Innovation Accounting
"We're making progress" is not a report. Innovation accounting makes early-stage
progress measurable in three steps: establish the baseline, tune toward the ideal, and
decide pivot or persevere. This skill runs all three — and refuses to let a learning
claim survive without a number attached.
The Three Steps
- Baseline. Run an MVP, measure where the engine stands today: conversion,
retention, churn — per cohort, worst-case honest.
- Tune. Iterate to move the baseline toward the ideal one experiment at a time.
Each iteration must move a named metric or it did not happen.
- Decide. When tuning stops working, the pivot-or-persevere meeting gets the
cohort evidence. Not before, not on mood.
Modes
Setup mode (default — new initiative or first time measuring)
- Establish the baseline: for the current product or MVP, list per-cohort values for
acquisition, activation, retention, referral, and revenue (see pirate-metrics-aarr
for bucket definitions). Where a number is unknown, mark UNKNOWN — unknown baselines
are findings, not gaps to fill with averages.
- Name the ideal: what each metric must reach for the engine to sustain the business.
If the ideal cannot be stated, the hypothesis is not ready for accounting.
- Copy
templates/learning-milestone-report.md and define the first 3 learning
milestones.
Report mode (user pastes progress claims or an update draft)
- Read
references/metrics-standards.md first.
- Number every claim of progress. Classify each: MILESTONE (a named metric moved by a
named experiment), INPROGRESS (experiment running, result pending, with a date), or
NOISE (activity without a metric attached — meetings held, features shipped that no
metric watched).
- Report the learning-to-activity ratio. Verdict — "Accountable" (80%+ of claims are
milestones), "Half-accountable" (50–79%), or "Activity theater" (below 50%).
- Convert each NOISE claim into the milestone it should have been.
Decide mode (measurement window closed)
- Assemble: baseline, each tuning experiment and its metric movement, current cohort
trend.
- Apply the rule: improvement across successive cohorts → continue tuning. Flat
cohorts after real tuning → hand off to pivot-or-persevere with the evidence pack.
- Output the evidence pack, not a recommendation to "keep pushing."
Hard Rules
- A learning milestone is falsifiable and metric-bound: "If [experiment], [metric]
will move from [baseline] past [threshold] by [date]."
- Every milestone names the decision it enables. A milestone that changes no decision
is a hobby.
- Cumulative metrics are inadmissible at every step — route through vanity-metric-audit
if they appear.
- The baseline is the worst honest number, not the best cohort. Tuning gets credit for
moving the worst, not the best.
Output Shape (Report mode)
| # |
Claim |
Class |
Metric moved |
Should have been |
Followed by the ratio, verdict, and rewritten milestones. Keep output in the user's
language. Never accept a learning claim whose metric is missing.
1---2name: innovation-accounting3description: Turn startup progress into accountable numbers with Ries's innovation accounting: establish a baseline MVP, tune cohorts toward the ideal, and make pivot-or-persevere decisions on evidence. Use when reporting progress on an unproven product, when 'we're learning a lot' needs to become a measurable statement, for investor or stakeholder updates on early-stage work, or when the user says innovation accounting, learning milestones, baseline metrics, or how do we measure progress before product-market fit. Not for products with established standard metrics like ARR or for vanity-metric cleanup (that is vanity-metric-audit).4---56# Innovation Accounting78"We're making progress" is not a report. Innovation accounting makes early-stage9progress measurable in three steps: establish the baseline, tune toward the ideal, and10decide pivot or persevere. This skill runs all three — and refuses to let a learning11claim survive without a number attached.1213## The Three Steps14151. **Baseline.** Run an MVP, measure where the engine stands today: conversion,16 retention, churn — per cohort, worst-case honest.172. **Tune.** Iterate to move the baseline toward the ideal one experiment at a time.18 Each iteration must move a named metric or it did not happen.193. **Decide.** When tuning stops working, the pivot-or-persevere meeting gets the20 cohort evidence. Not before, not on mood.2122## Modes2324### Setup mode (default — new initiative or first time measuring)25261. Establish the baseline: for the current product or MVP, list per-cohort values for27 acquisition, activation, retention, referral, and revenue (see pirate-metrics-aarr28 for bucket definitions). Where a number is unknown, mark UNKNOWN — unknown baselines29 are findings, not gaps to fill with averages.302. Name the ideal: what each metric must reach for the engine to sustain the business.31 If the ideal cannot be stated, the hypothesis is not ready for accounting.323. Copy `templates/learning-milestone-report.md` and define the first 3 learning33 milestones.3435### Report mode (user pastes progress claims or an update draft)36371. Read `references/metrics-standards.md` first.382. Number every claim of progress. Classify each: MILESTONE (a named metric moved by a39 named experiment), INPROGRESS (experiment running, result pending, with a date), or40 NOISE (activity without a metric attached — meetings held, features shipped that no41 metric watched).423. Report the learning-to-activity ratio. Verdict — "Accountable" (80%+ of claims are43 milestones), "Half-accountable" (50–79%), or "Activity theater" (below 50%).444. Convert each NOISE claim into the milestone it should have been.4546### Decide mode (measurement window closed)47481. Assemble: baseline, each tuning experiment and its metric movement, current cohort49 trend.502. Apply the rule: improvement across successive cohorts → continue tuning. Flat51 cohorts after real tuning → hand off to pivot-or-persevere with the evidence pack.523. Output the evidence pack, not a recommendation to "keep pushing."5354## Hard Rules5556- A learning milestone is falsifiable and metric-bound: "If [experiment], [metric]57 will move from [baseline] past [threshold] by [date]."58- Every milestone names the decision it enables. A milestone that changes no decision59 is a hobby.60- Cumulative metrics are inadmissible at every step — route through vanity-metric-audit61 if they appear.62- The baseline is the worst honest number, not the best cohort. Tuning gets credit for63 moving the worst, not the best.6465## Output Shape (Report mode)6667| # | Claim | Class | Metric moved | Should have been |68|---|-------|-------|--------------|------------------|6970Followed by the ratio, verdict, and rewritten milestones. Keep output in the user's71language. Never accept a learning claim whose metric is missing.
karozi