Mike KarrTo recreate the useful thinking behind a competitor ad, record the customer question it answers, then build a fresh answer using your own product evidence. Start a new brief before writing a script. This guide takes you from a public reference to an original 25-second video concept, a production handoff and a test you can interpret.
The worked example is hypothetical. It does not reproduce a real advertiser’s script or establish performance. A visible or long-running ad is a research reference; its profitability remains unknown without verified results.
1. Record the observation before the idea
Choose one public reference and save its URL, advertiser, date checked, market/filter and destination. Watch only what is available and mark missing information as unknown. For a research method, use our competitor ad analysis guide; this page begins with the production decision that follows.
Separate three notes: observation — the ad opens with a product being packed; interpretation — it may address uncertainty about capacity; new hypothesis — showing the exact contents of our pouch may help shoppers judge fit. Neither interpretation is a measured result.
Copy this reference record: source URL | advertiser | checked date | visible opening | customer question | proof shown | destination promise | unknowns | new product fact | own proof file. Write observations in plain words rather than copying dialogue.
2. Decide what belongs in the new brief
Use this as an editorial production matrix. It is not a legal clearance score. A high-level question can guide independent writing; distinctive execution needs its own review.
| Reference element | Decision for your brief | Concrete replacement |
|---|---|---|
| Customer question | Investigate it with your own buyers. | Use a question from your product support notes. |
| General format: question → demonstration | Plan an independent demonstration. | Show your feature, setting and proof sequence. |
| Dialogue, signature joke, scene sequence | Write from a blank brief. | New phrasing, opening situation and shot order. |
| Footage, music, graphics, face or voice | Use assets you own or have permission to use. | Own footage, appropriately licensed audio and authorised presenter. |
| Offer or customer result | Verify against your current product evidence. | Your actual terms; omit unsupported results. |
The U.S. Copyright Office distinguishes ideas from their expression. It also explains that fair use has no fixed permitted percentage. Changing colours or replacing a logo does not establish clearance. These are U.S. references, not a worldwide rule or an assessment of your ad. See what copyright protects and permission and fair use. Review rights and local requirements for the finished execution.
3. A worked example: from question to original script
Assume a reference answers “what fits inside?” for a travel product. Keep only that research question. Our invented product is a cable pouch with two elastic loops and a zipped pocket. Before using the copy, confirm those details on the real item and film the actual contents.
Fresh brief: audience — people packing a daily work bag; question — where do the cable and adapter go?; evidence — an uninterrupted packing shot; promise — see the layout and dimensions; destination — the pouch product page. The presenter explains the layout without claiming to be a customer.
- 0–4 s: overhead view of a work bag and loose accessories. Presenter: “Before you pack your work bag, give these small things a place.”
- 4–10 s: show the open pouch with labels on its two loops and pocket. “Here are two elastic loops and a zipped pocket.”
- 10–19 s: pack your own cable and adapter in one continuous shot. “Watch where this cable and adapter fit.” Keep the items visible and identify their sizes where relevant.
- 19–25 s: show the closed pouch beside a ruler. “Check the dimensions against what you carry.” Send the viewer to the matching product details.
Timing is an editing plan. Rehearse the local-language script and adjust it to the available footage. Do not keep a borrowed punchline, distinctive scene or unsupported claim simply to match the reference’s length.
4. Hand the brief to a UGC/avatar production workflow
- Prepare inputs: product URL or brief, buyer question, verified feature, own footage, destination and authorised presenter. Keep competitor media out of the asset folder unless its use is specifically cleared.
- Select a format: start from a ready UGC/avatar template in Kreopatra. Use a narrator to introduce the product detail. Build the concept from your fresh script; this workflow does not require importing or automatically cloning the source ad.
- Plan the proof insert: pair the presenter with authentic product footage in your edit when physical behaviour matters. Generated product details need review against the real item. Keep measurements and limitations readable.
- Review the output: check pronunciation, captions, product appearance and the destination promise. Brand narration should not imply a fabricated personal purchase, review or founder identity.
Production handoff: product URL | audience question | verified feature | new hook | proof filename and owner | presenter permission | audio licence | captions | destination | reviewer | remaining questions. Record actual filenames and permissions before export.
For alternative structures, use the video ad template library. Kreopatra’s core workflow is AI video ad generation from ready UGC/avatar templates; competitor research supplies an input to that workflow.
5. Review originality and run a focused test
Review the new ad beside the written brief first, then beside the reference. Can you explain each line with your product evidence? Does the ad have its own language, setting, presenter and proof? If the source’s recognisable execution remains central, rewrite or obtain the needed rights. Passing this review is an editorial decision, not a guarantee of legal clearance.
Produce a second opening around the same proof shot: “What goes in the small pocket of your work bag?” Keep the product demonstration, offer and destination constant. This asks whether a direct question or a routine opening brings more qualified visits.
- Choose a spend limit, evaluation window and business outcome from your own baseline before launch.
- Record audience, placement, delivery and spend for each variation. Unequal delivery can make a comparison inconclusive.
- Use attention and click measures for diagnosis, then inspect the intended conversion. If visits increase without qualified actions, review the promise and landing page before producing more variants.
Test record: concept | opening A/B | unchanged proof | destination | spend | intended outcome | observed outcome | uncertainty | next change. An inconclusive result is a valid record; public competitor visibility cannot fill the missing data.
Frequently asked questions
Can I copy an ad if I change the product and logo?
Those edits do not establish permission or originality. Write independently, use assets with the necessary rights and review the finished execution for your market.
How much of a competitor ad should I change?
There is no percentage rule that clears an ad. Start from the customer question and your product facts, then write and produce independently. Seek advice on a specific rights question.
Can an AI avatar deliver the new script?
Use an authorised presenter and accurate brand narration. Pair physical product claims with real evidence and avoid fictional testimonials or an implied relationship with the competitor.
Does a long-running reference mean it is a winner?
No. Public visibility does not establish spend, profitability or conversion performance. Treat the structure as a hypothesis and evaluate your own ad against a defined outcome.