THE GROWTH & ANALYTICS CARD · measure what matters; scale what works
WAKES WHEN a funnel is diagnosed, an experiment planned or read, or a test is about to stop early
THE LOOP
funnel · each step, the biggest leak
size --base --mde · the sample, before
run the test to its planned sample
test --a --b --plan · the verdict
scale the winner; write what was learned
NEVER
Call a test before its planned sample
Call chance a win
Count vanity numbers as revenue
IT MAKES funnel — each step's conversion, the whole, the biggest leak; test — rates, lift, z, p, the interval, SIGNIFICANT or NOT YET; size — the visitors each arm needs.
THE STANDARD measure what matters; the biggest leak first; the sample is set before the test; no first good day; the interval is the honest number; arithmetic, not judgement.
PROOF SELFTEST 11/11 · witnessed in Claude 2026-10-10 · Edition 1.
THE SEAL 98 of 100 · LEGENDARY · WITNESSED · Trigger 5 · Machinery 5 · Law 5 · Portability 5 · Proof 5
IT CAN BE WRONG an AI reading a text file: it can be wrong and can invent a number, a rule, a name or a date; check what matters against your own records and the source; the risk is yours. THE-CLAUSES.md rides in the packet and is part of the licence.
THE PUNCH · SHA-256 OF SKILL.MD · 1F77004DE6B9474BEC2E0E69
CITE THIS CARDTHE GROWTH & ANALYTICS CARD, Edition 1. Lund Studio LLC, Westminster, Colorado. https://lundr.tech/?view=growth-analytics-card. Punch 1F77004DE6B9474BEC2E0E69. Retrieved 2026-10-10.
What it is.
Brings arithmetic to growth decisions: funnel reads the counts at each step (visit, signup, trial, paid) and gives each step's conversion, the whole funnel's and the biggest leak — the step that loses the most people — and refuses a step larger than the one before; test reads an A/B result as conversions over visitors for each arm and gives the rates, the lift, the z score, the p value and the interval of the difference, with the verdict SIGNIFICANT or NOT YET, and holds the verdict when the planned sample has not been reached; size gives the visitors each arm needs to see a given difference from a base rate at a chosen alpha and power.
It carries a machine that shows its working: funnel — each step's conversion, the whole, the biggest leak; test — rates, lift, z, p, the interval, SIGNIFICANT or NOT YET; size — the visitors each arm needs — 11 checks in its selftest, every command on the record. It will never call a test before its planned sample; never call chance a win; never count vanity numbers as revenue.
Six laws, written into the card.
They are in the card itself, so the AI follows them.
MEASURE WHAT MATTERSRevenue drivers, not page views and followers; the funnel is read in people, step by step.
THE BIGGEST LEAK FIRSTThe step that loses the most people is named; that is where the next experiment goes.
THE SAMPLE IS SET BEFORE THE TESTThe visitors each arm needs are computed from the base rate and the difference worth finding, before a visitor is counted.
NO FIRST GOOD DAYA test is decided on its planned sample; a significant result before it is NOT YET, said plainly.
THE INTERVAL IS THE HONEST NUMBEREvery verdict carries the interval of the difference, so a small lift is not sold as a large one.
ARITHMETIC, NOT JUDGEMENTThe card computes on the counts it is given; what to scale is the owner's call.
What rides in the card.
In the card’s own voice, with a machine that shows its working.
THE GUIDEreferences/THE-DISCIPLINE.md
THE MACHINEgrowth.py · three commandsfunnel — each step's conversion, the whole, the biggest leak; test — rates, lift, z, p, the interval, SIGNIFICANT or NOT YET; size — the visitors each arm needs. Every command prints its working and refuses a bad input; 11 checks in its selftest, every command on the record.
What it does once your AI has it.
funnel Each step's conversion, the whole, the biggest leak
test Rates, lift, z, p, the interval, SIGNIFICANT or NOT YET
size The visitors each arm needs
On the record The funnel read with its biggest leak at the first step; the sample for a one-point lift on a five percent base computed at 8,158 a side; an early, small test held at NOT YET with the plan named; the full test read SIGNIFICANT with its interval clear of zero.
What it will never do.
It will never call a test before its planned sample; never call chance a win; never count vanity numbers as revenue.
The test is the two-proportion z test with its normal approximation, sound when each arm has a few dozen conversions; for very small counts read it as a rough guide. It assumes visitors are assigned at random and counted once. Running many tests at once, or reading one again and again, raises the chance of a false win; the planned sample is the guard the card keeps. Not financial advice. It is an AI reading a text file: it can be wrong, so check what matters against your own records and the source.
What’s in the buy.
CARD.md — the whole card: who it is, its laws, when it says no, what it carries · SKILL.md — the door your AI reads first · START-HERE.md — the first five minutes, for you
THE-DISCIPLINE.md — the guide in references/, in the card's own voice
scripts/growth.py — the machine: funnel · test · size, standard library only, selftest 11/11
WITNESS.md + SELFTEST.md — every command on the record, as it printed · MANIFEST.md, every file hashed · AUDIT.md, the twelve checks
MY-LEDGER.md — the card's memory, on your own machine · TIN.md · LICENSE-NOTE.md · THE-CLAUSES.md · HOSTS.md
card.svg + card-machine.svg + card.json · card.html, the card as one offline file · the doors for ChatGPT, Gemini, Cursor, Claude Code and MCP hosts (rules/, system/, mcp/)
PROOF · SELFTEST 11/11 · 3 COMMANDS ON THE RECORD · WITNESSED IN CLAUDE 2026-10-10 · EDITION 1 · PUNCH 1F77004DE6B9474BEC2E0E69
ON EVERY DOOR · A CARD FOR YOUR AI, A FILE FOR YOUR COMPUTER
In Claude: turn on code execution under Settings › Capabilities, then Customize › Skills › + › Create skill, and upload the zip as it is — START-HERE.md in the box walks you through it, and tells you how to keep the ledger. In ChatGPT, Gemini and the rest: a Project, a GPT or a Gem with the card’s files attached and the short door in the box pasted as its instructions. The words run anywhere an AI reads Markdown; the machines run wherever the host runs Python. Witnessed in Claude only. The host marks →
Take it.
Nº 1091 PROPOSED — LAUNCHING SOON
Sold as is under the term printed. A card instructs an AI you operate; results depend on your model, your files and your judgement. Not medical, legal, tax or financial advice, never the person’s boss, and not a licence to practise. The terms →
One payment. The file is yours to keep and keeps working at its edition; new editions issued in the year after you buy reach you by email as replacement files. Tax is added at checkout where it applies. Nothing auto-renews.
A TERM, NOT A SUBSCRIPTION: NOTHING RENEWS ITSELF AND NOBODY IS CHARGED AGAIN. WHEN A TERM ENDS THE FILES STAY ON YOUR COMPUTER AND STAY READABLE — FREEZE, NOT LOSS.
THE MACHINE · IN THE OPEN · FROM THE WITNESSED RUN OF 2026-10-10
growth.py, every command showing its working.
A small Python program, standard library only. Your AI runs it; you can too. Every block below is pasted from the card’s own WITNESS.md, as printed.
funnel
each step's conversion, the whole, the biggest leak
test
rates, lift, z, p, the interval, SIGNIFICANT or NOT YET
size
the visitors each arm needs
selftest
11 checks that prove the machine; it must print 11/11
funnel · EACH STEP'S CONVERSION, THE WHOLE, THE BIGGEST LEAK
$ python3 scripts/growth.py funnel visit=10000,signup=1200,trial=600,paid=90
visit → signup 12.00% lost 8,800
signup → trial 50.00% lost 600
trial → paid 15.00% lost 510
the whole: 0.90% of the 10,000 at visit reach paid · the biggest leak: visit → signup (8,800 lost)
size · THE VISITORS EACH ARM NEEDS
$ python3 scripts/growth.py size --base 0.05 --mde 0.01
8,158 visitors each arm (16,316 in all) to see 5.00% → 6.00% at alpha 0.05 with power 0.8
test · RATES, LIFT, Z, P, THE INTERVAL, SIGNIFICANT OR NOT YET
$ python3 scripts/growth.py test --a 20/1000 --b 24/1000 --plan 8158
A 2.00% (20/1000) · B 2.40% (24/1000) · lift +20.0%
z 0.61 · p 0.5420 · 95% interval of B − A: -0.89 to +1.69 points
the planned sample is 8,158 a side and the test has 1,000; decide on the plan, not on the first good day
NOT YET — no difference that the counts can tell from chance
test · RATES, LIFT, Z, P, THE INTERVAL, SIGNIFICANT OR NOT YET
$ python3 scripts/growth.py test --a 410/8200 --b 505/8210 --plan 8158
A 5.00% (410/8200) · B 6.15% (505/8210) · lift +23.0%
z 3.21 · p 0.0013 · 95% interval of B − A: +0.45 to +1.85 points
SIGNIFICANT at 0.05
THE NUMBERS, NAMES AND DATES IN THESE RUNS ARE THE STUDIO’S OWN EXAMPLES, INVENTED FOR THE RECORD; NO REAL COMPANY, PERSON OR PROPERTY IS NAMED.
QUESTIONS AT THE TABLE
Asked before the wire.
What does it need from me?
Funnel · each step, the biggest leak. It asks first for the counts at each step, and for a test the base rate, the difference worth finding and the two arms' results.
Does it send anything anywhere?
No. The machine is a standard-library Python program that runs on your own computer and sends nothing anywhere; the words run wherever an AI reads Markdown. No server, no account, no key.
What is on the record?
The self-test (11/11) and every command run on real example inputs, pasted into WITNESS.md as they printed; the punch on the face is the SHA-256 of SKILL.md, so a downloaded copy can be checked against this room.
Can it be wrong?
Yes — it is an AI reading a text file. The test is the two-proportion z test with its normal approximation, sound when each arm has a few dozen conversions; for very small counts read it as a rough guide. It assumes visitors are assigned at random and counted once. Running many tests at once, or reading one again and again, raises the chance of a false win; the planned sample is the guard the card keeps. Not financial advice.
What will it never do?
It will never call a test before its planned sample; never call chance a win; never count vanity numbers as revenue.
