The premise: paid social is a workflow, not a channel
Most operators treat Meta and TikTok the way they treat Google: pick a tactic, set a budget, run for a quarter, evaluate. That model breaks immediately on paid social for one reason — accounts die. Not occasionally. Routinely. A licensed operator running €200K/month on Meta should expect to lose 2–4 BMs per quarter to risk-team review, even with verified status.
Once you accept that account loss is a baseline operational fact, paid social stops being "ad creative + budget" and becomes a workflow with three pillars: pre-warmed inventory, pre-staged creative cohorts, and pre-pass compliance. The creative is the easy part.
Pillar 1 — Pre-warmed BM inventory
Verified Business Manager status (and its TikTok equivalent — TikTok Spark approval and partner-merchant tier) is the foundation. Without verification, suspension reactivation can take 14–28 days. With it, 24–72 hours. The difference between those numbers is the difference between scaling and running in place.
Pre-warmed inventory means at any given moment we have:
- 2–3 verified active BMs running spend in each market.
- 2–3 verified BMs in warm-up (low-spend, low-velocity, building Meta trust).
- 1–2 verified BMs idle, ready to absorb spend within four hours of an active suspension.
This is approximately 6–8 BMs per primary market. It is overhead. It is also the reason our scale operators do not have weeks-long acquisition outages when one BM cycles. The math on the cost of overhead vs the cost of an outage is not close.
Pillar 2 — Pre-staged creative cohorts
Creative cohorts are pre-approved sets of 6–12 ads with shared theme, structure, and compliance treatment, ready to deploy on any active BM within 30 minutes. A cohort lives independently of the BM running it.
The benefit is direct: when BM A goes down at 09:14 and BM B picks up spend at 12:00, you do not pause to "rebuild" the campaign on B. You promote a cohort that has already passed your internal compliance pre-pass to B and resume spend. Daily delivery does not collapse.
For a €500K operator we run 4–6 active cohorts per market at any time, with 8–12 in the pre-stage queue. Cohorts get rotated out every 2–3 weeks based on creative fatigue, not compliance.
Pillar 3 — Pre-pass compliance
Most operators we audit have a "submit, see what gets rejected, revise" cycle with paid social. That cycle costs 18–36 hours per round and at scale produces enough rejected ads to slow your manufacturing pipeline by 40%.
Pre-pass compliance moves the review upstream. Before any asset enters the production queue, it gets reviewed against:
- Platform policy — Meta's gambling policy, TikTok's regulated industries policy, by market.
- Local regulator rules — UKGC tone-and-content rules, MGA responsibility messaging, SGA risk-warning placement, AGCO display standards.
- Operator licence terms — most casino licences carry their own constraints on advertising claims, demographic targeting, and bonus framing.
The result: 92–96% of submitted ads pass platform review on first submission, vs the operator average around 60%. The cost is one full-time compliance reviewer per market cluster. The benefit is creative ships at the volume you need.
The four creative cohort archetypes
Across markets, four cohort archetypes consistently outperform on Meta & TikTok for licensed iGaming. We rotate among them based on the market and the operator:
| Archetype | Hook | Best for |
|---|---|---|
| Demonstration | "Watch this user spin / win" UGC-style | Slots-led operators · TikTok-heavy |
| Authority | Talking-head, expert framing, calm tone | DACH · UK · regulated tone markets |
| Comparison | "Vs other casinos" grid / table | LATAM · Tier-2 · price-sensitive |
| Lifestyle | Aspirational, contextual placement | Sportsbook-led · younger demos |
Each archetype runs as a cohort with 8–12 variants. We measure CTR, CPA, and D7 retention separately and rotate weights based on market response, not gut feel.
What kills paid social accounts (avoidable)
The four most common avoidable causes of BM suspension we see:
- Cloaking-adjacent landing pages. Even if not actually cloaking, dynamic content, geofenced redirects, or aggressive AB-test variants trigger Meta's risk team. Solution: stable, single-version landing pages with regional content rendered server-side, not via JS redirect.
- Aggregate spend velocity. Going 0 → €5K/day overnight on a fresh BM. Solution: warm BMs over 14 days, scale velocity 2× per week max.
- Demographic targeting that triggers review. Even when allowed by license, age and interest combos that look "predatory" trigger review. Solution: broad targeting + creative-side audience filtering.
- Account-history reuse. Reusing personal Facebook profiles or thin BMs. Solution: pristine BMs, dedicated business managers, no shortcuts.
What good looks like at scale
For an operator running €300K+/month on paid social with this framework, target metrics are:
- BM survival rate: ≥ 4 months mean lifespan
- Creative pre-pass rate: ≥ 92% first-pass approval
- Cohort time-to-deploy after BM rotation: < 4 hours
- D7 retention from paid social cohorts: within 5% of organic baseline
- Account share of total monthly spend on a single BM: ≤ 35%
The audit question
If you are running paid social for a licensed operator at scale, ask your team one question: "if our top BM was suspended at 09:00 tomorrow, how long until we are spending the same daily volume on a different BM?" If the answer is more than 8 hours, you do not have a paid-social workflow — you have a paid-social campaign that is one suspension away from a bad month.
The fix is not faster reactivation. It is the workflow above. We cover the broader scaling architecture in our FTD scaling piece; this article is the paid-social-specific layer of that.