1. One ad set per market
Meta spends where results are cheapest at that moment. If several countries share one ad set, expensive and cheap markets are averaged together and you can't tell which one is working. Separate ad sets let us read each market on its own and move budget on purpose.
In one Arabic-language campaign, we split the audience into four regions: Egypt, Saudi Arabia, the UK and Europe. Compared with the campaign average, Egypt brought 3.2× the leads for the same budget, Europe 0.8×, Saudi Arabia 0.7× and the UK 0.3×. In a single mixed ad set, those differences would have been invisible.
2. Keep the creative the same when comparing markets
All four regions ran the same video. When only the market changes, the difference in results comes from the market, not from the ad. Our rule: test creative inside one market, and test markets with one creative.
3. Language can matter more than country
For the same project and budget, Arabic-speaking expats in Europe brought 2.2× the leads of English-language international buyers. The language of the ad decides who you really reach, so we treat Arabic-speaking expats as their own audience instead of folding them into an English campaign.
4. Expect very different auction prices
Reaching people in the UK cost about 8.8× more than reaching people in Egypt — same campaign, same video. A higher cost per lead in an expensive market is not automatically a bad result. We compare it with other campaigns in the same market, and with how many of those leads sales can actually qualify.
5. One campaign per country for native-language markets
For Czech and Polish buyers, we built one campaign per country, with video creative and broad targeting that uses investment and real-estate interests as signals (Advantage+ audience). Compared with earlier campaigns in the same markets, they brought 2.1× and 2.4× the leads for the same budget.
6. Stop weak tests early
Not every market works. Denmark brought only 2 leads across two campaigns, and a UK test was stopped within two days with none. We decide before launch how much a test may spend before we judge it — and we stick to it.
Checklist
- One ad set per market, so each one can be measured on its own.
- The same creative across markets when you compare them.
- Separate audiences for each language, not just each country.
- Judge each market against its own auction cost.
- A spend limit for every test, set before launch.
- Lead quality checked with sales before you scale.