Define success metrics and guardrails for a feature launch
A primary metric, supporting signals, guardrails, and targets agreed before launch so the result can be judged honestly.
The prompt
Help me define how we will judge whether a feature succeeded, before it launches.
Feature and the problem it solves: [DESCRIPTION]
Target users: [SEGMENT]
Business goal it supports: [e.g. retention, expansion revenue, lower support cost]
Data we can measure today: [EVENTS, DASHBOARDS, TOOLS]
Current baselines: [NUMBERS WITH DATE RANGES - or "unknown"]
Launch approach: [full release / staged rollout / A/B test]
Produce a measurement plan:
1. Primary success metric: one metric, exact definition (numerator, denominator, time window, who is included), and why it reflects the user outcome rather than just usage.
2. Leading indicators (2-3) visible in the first 1-2 weeks.
3. Guardrail metrics (2-4) that must not get worse, e.g. performance, support volume, conversion elsewhere, cannibalization of another feature.
4. Targets: a minimum result that counts as success and a result that would make us roll back or rethink. Base them on my baselines; where none exist, say how to set them after 2 weeks of data.
5. Review schedule: when we check, who decides, what happens at each outcome.
6. Pitfalls specific to this feature: novelty effects, seasonality, mix shift, vanity metrics.
Format as a one-page table plus short notes. Do not invent baselines; mark [ADD BASELINE].
Highlighted [brackets] are blanks to replace with your own details.
Fill in the blanks and use this promptMore Data & Analysis prompts
What to fill in
- DESCRIPTION
- SEGMENT
- e.g. retention, expansion revenue, lower support cost
- EVENTS, DASHBOARDS, TOOLS
- NUMBERS WITH DATE RANGES - or "unknown"
- full release / staged rollout / A/B test
- ADD BASELINE
How this prompt is built
This prompt follows the Few-Shot structure (Few-shot prompting). A few-shot prompt includes one or more worked examples of the input and the output you want, so the model can copy the pattern instead of guessing it.
Related prompts
- Explain my Excel model to a non-technical executiveTurns a complex spreadsheet or model into a plain-English walkthrough an executive can follow, trust, and question.
- Forecast the next few months from historical data in ExcelBuilds a simple, explainable forecast from your past numbers, with trend and seasonality, a range, and the Excel steps to reproduce it.
- Run a fair lending review of loan pricing and approval dataA structured fair lending analysis plan and readout that tests for disparities in approvals, pricing, and exceptions.
- Turn a vague data request into a clear analysis planConverts a fuzzy stakeholder ask into a scoped plan with the real question, data needed, deliverable, and clarifying questions.