Ad experiment template for paid ads
Name the hypothesis, the test, and the decision — so the experiment earns the ROAS, not the guess.
- Google Ads
- Meta Ads
- LinkedIn Ads
- Notion
- Google Sheets
Plan an experiment for our top Meta ad.
- Pulled the adsTop ad: 2.8% CTR, $52 CPA, 4.1:1 ROAS
- Named the hypothesis"Pain-led hook earns 20% more clicks than feature-led"
- Named the testA/B: pain hook vs feature hook, 200k each, 7 days
- Named the decisionIf pain wins >15% CTR, scale to $20k/day; else hold
The plan named the hypothesis (pain-led earns 20% more), the test (A/B 200k each, 7 days), and the decision (if pain wins >15%, scale to $20k/day). The test shipped and the pain hook earned 3.4% CTR vs 2.1% feature — a 62% lift. The decision scaled and the ROAS moved from 4.1:1 to 5.2:1. The template earned the ROAS because it named the hypothesis, not ran the tests.
An ad experiment template is the plan that earns the ROAS. It fixes the hypothesis, the test, and the decision — so the experiment reads as a strategy, not a guess. A paid lead who runs the tests earns the guess; an ad experiment template that names the hypothesis and the test earns the ROAS and the next quarter.
What is an ad experiment template?
A reusable plan shape for the experiment — the hypothesis, the test, the decision — that earns the ROAS and the next quarter. The template fixes the shape so the experiment reads as a strategy, not a guess.
Build the plan
- Hypothesis named — the one bet the experiment earns
- Test named — the one swap the experiment ships
- Decision named — the one call the experiment earns
Ship the experiment
- Name the hypothesis, do not run the tests
- One test per hypothesis, not five
- Set the decision before the test
What hypothesis does an ad experiment template name?
The one bet the experiment earns — the angle, the audience, the offer — that earns the ROAS. The hypothesis is the part that earns the ROAS, because a paid lead who runs the tests earns the guess; a paid lead who names the hypothesis earns the ROAS.
What test does an ad experiment template name?
The one swap the experiment ships — the A/B, the split, the holdout — that earns the ROAS. The test is the part that earns the ROAS, because a hypothesis without the test earns the count; a hypothesis with the test earns the decision and the ROAS.
What decision does an ad experiment template name?
The one call the experiment earns — the scale, the cut, the hold — that earns the next quarter. The decision is the part that earns the next quarter, because a hypothesis without the decision earns the test and loses the quarter; a hypothesis with the decision earns the ROAS.
How the ad experiment template fits your stack
The agent reads your ads and the performance, drafts the hypothesis, the test, and the decision, and ships the plan. It pairs with the creative testing matrix and the creative iteration plan templates.
- Google Ads
- Meta Ads
- LinkedIn Ads
- Notion
- Google Sheets
Who uses this ad experiment template
- Paid leads
- Name the hypothesis, do not run the tests.
- Growth marketers
- One test per hypothesis, not five.
- Founders
- Set the decision before the test.
How to run this ad experiment template in Metaflow
Pull the ads
The hypothesis, the test, the decision.
Name the hypothesis
Agent names the one bet.
Name the test
The one swap the experiment ships.
Name the decision
The one call the experiment earns.
What you provide
- Ads
- Performance data
- Test capacity
What you get back
- Experiment plan
- Hypothesis named
- Test named
- Decision named
Why use this ad experiment template?
Names the hypothesis, does not run the tests
One test per hypothesis, not five
Sets the decision before the test
Pairs with the testing matrix and iteration plan templates
Ad experiment template FAQs
How do you plan an ad experiment?
Name the one hypothesis, the one test, and the one decision. The plan reads as a strategy, not a guess, which is why the experiment earns the ROAS, not the count.
What is an ad experiment?
The plan that earns the ROAS — the hypothesis, the test, the decision. A paid lead who runs the tests earns the guess; a paid lead who names the hypothesis and the test earns the ROAS and the next quarter.
How do you set the decision before the test?
Name the one call the experiment earns — the scale, the cut, the hold. A hypothesis without the decision earns the test and loses the quarter; a hypothesis with the decision earns the ROAS, because the decision is what earns the next quarter.
What makes an ad experiment convert?
A hypothesis that earns the ROAS, a test that earns the decision, and a decision that earns the next quarter. The experiment that runs earns the guess; the experiment that names the hypothesis earns the ROAS.
Similar Google Ads and Meta Ads templates
Key takeaways
- Names the hypothesis, does not run the tests
- One test per hypothesis, not five
- Sets the decision before the test