Ad Creative Testing: Write the Next Brief From Results

“Make more like the winner” is not a useful brief if you don’t know why it worked. Turn ad results into one question worth testing.
Ad Creative Testing: Write the Next Brief From Results
Written By
Nitin Mahajan
Published on
November 20, 2026

You open an ad report. One video has a 38% hook rate. Another has 14%. The obvious next brief is “make more like the first one.”

But in a nutrition-brand report we reviewed, the 38% video recorded four purchases and 0.88 ROAS. The 14% video recorded nine purchases and 2.47 ROAS. Ranking by hook rate alone would put the higher-reported-ROAS row second.

That does not make the second video a proven winner. It makes the next question more useful: what should we check before commissioning another batch? Here is a five-step ad creative testing process for turning that question into a specific brief.

By Nitin Mahajan · Last updated 19 November 2026

The source report covers 25 January to 23 February 2026 and has unequal spend and exposure, small purchase counts and unverified attribution settings. Read the complete two-row comparison in the nutrition-brand case study. Here, use it to write a next-test brief; for the wider testing framework, read the ad creative testing guide.

1. Write the decision before the brief

Use a sentence such as: “We need to learn whether showing this everyday problem makes the feature easier to understand.” Choose an outcome relevant to that question and the business. Views can diagnose attention; they do not settle acquisition economics.

2. Separate a concept from a variation

A concept changes the customer situation, argument or demonstration. A variation changes an execution detail inside that idea. If the core explanation is weak, more colour, caption and opening-line versions may polish the same unanswered question.

Record the concept in ordinary language. For the app, “show the task taking too long before introducing the shortcut” gives the creator a clearer job than “make a better hook.”

3. Give the creator useful evidence

Provide the product behaviour, supported claims, relevant customer questions and the intended demonstration. Use AI to help organise the material and draft alternatives, then have someone check the facts and whether each alternative tests the intended idea.

Looking at a category ad can suggest a hypothesis. Its presence in an ad library does not prove it is profitable, and it does not establish that the same idea will work for your product.

4. Keep a comparison you can explain

Where possible, keep the offer, destination, conversion event and relevant delivery conditions comparable. Record differences. Choose the observation window and decision criteria from your economics and available data; avoid universal success thresholds.

Count creator, rights, editing, internal work and media costs. Record whether the test was intended to assess the whole package or one change so the review uses the right language.

5. Turn the result into the next brief

Use a record with the question, change, observed result, limitations and next decision. “The task-first demonstration deserves another test with clearer screen detail” gives the next commission a purpose. “Try five more creators” leaves the problem unresolved.

Stopping can be a useful outcome too. If the offer or product experience needs work, another production batch may be the wrong next investment.

Put the example to work

Open the working decision resource. It includes fictional inputs, editable decision fields and a local export. Keep your own evidence and the unresolved questions beside the decision.

Where QuickAds fits

At QuickAds, the creator search should support the demonstration you want to test. Use relevant work and audience evidence to explain why a candidate suits that brief, then carry the test question into review. AI can help organise candidate and brief information; a person still approves the claims and interprets the outcome.

Review the creator workflow against your next test. The useful output is a reason for the next decision, with the evidence behind it.

Copy the worked example and blank decision sheet.

Questions before the next decision

What is the difference between a creative concept and a variation?
A concept changes the customer situation, argument or demonstration. A variation changes an execution detail within that idea. Record which question the next commission tests.

What should the next creative test brief contain?
State the question, proposed change, supported product facts, comparison, observation window, costs and decision rule. Preserve meaningful delivery differences beside the result.

Keep the next decision connected

Continue with the broader ad creative testing guide and the creator workflow · influencer measurement.

About the author

Nitin Mahajan is the founder of QuickAds, which publishes this guide and sells related software and services.

Nitin Mahajan, founder of QuickAds

Create Ads Like a Pro in Minutes – No Experience Needed!

Discover how easy it is to create scroll-stopping ads with the power of AI and a massive ad library!

Frame 1171276202.Frame 1171276203.
Smiling bald man with glasses wearing a light gray collared shirt against a white background.
Nitin Mahajan
Founder & CEO
Nitin is the CEO of quickads.ai with 20+ years of experience in the field of marketing and advertising. Previously, he was a partner at McKinsey & Co and MD at Accenture, where he has led 20+ marketing transformations.
Transform Your Ads In Seconds - Try QuickAds for Free

Access Our Massive Ad Library & AI Ad Making Tools Today

Image asset.
Image asset.
Image asset.