Meta is still the widest paid reach available in Egypt and a dependable prospecting engine across the Gulf. What changed is that the measurement stopped being free.
Since iOS tracking restrictions, browser-only pixel data under-reports and the platform fills the gap with modelled conversions. Accounts that never adapted are optimising on a partly imaginary signal. The ones that adapted with a properly deduplicated Conversions API setup have a genuine cost advantage over competitors who did not.
What the work covers
- Conversions API implementation, server-side, deduplicated against the pixel
- Event match quality: the score most accounts ignore and that directly affects delivery
- Campaign structure: prospecting, retargeting and retention separated properly
- Audience strategy: broad versus interest versus lookalike, tested rather than assumed
- Creative testing framework: the variable that moves Meta performance
- Arabic ad copy written natively per market, not translated
- Catalogue and dynamic product ads for e-commerce
- Attribution settings and how they compare to your own back-end numbers
- Frequency and creative fatigue monitoring
- Incrementality testing where budget justifies it
Creative is the lever, not targeting
Meta’s targeting has been getting progressively more automated for years. The practical consequence is that creative now explains most of the variance in performance between two accounts selling the same thing at the same price.
Volume of concepts, not variations
Twenty colour variations of one idea is one test. Five genuinely different angles is five tests. Most accounts run the first and wonder why nothing improves.
Native Arabic, per market
Egyptian and Gulf audiences respond to different register, humour and dialect. One MSA version for all three markets underperforms all three.
Fatigue is faster than you think
In small markets, frequency climbs quickly. A creative refresh cadence planned in advance beats reacting after CPMs have already risen.
Test structurally
One change per test, enough budget to reach significance, and a written record of what you learned. Most accounts test constantly and learn nothing.
Frequently asked
Do we still need the pixel if we have the Conversions API?
Yes, both, deduplicated. The pixel captures browser-side signals the server misses and vice versa. Running only one leaves match quality lower than it needs to be, which shows up as worse delivery.
Meta says we made 40 sales but our system shows 22. Which is right?
Yours. Platform attribution counts view-through and modelled conversions using its own windows. Use your back end as truth and treat platform numbers as a directional optimisation signal, not as revenue reporting.
Should we run Advantage+ campaigns?
For e-commerce with a catalogue and steady conversion volume, they usually perform well. For lead generation, or for accounts with few conversions, manual structure still gives you more control over where the money goes.
How much creative do we need?
More than you think and less than agencies sell. A practical minimum is three to five genuinely different concepts per audience per month, refreshed before frequency climbs rather than after performance drops.