A Meta creative research system starts with customer evidence: objections, use cases, product facts and questions heard before purchase. It turns each insight into a testable concept, rather than producing many visual variants of the same unsupported promise.
How to make the decision
Use the following sequence for a UAE English-language business. Keep a copy of the inputs and the reason for each decision so a later reviewer can reproduce it. The right answer depends on the actual offer, audience and data available; the example is a planning exercise, not a claimed Megvi client outcome.
| Step | What to do |
|---|---|
| Gather permissioned evidence | Collect anonymized support questions, review themes, sales notes and product documentation. Record where each claim comes from and whether it can be shown publicly. |
| Write a hypothesis | State the audience, objection, message and expected response. A “better creative” hypothesis is too vague to learn from. |
| Brief distinct concepts | Pair each idea with the proof asset it requires: demo footage, specification, founder explanation or verified customer quote. Note production rights. |
| Review outcomes by concept | Look at attention, qualified actions and sales feedback with spend and delivery context. Keep a record of what the next iteration will change. |
Worked example
Illustrative objection: “Will this desk fit a small apartment?” Concept A shows measured dimensions in a room; Concept B walks through the packing and setup. Each addresses a different uncertainty. The brief requires real product measurements and licensed footage before any claim is published.
Where this approach can fail
Customer comments are research inputs, not automatically approved testimonials. Remove personal details and obtain consent before using identifiable quotations or footage.
Review checklist
For each concept, verify source, claim, asset rights, CTA, destination and measurement event. Score evidence on a simple 0–2 scale: absent, plausible but unverified, or documented and approved. A concept with a strong hook but no approved proof remains in research. Retire concepts with no distinct learning question.
Ask an owner to confirm the inputs and assumptions before using this method to change a live campaign, customer record or product recommendation. Keep the original version of any measurement or creative asset so the next review can separate what changed from what only appears to have changed.
What to do next
To separate two concepts drawn from one customer objection, use Comparing Problem-Led and Demonstration-Led Ad Concepts. If you need help applying the method to a live account or store, explore Megvi’s Meta Ads service and bring the current page, campaign or data flow plus the result you want to improve.
Ahmed Elneil
https://megvidigital.comAhmed Elneil is a certified digital marketer and entrepreneur with over 5 years of experience helping brands grow through Search Engine Optimization (SEO), paid advertising, and CRM automation.