Is this A/B test template free?

Yes. Enter your email and we’ll send you the CSV instantly — it imports into Google Sheets, Excel, or Numbers at no cost.

How many variables should I test at once?

One. Changing a single element per test keeps the result attributable. If you must vary several things at once, you need a multivariate setup and far more traffic to read it reliably.

How do I know if my A/B test result is significant?

Set a minimum sample size and confidence level (commonly 95%) before launching, then check the numbers against a significance test. Our A/B Test Significance Calculator does this for you — the template links to it.

What should I track for ad creative testing?

Capture the hypothesis, the single variable changed, the primary metric (hook rate, CTR, CVR, CPA, or ROAS), sample size, the winner, and the learning. The template has a column for each.

Why log losing tests?

A losing or inconclusive test is still knowledge. Documenting it stops the team from re-running the same idea and sharpens the next hypothesis — that compounding is the whole point of a tracker.

Free A/B Test Tracker Template (Google Sheets) | Ad & Creative Testing Log - performance_tracking Template

Overview

Run disciplined experiments across ads, landing pages, and email — one row per test, capturing the hypothesis, what changed, the primary metric, sample size, result, and the decision so every test produces a documented learning.

What's Inside

Test ID, channel/surface (Meta, Google, landing page, email), and start/end dates.
Hypothesis and the single variable changed (the “what” and the “why”).
Control vs. variant: the primary metric (CTR, CVR, CPA, ROAS), sample size, and confidence.
Result, winner, % lift, and statistical-significance check.
Decision and learning: ship / iterate / kill, plus the insight to carry into the next test.

How to use this template

How to use the A/B test tracker template:

Make a copy. Import the CSV into Google Sheets (File → Import) or open it in Excel, then save your own copy.
Write the hypothesis. State what you believe and why, e.g. “A benefit-led headline will beat the feature-led control on CTR because it speaks to outcome.”
Change one variable. Log the single element under test (headline, hook, thumbnail, CTA, audience). One variable per test keeps results attributable.
Define success up front. Set the primary metric and the minimum sample size / duration before you launch — not after you peek at results.
Run and record. Enter control vs. variant numbers, then check significance before declaring a winner (use the linked significance calculator).
Decide and document. Mark ship / iterate / kill, the % lift, and the one-line learning.
Compound it. Feed each winner into the next test so your creative and funnel improve continuously instead of resetting.

Pro tips

One variable at a time: Test a single element per experiment so a win is actually attributable to the change.
Decide significance first: Set your sample size and confidence threshold before launch to avoid calling winners on noise.
Log the losers: A “no difference” or losing test is still a documented learning — record it so the team doesn’t re-run it.
Prioritise by impact: Test the highest-leverage elements first (hook, offer, audience) before polishing minor details.

Example uses

Meta creative test

Three hook variations against one control, judged on 3-second view rate and CTR, with the winner promoted to the next round.

Landing page CVR test

Benefit-led vs. feature-led hero, measured on conversion rate with a pre-set sample size and significance check before shipping.

Works with

Google SheetsMicrosoft ExcelApple Numbers