Research competitor ads and build an original creative test
Find public ad examples, record observable hooks and proof, separate facts from assumptions and create an original video brief for editing.
Make the public research record reproducible
Match the research conditions
Choose the market, advertiser identity and observation date before comparing ads. Save the public source link and visible status. Similar brand names and different countries can produce a misleading sample, so resolve those differences before counting creative patterns.
Describe before judging
Write what the viewer can see and hear: an overhead desk shot, a sorting action, a price card. Keep phrases such as "likely reduces uncertainty" in a separate hypothesis field. Another reviewer should be able to find the same evidence at the recorded timestamp.
Keep outcomes separate
An active ad or repeated concept can suggest something to investigate. It does not reveal private profitability. If a source displays a metric, record its exact label, scope and date rather than translating it into a different measure of success.
Compare observable creative choices
A problem-first opening
One ad begins with cables covering a desk; another starts with a product pack shot. Record how quickly each establishes the problem. Your test could compare those opening structures using the same owned footage, without borrowing the competitor’s script or design.
A demonstration with limits
Note whether the ad visibly shows the organizer holding the items it claims to hold. A fast cut may hide the placement step. For your version, film a continuous action so viewers can inspect what fits, and keep unsupported capacity claims out of the narration.
An offer with a destination
Record the offer as displayed and whether the final action is to shop, compare or request information. Check your own landing destination before filming a similar type of invitation. A creative idea should not send viewers to an expired price or a different product.
Turn one finding into your own video
Collect a comparable sample
Search a public ad library using the advertiser or relevant keyword. Keep five accessible examples with the same research conditions. Record missing fields as unavailable and revisit the source if the ad changes; do not silently replace an old observation with a new one.
Write one original hypothesis
For the desk-organizer example, test whether showing the sorting action earlier improves understanding of the product’s use. Specify the opening change, the unchanged footage and the result you will measure in your own authorized campaign data. Avoid changing the offer at the same time.
Produce and verify the variant
Use verified selling points in CapCut’s Ad script workflow, review the draft and add your own media. Adjust timing and inspect the export against the brief. Name the files by variant and keep the approved offer identical so the comparison remains interpretable.
Frequently Asked Questions
What complete prompt or practical brief can I copy?
Analysis brief: Review five publicly accessible desk-organizer ads from the same market. Record source URLs, advertisers, dates, formats and durations. Describe the 00:00–00:03 opening; timestamp demonstrations, evidence, offers and calls to action. Summarize without copying; separate observations from hypotheses. Do not infer ROAS, conversions, exact spend or targeting from an ad’s age or appearance. Compare when the organizer first solves a visible problem. Propose one original 20-second test using our own product footage: an untidy desk, a visible sorting action and the resulting arrangement. Preserve our specifications and offer. Return a comparison sheet, hypothesis and original shot list.
Where can I find competitor ads?
Can I tell which competitor ads are profitable?
How do I use the research without copying?
What if the sample does not support a clear conclusion?
Edit an original ad from your findings
Bring your approved product facts, owned footage and one creative hypothesis into CapCut, then review the exported variant against the research brief.