The Offer and Pricing tools: let the AI suggest, let the person pick, let code do the arithmetic

The problem behind it

After a market is chosen, the business cycle needs three things: an offer, a price, and a list of the riskiest guesses behind them. In October these were four tiles on the Strategy page marked "Coming in Phase 4". On 23 October I asked for them to become working tools.

The danger with AI here is plain. Ask a model for a price and it will give you a confident one. It has no idea what your cost to deliver is, or what a customer costs you to win.

How it works

The tools follow the Hormozi books I rely on ($100M Offers for the offer, $100M Money Models for the ladder). From the help article:

1. Offer. From the chosen market, the AI writes the dream outcome with every obstacle in four groups, a solution for each, how to deliver it, then three different offers (stack, guarantee, bonuses, scarcity, urgency, name) scored on the value equation. Yukti suggests one. A person picks.
2. Pricing. From the picked offer, the AI works out what the result is worth to the buyer and what buyers pay today, then writes three price options with an anchor and what could follow later. Unit economics use only two numbers that you type: cost to deliver and cost to win a customer.
3. Assumptions register. The riskiest guesses behind market, offer and price, each with a cheap test (a few thousand rupees), the signal that proves it true or false, and a status you update.
4. Strategy pack. No AI. The chosen market card, offer, price and five riskiest assumptions on one printable page, with an export. Each block says ready, draft or not done yet.

The decisions

Code decides the arithmetic and the order. Value score is dream outcome times likelihood divided by time times effort, from the AI's scores of 1 to 5. Offers are ranked by it. Margin, lifetime value, payback and "does the first sale cover the cost to win" come only from numbers the person typed. A missing number is named, not guessed. The register is ordered with proved-false first, then open ones by risk.

Yukti advises, the person picks. Moving on from the Offer needs a picked offer. That is the only guard. Skipping Pricing or Assumptions is allowed and the pack says what is missing. Nothing blocks tasks, flows or approvals.

One machine, not two. The Strategy tools run on the same tables and runner as Marketing Ops, with new run kinds. Building a second copy would have broken our rule against duplicate homes for the same thing.

Guesses are marked. AI guesses carry a "to verify" mark.

Practitioners first. The prompts follow named practitioners rather than generic advice. That is a choice, not a proof of quality. I have not tested whether the offers it writes convert.

The part that changed in November

The content tests exposed a weak link between offers and writers. The Lab's human review found that Conversion was the weakest score and that offers and calls to action were muddled in four of five review batches: a board-game session, a Studio app, workshops, coaching and a free business check all mixed together. So on 3 November I decided the offer needed a compact "offer card" with a value ladder: one line per rung (name, who, price, next step and link), at most five rungs, about 900 characters, so a writer can pick the rung that fits a piece. I also said the offer should read from the Offer and Pricing tools. On 4 November both tools moved into the main menu under Strategy.

The retention thinking also changed what the ladder is for. The ladder is no longer just a path to the core offer. After it come retention rungs: membership, community, advanced levels, renewals, alumni. The Pricing tool is where lifetime value, cost to win, the ratio of the two, payback months and churn assumptions per rung belong.

What went wrong

On 31 October the Pricing tool was noted as "exists but has NOT been reviewed yet", and the Market Finder did not use it. I do not have a recorded review. I would not call the Pricing tool done.

What's next

The offer card is built into the content flow plan (the writer receives it in a short brief; the judge scores Conversion against it). The retention rungs and LTV panel are specified. A review of the Pricing tool to 9.5 out of 10 is on my goals list, number six.

Numbers at a glance