---
title: Experiment Planner — Traction, Incrementality, A/B Test Design | RGM®
url: https://realgrowthmatters.com/tools/experiment-planner/
updated: 2026-06-10
source_html: https://realgrowthmatters.com/tools/experiment-planner/
---

### Step 1 — Pick your test type

Test type

A/B test on conversion rate (CRO, landing page, ad creative)
Incrementality test (geo holdout for paid channel)
Traction test (rapid early-stage validation)

What you're testing (descriptive)

### A/B conversion test inputs

Baseline conversion rate (%)

Your current CVR — what control performs at

Minimum detectable effect (MDE %)

Smallest relative lift you care about (10% = lift baseline by 10% relative)

Statistical significance threshold (%)

95% standard; 90% acceptable for low-risk tests

Statistical power (%)

80% standard; 90% for higher-stakes tests

Daily traffic to test surface (visitors)

Test variants (including control)

### Incrementality test inputs

Channel being tested

Meta (Facebook + Instagram)
TikTok
Google Search
YouTube
Programmatic Display
Connected TV
Audio / Podcast

Expected lift % (your hypothesis)

% of conversions you expect to be incremental

Geo markets available for test

DMAs in US; metros internationally. Need 20+ for good power.

Monthly conversions in control region

Lower = need longer test for statistical power

Channel monthly spend ($)

Test design

Geo holdout (50/50 split)
Matched market pairs
PSA ghost ads (platform-native)
Switchback (on/off in same market)

### Traction test inputs

Test budget ($)

Smallest meaningful spend to validate an idea

Channel

Meta (most signal-rich for testing)
TikTok
Google Search
LinkedIn (B2B)
Email / SMS to owned audience

What you're trying to learn

Whether an audience converts
Which message resonates
Whether an offer drives action
Which ad format works (video vs static, etc.)
Whether a creator amplifies our message

Number of variants to test

## Test plan

**Test design reasoning**

### Operating checklist

[Open blank Google Sheet ↗](https://sheets.new)
[Open blank Google Doc ↗](https://docs.new)

After download: open the file above, click cell A1 (or click into the doc), and paste (Cmd/Ctrl+V).

Free training · Experimentation discipline

**From hunches to validated learning**

A good test design saves months. A bad one costs months. RGM training walks through hypothesis writing, prioritization (ICE/RICE), power analysis, and how to read results honestly. Free, no signup.

[Start the training →](/training/)

### Pair with these reads

[**Incrementality testing**Geo holdouts, PSA tests, matched markets.](/learn/frameworks/incrementality-testing/)
[**ICE prioritization**Pick which experiments to run.](/learn/frameworks/ice-prioritization-scoring/)
[**Lean Startup**The discipline behind validated learning.](/learn/frameworks/lean-startup-build-measure-learn/)

Methodology & sources

1. Lewis, R. A. & Rao, J. M. (2015). "The Unfavorable Economics of Measuring the Returns to Advertising." Microsoft Research / NBER.
2. Standard sample-size formulas via Evan Miller's A/B test calculator methodology.
3. Meta Conversion Lift documentation.
4. Google Conversion Lift documentation.
5. Common Thread Collective & Tinuiti incrementality testing case studies.
6. Croll, A. & Yoskovitz, B. (2013). *Lean Analytics*. O'Reilly.