AnalyticsAttribution, incrementality and testsAdvancedDesigns a test

Incrementality test design: holdout or geo experiment

Designs an incrementality test plan with hypotheses, units, duration decisions and pre-registered success criteria, without inventing sample sizes.

For: Performance marketers, Marketing analysts, Growth marketers · Works with: Any capable chat model (ChatGPT, Claude, Gemini, others)

Public beta. This prompt was drafted with AI assistance and checked by automated rules, but it has not been reviewed or tested by a person yet. Treat it as a starting point and check the output. It is hidden from search engines while in beta. Use the “Was this prompt useful?” box to tell us what works.

Your prompt

Act as a marketing-science consultant. Design an incrementality test for paid social prospecting with the goal of answering: "How many of our paid-social conversions would have happened without the ads?". Context: business model DTC skincare with repeat purchase; typical conversion lag mostly within 7 days; available data order data by postcode, platform reports; constraints can pause ads in some regions for 4 weeks; cannot hold out the whole country; approximate weekly conversions about 600. Deliver: 1. Recommended design (user-level holdout, geo holdout, matched markets, ghost ads/PSAs if applicable, or pre/post) with reasons and the biggest threats to validity for my context. 2. Hypothesis, primary metric, secondary metrics and the exact test and control definitions. 3. The inputs that determine duration and sample size (baseline rate, minimum detectable effect, variance, conversion lag), and instructions for calculating them with a power calculator; do not state a required sample size as fact. 4. Pre-registration checklist: success criteria, stopping rules, what we will do for each outcome (positive, null, negative), and what will be logged. 5. Risks: contamination, spillover, seasonality, platform delivery changes, changes during the test. 6. How to read the result, including uncertainty and what would not be concluded. 7. What data is missing. Rules: do not invent baseline numbers, effect sizes or results. State assumptions; if information is missing or not provided, list questions. Verify the plan is consistent with the constraints I gave. Distinguish correlation from causation: the design exists to support causal inference, and says where it cannot.

Customize

The fields start with example values so you can see how the prompt works. Everything stays in your browser.

What is being tested.

Business question.

Context.

Delay to conversion.

What you have.

Practical limits.

Rough baseline for context.

Example of what to expect

Illustrative only. It describes the kind of result this prompt aims for; real output varies by tool, model and run.

A recommended design with threats to validity, hypothesis and metrics, inputs that determine sample size and duration with calculator guidance, a pre-registration checklist, risks and reading guidance. (Illustrative.)

Expected format: Design, hypothesis and metrics, power inputs, pre-registration, risks.

How to use it

  1. Pre-register your decision rules.
  2. Calculate power with a proper tool.
  3. Avoid changing campaigns during the test.

Limitations

  • Small businesses may lack the volume for precise results.
  • Spillover and contamination can bias geo tests.

Platform notes

Any capable chat model (ChatGPT, Claude, Gemini, others)

Official docs read · checked 2026-10-11

Works in any capable chat model. Paste a small anonymised sample and recompute the key numbers yourself: chat models can misread columns or miscalculate, and they cannot run your statistical tests reliably.

General notes on Any capable chat model (ChatGPT, Claude, Gemini, others)
  • These prompts are plain text and work in any chat assistant. For analytics prompts, paste a small, anonymised export; a chat model can misread columns or miscalculate, so recompute key numbers yourself.

Source, license and attribution

Origin
Original by MarketerTools
Publisher
MarketerTools
License
Original work by MarketerTools, free to copy and use. Informed by the linked documentation; no third-party text is reproduced.
Platform assumptions checked
2026-10-11

Documentation and references behind this prompt

These informed the structure and the platform notes. A reference is not a license, and no third-party prompt text is copied here.

Spotted an attribution error, or are you a source owner with a request? Use the chat button at the bottom right and quote prompt ID ana-034. We will correct or remove it.

Related prompts

  • AnalyticsIntermediateAnalyses your data

    A/B test result evaluation with statistical guardrails

    Interprets an A/B test from raw counts with checks for sample-ratio mismatch, peeking and practical significance, without inventing a significance result.

    Any capable chat model (ChatGPT, Claude, Gemini, others) · Google Sheets / Excel