Arabic Card Games and the Meta UA Gap: Why "Interested in Card Games" Misses Tarneeb, Trix, and Baloot Spenders

If you run paid UA for an Arabic card-game studio, you already know the targeting on Meta and TikTok is thin. You select "Interested in Card Games" or "Interested in Mobile Gaming," watch CPI come in cheap, and then watch D7 ROAS quietly fall apart. The audience converts — they just don't spend.

This is not a creative problem. It is a data problem. And it is sharper in the Gulf than almost anywhere else.

The interest category is not the player

Meta's "Interested in Card Games" segment is built on inferred signal: people who clicked an ad about cards, liked a Solitaire page, watched a poker reel. None of that is a behavioral statement that the user actually plays — let alone pays inside — a card game.

Now layer the GCC complication. Tarneeb, Trix, Hand, and Baloot are not in Meta's interest taxonomy as named genres. They live as undifferentiated "card games," which means a Riyadh adult who plays Baloot four hours a week, a Cairo teen who watched a Solitaire ad once, and a Dubai office worker who liked a poker meme all share the same targeting label.

When you seed a lookalike from "Interested in Card Games" in Saudi Arabia, you are seeding from that mixed bucket. The lookalike inherits its noise. That is why your Baloot installs convert and don't stick.

What the platforms can't see

Meta and TikTok build interest segments from what users do on Meta and TikTok — likes, watches, clicks, page follows. They do not see what users do inside other apps. They cannot tell you which devices in Riyadh are running a card-game session every evening, which ones have spent inside a card-game app this quarter, and which ones are merely casual genre-curious.

The result is a structural ceiling. Even with perfect creative and a generous Saudi budget, your audience quality is capped by a signal that was never built to identify Arabic card-game spenders.

What behavioral data changes

Onmo Signal scores 300M+ device IDs across 22 countries — including 48M behaviorally-scored Saudi device IDs and 58M+ across the GCC — using actual app-level behavior across 170,000+ apps and 4B+ events. That is a different signal entirely:

• Which devices have card-game apps installed and actively running
• Which of those are top-quartile spenders by recorded in-app spend
• Which devices show high engagement intensity and session frequency
• Which run on premium hardware (a strong proxy for income and ARPU)

For an Arabic card-game UA team, the practical move is to take the GCC card-game spender cohort, upload it to Meta as a Custom Audience in one click, and use it as a lookalike seed — or as a direct retargeting layer. The seed is no longer "people interested in cards." It is people who actually play and pay inside card-game apps in your market.

Why this matters for the studios scaling MENA right now

The MENA card-game category is consolidating: Jawaker, Spyke, and a wave of social-casino and Arabic card studios are all bidding on the same Meta and TikTok inventory with the same targeting signal. Whoever has a cleaner audience seed wins the Saudi spender at a lower CAC.

Behavioral seed quality is the variable that does not show up in your dashboards but determines whether your Smart+/Advantage+ campaigns scale or stall. Algorithms only optimize toward the signal you give them. Give them a noisy interest cohort and you get a noisy lookalike. Give them top-quartile in-app spenders inside your category and the algorithm does what it is supposed to do.

The catalog vs the custom build

Most audience-data platforms make you wait two weeks for a custom build, charge a five-figure minimum, and ship something you cannot inspect. Signal exposes 273 pre-built segments — by category, geography, and quality tier — that you browse in a self-serve catalog and upload to Meta, TikTok, Snapchat, or Google in one click. There is no SDK to integrate, no minimum spend, and right now no upload limits.

You can run an A/B test against your current Meta interest seed this week, not next quarter.

What to test first

If you run UA on a card-game app in the GCC:

1. Pull your current best-performing ad set on Meta. Note the audience seed and the D7 ROAS.
2. Upload a behavioral GCC card-game spender segment from Signal as a Custom Audience.
3. Build a 1% lookalike from the behavioral seed in the same geo.
4. Run the new ad set against your control with the same creative and budget split.
5. Compare match rate, CPI, and D7 ROAS.

The point is not that we win every test. The point is that you should be testing your data source — not just your creatives.

Try it

Signal is free right now — no upload limits, no minimum spend, full self-serve. The free window is closing within ~2 months as we move to a paid tier.

If you are running paid UA for an Arabic card-game title, the test is cheap and the upside is structural. Start at https://signal.onmoai.com.