Creative Pre-Test Planner

How to use it

1. 1Copy the prompt into ChatGPT, Claude or Gemini.
2. 2Fill the bracketed inputs (campaign, creative routes, decision, budget, timeline).
3. 3Answer its clarifying questions about the decision at stake.
4. 4Iterate: ask it to draft the questionnaire outline for your chosen method.

See it in action

Pre-Test Plan — Buttercup Bakes (excerpt)

The decision: Pick one route for Q4, or kill both if neither clears the bar. Not: polish executions, test media weight.

What to measure:
1. Branding — do viewers know it is Buttercup Bakes without the end card?
2. Comprehension — one sentence back: what is this ad telling you?
3. Distinctiveness — does it feel like the category, or like this brand?
4. Persuasion — shift in likelihood to add to the next basket.

Method recommendation: Monadic online test, two cells (one route each), plus a small qual layer of 8-10 interviews to explain the numbers. Fits four weeks and the budget; sequential exposure would save money but adds order bias.

Cell design: Each respondent sees one route only. Category buyers screened for refrigerated dough or adjacent baking purchases; read parents and solo households separately — the routes target different occasions.

Decision rules (agree these BEFORE fieldwork): The winning route must lead on persuasion AND clear the branding bar. If both fail branding, neither launches as-is.

Traps flagged: testing polish instead of the idea; letting the team reinterpret a losing score as 'polarizing.'

Want me to draft the questionnaire outline for the monadic test?

The prompt

Here's the start of the prompt. Download the free bundle for the full, ready-to-paste version — plus the installable Claude Skill and Custom-GPT instructions.

# Role
You are a senior consumer insights consultant who has pre-tested hundreds of food and beverage campaigns. You believe a pretest exists to make one decision cheaply before media money makes it expensive — and you refuse to run research whose decision rules were not agreed before fieldwork.

# Context I'll provide
- Brand and campaign: [BRAND / CAMPAIGN]
- Creative routes or executions to test: [ROUTES — describe each in 1-2 sentences]
- The decision this test must make: [DECISION e.g. pick a route, go/no-go, fix or kill]
- Budget and timeline: [BUDGET / WEEKS AVAILABLE]
- Target audience: [AUDIENCE]
- Prior creative learnings: [WHAT TESTED WELL OR BADLY BEFORE] (optional)

# Your task

Frequently asked questions

What is a creative pre-test?

A creative pre-test is research run before a campaign launches to check whether the advertising works: whether people recognize the brand, understand the message, find it distinctive, and become more likely to buy. This skill designs the full test — measures, method, cells, and decision rules — so the results settle the route decision instead of starting a new argument.

How is this different from the Concept Screening Survey skill?

Concept screening tests product ideas — would people buy this product at all. Creative pre-testing tests the advertising for a product you have already committed to — does this execution communicate, brand, and persuade. This skill assumes the concept is settled and focuses on choosing or fixing creative routes before media money is spent.

Which AI assistants can run this prompt?

Any capable chat model handles it — ChatGPT, Claude, or Google Gemini. The prompt is model-agnostic, and insights teams often save it as a Custom GPT or a reusable Claude Skill so every pretest in the company starts from the same design discipline instead of a blank page.

What's the most common pretest mistake this helps avoid?

Agreeing the decision rules after the results arrive. When teams see scores first, losing routes get reframed as 'polarizing' and weak branding gets excused as a craft fix. This skill forces pass/fail logic, tie-breakers, and the treatment of negative reactions to be written and circulated before fieldwork, while everyone is still objective.