Two creative variants with very different spend cannot be compared by raw conversions alone. Delivery systems may also have shown them to different people and placements. Start by describing that imbalance, then look for a decision that is robust enough to justify a controlled retest.
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 |
|---|---|
| Build a comparable view | Record spend, impressions, reach, placements, frequency, clicks and qualified actions for each asset over the same period. Note start dates and audience differences. |
| Use rates with counts | Calculate outcomes per thousand impressions or per unit of spend, but show the underlying counts. One conversion from tiny spend is not a stable winner. |
| Inspect delivery context | Check whether one ad received broad delivery while another was confined to a small retargeting pool or was introduced later. |
| Decide on a fair follow-up | If a concept looks promising, give it a controlled test with comparable conditions. If evidence is weak, keep learning rather than scaling immediately. |
Worked example
Synthetic example: A spends AED 4,000 for 20 qualified actions; B spends AED 400 for 3. B has a lower apparent cost per qualified action, but only three outcomes and possibly different delivery. The decision is to test B under comparable audience and placement conditions, not declare it ten times better.
Where this approach can fail
Platform-reported conversions can be modeled or attributed differently from CRM outcomes. Reconcile definitions and lag before treating a small rate difference as a business result.
Review checklist
Archive the date range, attribution setting, asset, placement mix, spend and CRM-qualified outcomes. State what minimum evidence would change the budget decision.
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
For a store-owner summary of what each concept taught, use a monthly creative learning report. Feed the result back into the next evidence-led brief through Building a Meta Ad Creative Research System From Customer Evidence. 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.