Why 122 Million Puzzle Game Players Beat Meta Interest Targeting

If you run UA for a mobile game, you have probably built campaigns targeting Meta's "interested in puzzle games" audience. And you have probably noticed the results are... fine. Not great. Just fine.

Here is why: Meta's interest categories do not mean what you think they mean.

What "Interested in Puzzle Games" Actually Means

When Meta labels a user as "interested in puzzle games," that could mean any of the following:

• They watched a puzzle game video ad for 3 seconds
• They liked a Facebook page about mobile gaming
• They clicked on a Candy Crush article
• They visited a gaming review website
• An algorithm inferred they might like puzzles based on other behavior

None of these signals tell you whether someone actually plays puzzle games. None tell you whether they spend money inside apps. And none tell you whether they are on a device that indicates premium purchasing power.

Interest categories were designed for brand advertising — helping Nike find people who like running, or Coca-Cola find people who drink soda. They were never built to find high-LTV mobile gamers.

The Gap Between "Interested" and "Spends Money"

This is where most gaming UA budgets leak. You are paying to reach people who Meta thinks might be interested in your game category, but the signal quality is low.

Consider the difference:

Interest targeting: "This user might like puzzle games based on their browsing behavior."

Behavioral data: "This user has puzzle gaming apps installed, has been active in the last 30 days, and is in the top 20th percentile of in-app spenders on a premium device."

These are fundamentally different signals. The first is a guess. The second is observed behavior.

What 122 Million Puzzle Game Devices Looks Like

Onmo Signal tracks behavioral data across 300M+ device IDs and 170,000+ apps. For puzzle games alone in the US, that is 122 million devices with actual behavioral scoring.

These are not inferred interests. They are devices where we can see:

• App category activity — which gaming categories the user is active in
• Spend behavior — whether they make in-app purchases and at what level
• Device quality — premium hardware as a proxy for purchasing power
• Engagement depth — casual usage vs power engagement across multiple apps

This data comes from 4 billion+ behavioral events, scored and segmented into 273 pre-built audiences you can browse and upload immediately.

Why This Matters for Your UA Campaigns

When you upload a behavioral audience as a Custom Audience to Meta, TikTok, Snapchat, or Google, two things happen:

1. Direct targeting improves. You are starting with users who have demonstrated real gaming behavior, not inferred interest. Match rates vary by platform, but the seed quality is categorically different from interest categories.

2. Lookalike quality improves. A lookalike built from "users who actually spend money in puzzle games" will find better prospects than a lookalike built from "users Meta thinks might like puzzles." The seed defines the output.

This is not a marginal improvement. It is a structural change in how you build your targeting foundation.

Beyond Puzzle Games

The same behavioral depth exists across gaming categories:

Each category has segments scored by spend behavior, device quality, and engagement intensity. And for teams targeting the Gulf region, there are 48M Saudi devices and 58M+ across the GCC — a depth that most US-centric platforms cannot match.

How to Try It

Onmo Signal is self-serve and currently free — no upload limits, no minimum spend.

1. Sign up at signal.onmoai.com
2. Connect your ad account (Meta, TikTok, Snapchat, or Google)
3. Browse 273 pre-built segments filtered by category, geography, and quality tier
4. Upload to your ad account in one click
5. Use for direct targeting or as lookalike seeds

The free access window will not last — a paywall is coming within the next couple months. If you are running gaming UA on any budget, it is worth testing behavioral audiences against your current interest targeting to see the difference firsthand.

---

Onmo Signal is a behavioral audience data platform built for mobile gaming publishers. 300M+ device IDs scored by real app behavior. Try it free at signal.onmoai.com.