Meta's Special Ad Categories framework removes most of the granular targeting iGaming operators built their 2018–2022 playbooks on. Detailed interest targ
Meta's Special Ad Categories framework removes most of the granular targeting iGaming operators built their 2018–2022 playbooks on. Detailed interest targeting on "casino games" or "sports betting." Demographic refinement below the legal age and gender allowed by the operator's license. Behavioral targeting on prior gambling activity. All restricted in 2026 across most jurisdictions where Meta runs gambling ads. The operators still winning on Meta have not gone back to those tactics — they have replaced them.
This is what Special Ad Categories actually means for casino and sportsbook campaigns in 2026, what still works, and how to design audiences that perform inside the framework.
What Special Ad Categories is
Special Ad Categories is Meta's policy framework that restricts targeting in advertising categories Meta has identified as regulated or sensitive — housing, employment, credit, social issues, and gambling in most jurisdictions where Meta permits gambling ads.
The framework restricts:
- Detailed interest targeting (interests, behaviors, certain demographics)
- Custom audience refinements that could exclude protected groups
- Location targeting below 15-mile minimum radius
- Demographic targeting below the legal minimum age and gender allowed by the operator's license
The framework still permits:
- Geographic targeting at country/region level (above the 15-mile minimum)
- Age targeting at the legal minimum and above (21+ in US, 18+ in most EU, 19+ in parts of Canada)
- Lookalike audiences seeded from FTD or value cohorts
- Custom audiences from CRM (email/phone hashed lists)
- Advantage+ audiences (Meta's algorithmic audience-discovery layer)
- Contextual placement targeting (sports content, gaming content, news)
Why it applies to gambling in 2026
Meta's stated rationale is that gambling advertising in many jurisdictions is subject to specific audience-protection rules — minimum age, restrictions on targeting populations at risk for gambling harm, restrictions on segmenting that could appear discriminatory. Special Ad Categories standardizes Meta's targeting controls to match those rules.
The practical result for operators is that the granular audience strategies that worked in 2020 — targeting "interest: poker" combined with "interest: cryptocurrency" combined with "behavior: small business owner" — are unavailable. Operators have to build audience strategy out of the tools that remain.
What still works inside Special Ad Categories
Lookalike audiences from FTD cohorts
The highest-ROI audience for most operators in 2026. Pattern:
- Upload top-LTV cohort (top decile by NGR-LTV12) as a Custom Audience via Customer File or CAPI customer list endpoint
- Build a 1% Lookalike for tight match, 2–3% Lookalike for volume
- Refresh seed every 30 days as cohort behavior data accumulates
LAL audiences seeded from value cohorts (not blended FTD lists) typically deliver CPA 30–45% below blended Advantage+ audiences in 2026 Meta audits.
Custom audiences from CRM
Existing-player retargeting, reactivation, cross-sell, and VIP nurture all run inside Special Ad Categories rules with hashed-list custom audiences. Operators with mature CDP integration get this to scale; operators relying on manual list uploads see degraded match rates.
Advantage+ shopping campaigns
Meta's algorithmic audience-discovery campaign type. Performance varies by market and operator data maturity. Pattern:
- Operators with 12+ months of pixel/CAPI data in a market see Advantage+ deliver competitive CPA to manual audience builds
- Operators new to a market often see Advantage+ underperform until conversion volume passes ~500/month threshold
- Advantage+ benefits most from healthy CAPI signal — pixel-only operators see weaker Advantage+ performance
Contextual placements
Sports content categories (relevant for sportsbook), gaming content categories (relevant for casino), news placements. Less powerful than detailed interest targeting in 2020 terms, but still meaningfully shapes who sees the ad and supports Meta's algorithm in finding lookalike-similar users.
Geographic targeting
Country, region, city above the 15-mile minimum radius. For US state-licensed operators, the practical structure is state-level targeting with cross-state geofencing handled at the operator's product layer (Meta will serve to state borders, the operator's product handles the actual eligibility check).
What does not work and what to replace it with
| Old (pre-Special Ad Categories) | New (Special Ad Categories) |
|---|
| Interest: Casino games, Sports betting, Poker | Contextual placements + Lookalike from value cohort |
| Behavior: "Engaged with gambling content" | Custom Audience from CRM + Lookalike from CRM |
| Demographic targeting below legal age (e.g., 25–45) | Legal minimum age + (no upper limit beyond Meta minimum 65+ excludable) |
| Income or homeowner status targeting | Lookalike from value cohort (proxies income via FTD behavior) |
| Cross-app behavior segmentation | First-party CDP signals routed through CAPI |
| Sub-15-mile radius targeting | 15-mile minimum + product-side geo-fencing |
Audience design pattern for a new market launch under SAC
For an operator entering a new regulated market on Meta with no historical data:
**Phase 1 (weeks 1–4): Seed the algorithm.** Run Advantage+ shopping campaigns with broad geo (country level), legal-minimum age, and contextual placement bias toward sports or gaming content. Goal: generate enough conversion volume (target 500+ FTDs in pixel/CAPI data) for the algorithm to learn.
**Phase 2 (weeks 5–8): Build first lookalikes.** Once 500+ FTDs are in the data, build 1% LAL on top-decile by deposit value, 2% LAL on broader FTD cohort. Test each LAL against ongoing Advantage+ baseline.
**Phase 3 (weeks 9–12): Layer retargeting.** Once CRM has 1,000+ active depositors, upload segmented Custom Audiences for reactivation (45-day-dormant), cross-sell (sportsbook-only players for casino offers), and VIP nurture.
**Phase 4 (ongoing): Maintain.** Refresh LAL seeds every 30 days, refresh CRM Custom Audiences every 14 days, run creative refresh against fatigue.
Where Special Ad Categories breaks operator habits
Three common operator instincts that fail under SAC:
**Trying to exclude older players.** Operators sometimes want to exclude 55+ as "not the casino target." Special Ad Categories prevents demographic exclusion above legal age. Adapt by letting creative and contextual placements self-select instead.
**Targeting hyper-local geo.** Operators sometimes want to target single neighborhoods around their retail venues. The 15-mile minimum radius prevents this. Workaround: build retail-tie-in campaigns at city level with creative that calls out the retail venue.
**Cross-referencing with sensitive segments.** Operators sometimes design audiences from "low-income" or "credit-constrained" inferences. Even where Meta would otherwise permit the data, Special Ad Categories prohibits the exclusion or refinement. Adapt by focusing acquisition on responsible-target audiences and using CRM for individual-level safer-gambling controls post-acquisition.
How Basher designs audiences for gambling on Meta in 2026
We design audience strategy around lookalikes seeded from operator value cohorts, CRM-derived custom audiences for retargeting and reactivation, Advantage+ for top-funnel volume, and contextual placement bias to support algorithmic discovery. Country fragmentation, gambling permission compliance, and CAPI implementation are pre-requisites; audience design is the layer on top.
For operators rebuilding their Meta audience playbook under current Special Ad Categories rules, [contact Basher](/contact/).
Related reading
- [iGaming Meta Ads Compliance 2026](/resources/guides/igaming-meta-ads-compliance-2026/)
- [Why Meta Rejects Casino Ads — 7 reasons](/article/why-meta-rejects-casino-ads-7-reasons-2026/)
- [Meta Business Manager Setup for iGaming Operators](/article/meta-business-manager-setup-igaming-operators-2026/)