Build An Offer Scorecard

Score an existing offer across its weakest dimensions to turn vague dissatisfaction into a ranked, evidence-backed fix list.

Pro involved Offer Improvement Recipes

What it adds

Fixes the situation where an offer underperforms but no one can say which part is failing, so effort scatters across cosmetic changes. The AI builder rates the offer against defined dimensions using evidence from the source files, records the reasoning behind every score, and sequences fixes by impact against effort. The outcome is a diagnostic baseline that makes the next improvement obvious and measurable.

What your builder is told to do

The actual instructions, in order.

1
Inspect the source files and establish the scoring dimensions: clarity of promise, strength of the target definition, believability of the proof, pricing logic, risk reversal, offer structure and packaging, and the clarity of the next action.
2
Score each dimension on a fixed scale and write one evidence line per score, citing the specific source file and passage that justifies it. Any score without a citation reverts to unknown.
3
Mark dimensions as unknown where the source files provide no evidence, and list what data would be needed. Unknowns are findings, not failures.
4
Rank the weakest scored dimensions by expected impact on the buying decision, then divide the fixes into structural changes and copy changes so effort can be estimated realistically for each.
5
Select the single highest-impact fix and specify it concretely enough to hand to a builder — what changes, where, and what the before and after states are.
6
Establish the measurement baseline now: record the current conversion rate, average order value, refund rate, and objection mix, so the next change can be judged against numbers rather than impressions.
7
Specify responsive behavior for the scorecard artifact itself: on mobile, one dimension per row with score, evidence line, and rank stacked; on tablet, a two-column grid pairing each dimension with its evidence; on desktop, a full table with dimension, score, evidence citation, fix type, and rank visible in one view.
8
Define tracking and rollback: after the selected fix ships, rescore only the affected dimension against the same evidence standard and compare against the recorded baseline. If the baseline metrics do not move within a full traffic cycle, revert the change and address the next-ranked dimension rather than iterating on a fix the evidence does not support.

Edge cases it handles

The things an agent skips when you only say "build a build an offer scorecard".

Do not score a dimension from intuition; if the source files carry no evidence, mark it unknown and say what would resolve it.
Preserve all factual claims while scoring; this recipe assesses and plans, it does not rewrite the offer.
Where two dimensions score equally low, prefer the structural fix over the copy fix, since copy applied to a broken structure rarely holds.
Avoid scoring against an idealised offer; score against what the evidence in the source files shows buyers responding to.
Do not bundle several fixes into one release, or the rescore cannot attribute any change to a cause.

Definition of done

Your builder is required to check every one of these before reporting the work finished.

Every dimension carries a score, an evidence line, and a citation to a specific source file passage, or is marked unknown with the missing data named.
Weak dimensions are ranked by expected impact and split into structural versus copy fixes.
The single highest-impact fix is specified concretely with defined before and after states.
A measurement baseline of conversion, order value, refund rate, and objection mix is recorded before any change.
Mobile, tablet, and desktop presentations of the scorecard are each specified.
A rescoring method and a revert-and-move-on rule are written down before the first fix ships.
The builder reports which source files it used and lists every dimension left unknown along with the data required to resolve it.

Related offer files

Offer Improvement Recipes

Create An Order Bump

Insert a single low-friction add-on at checkout that completes the purchase without pulling attention off the main decision.

Offer Improvement Recipes

Split Into Tiers

Break a single take-it-or-leave-it offer into tiers that capture buyers at different budgets without gutting the core promise.

Offer Improvement Recipes

Repackage As A Bundle

Merge scattered standalone products into one coherent bundle solving a complete problem, priced below the sum of its parts.

How it works

1

Copy the link

Grab the Markdown blueprint URL for this offer type.
2

Give it to your builder

Paste it into Claude Code, Cursor, Codex, or whatever AI builder is already working in your app.
3

It maps, then builds

Your agent reads the blueprint first, then builds the offer into your app around the product, audience, and stack.

Works with your stack

These blueprints are written to adapt. They tell the agent to detect your framework, match your existing design system, and use your source files instead of guessing the offer.

Need it tighter than that? Customize the offer file and tell it exactly which source docs and stack to use.