START WITH THE PRODUCT STEPS
Read START-HERE.txt first. It states the actual source acquisition, access/privacy limits, manual field mapping and exact next prompt. The original file templates and synthetic examples below are supporting review aids. Some paths need separately supplied documents or source text; a blank worksheet is not a native export or completed integration.

M03 - Plan a capped Meta creative experiment

Status: Proposed / not tested

WHAT THIS KIT CONTAINS
Exact input scaffolds and synthetic examples for an offline pilot. Header, JSON-shape and fixture-consistency checks validate these files only; they do not test an assistant, product integration, legal decision, PDF inspection or live workflow.

PREREQUISITES
1. Two attorney-approved creative IDs and one defined business question/primary metric
2. Staff-supplied all-in budget, experiment portion, currency, dates, fees/taxes and stop authority
3. Current account eligibility evidence and a privacy-safe measurement path
4. Existing aggregate result export only if reading results; no personal lead/case data

SETUP
1. Use the separate synthetic example inputs first. Fill a copy of the blank template or input checklist for your own preparation; do not mix fixture information with real records.
2. Give only the selected synthetic input sections/files and the prompt below to an existing firm-approved private workspace, or perform the clerical comparison manually. Keep expected-result notes separate from the input. For planning checklists, leave unknown requirements explicit instead of inventing them.
3. Compare the draft output with the expected fixture result. A reviewer checks original evidence and all exceptions before any separately approved action. Retain the version, scope and source references with the review.

INPUT RULES
This recipe takes a document brief and optionally a platform-native report. No universal Meta export schema or import blueprint is defined. Use the checklist to preserve the report’s actual column labels and confidence/result fields, including absent/unknown values.

FILES
- INPUT-CHECKLIST.txt: Input preparation and review instructions
- SYNTHETIC-BRIEF-AND-RESULTS.txt: Separate synthetic planning and result-reading examples

COPYABLE OFFLINE PROMPT
Draft an offline experiment brief comparing only the two approved Meta creative IDs supplied. State one question, the one variable changed, primary metric and source, fixed review dates, allocation assumptions, and a maximum experiment cost that does not exceed our supplied allowance. Reserve room for any supplied fees/taxes; if these or platform eligibility requirements are unknown, mark launch blocked pending verification. Include a stop rule for the supplied cap, a broken measurement path, a policy/privacy issue or the planned end date. Staff alone execute stops. Explain what must stay constant and why an ordinary unequal-delivery report is not automatically a randomized A/B test. Give a result template with 'inconclusive' and no automatic scaling. If a results export is included, report observed arithmetic and the platform's actual result fields without inventing confidence, sample size, statistical power or a winner. Do not start a test, change spend or send conversion data.

INPUTS
(1) two already attorney-approved creative IDs; (2) one business question and one primary metric definition, including the source and attribution window; (3) the firm's total all-in allowance, currency, maximum experiment portion, dates, expected fees/taxes if known, reviewer and stop authority; (4) current account screenshots for eligible experiment options and any displayed requirements; (5) aggregate historical volume if available, or “unknown”; (6) optionally an existing experiment export with variant, spend, primary outcomes, dates and the platform's actual confidence/result fields. Do not provide raw leads or case outcomes tied to people.

OUTPUT
an approval-ready experiment brief and results-reading template that can return “inconclusive.

Use only these synthetic or firm-approved inputs and preserve source row/page IDs. Keep missing or conflicting evidence unresolved. Return a draft for the named reviewer. Do not send messages, connect accounts, change records or permissions, move money, or make legal decisions. Keep source files unchanged. List missing prerequisites before continuing.

EXPECTED SYNTHETIC FIXTURE RESULT
Planning fixture: total allowance GBP 900, experiment allowance GBP 200 all-in, unknown fees/taxes and missing eligibility capture. Launch blocked. Results fixture: A GBP 100 / 2 outcomes = GBP 50.00; B GBP 100 / 3 = GBP 33.33. Total GBP 200 reaches the ceiling. Confidence is not supplied; result inconclusive, no winner or scaling.

LIMITS
- No account setup, launch, budget changes, stops, Pixel/CAPI setup, conversion uploads or automated scaling. A planning cap does not prove Meta billing controls enforce an all-in ceiling. Counts alone do not establish a winner.
- Synthetic data only in the delivered files. No credentials, real client records, live-action scripts, API connection or Make blueprint is included.
- Real use requires firm-approved access, confidentiality, processing terms, retention, product access and reviewer authority. Pseudonymous IDs alone do not anonymize real records.
