test-growth-channels.md): when an objective is about acquisition, use this to pick and test the channels that feed it.
→ [Make High Stakes Marketing Decisions](./make-high-stakes-decisions.md): when scoring initiatives leaves a genuinely hard call, run it through a structured decision to commit with confidence.