Skip to content

Methodology

How Compliance Risk works

Optimus Pass scans ad copy for policy-risk language across 10 platforms using 65risk detectors. Here's how the scoring, weighting, and detection actually work under the hood.

Last updated: September 2026

OPTIMUS / METHODOLOGYAnswer block

How is the Optimus Pass Compliance Risk calculated?

Compliance Risk (higher = more detected risk) comes from 65 risk detectors across claim language, personal wording, platform safety, restricted categories, and landing-page review signals. The active ruleset currently contains 65enabled detectors. Compliance Risk is driven primarily by the strongest detected risk signals. Higher scores mean more detected compliance risk; the score is not a probability that an ad will be rejected. Beyond keywords, the engine also catches semantic patterns — recovery arcs, transformation claims, personal-attribute framing, emotional frustration targeting, and unsubstantiated authority — that no single banned word would ever trigger. Genuinely prohibited offers can reach maximum risk. The score helps you spot risk patterns worth reviewing. It doesn't predict whether a platform will approve or reject your ad.

OPTIMUS / METHODOLOGYAnswer block

How do ad review policies differ across platforms?

Every platform enforces different rules. Something that gets rejected on Meta might fly on TikTok, and vice versa. Because of this, Optimus Pass reports two separate signals: the Compliance Risk number above (platform-agnostic — it flags risky language regardless of destination) and a separate platform-aware enforcement estimate that reflects how strictly each platform is observed to treat different categories of claims. Keep in mind these estimates are directional, based on observed policy patterns and enforcement signals. Actual enforcement depends on your account history, ad context, and region.

Registry

65 Risk Detectors

Each detector is a pattern — a term, phrase, or semantic structure — that correlates with higher ad-review risk on one or more platforms.

Clinical Language

Patterns in this category are directional review signals and may require contextual review.

Medical Claims

Patterns in this category are directional review signals and may require contextual review.

Overpromise

Patterns in this category are directional review signals and may require contextual review.

Disease Claims

Patterns in this category are directional review signals and may require contextual review.

Exaggeration

Patterns in this category are directional review signals and may require contextual review.

Outcome Claims

Patterns in this category are directional review signals and may require contextual review.

Violence / Self-Harm

Patterns in this category are directional review signals and may require contextual review.

Hate / Harassment

Patterns in this category are directional review signals and may require contextual review.

Platform Safety

Patterns in this category are directional review signals and may require contextual review.

Adult / Sexual Content

Patterns in this category are directional review signals and may require contextual review.

Financial Claims

Patterns in this category are directional review signals and may require contextual review.

Patterns

What are Semantic Patterns?

Narrative patterns catch implied claims that no single word triggers.

Recovery Arc:

"I finally feel like myself again" — zero banned terms, but implies recovery from a negative state.

Transformation Claim:

"See results in 7 days" — implies a specific outcome within a timeframe.

Personal-Attribute Framing:

"For people who struggle with focus" — targets a personal attribute, which some platforms restrict.

Emotional Frustration Targeting:

"Tired of low energy?" — exploits a negative emotional state as a hook.

Unsubstantiated Authority:

"Doctors recommend" — implies endorsement without evidence.

Scoring

What does the Compliance Risk score mean?

OPTIMUS / SCORING0–100 · higher = more risk

Directional, not a verdict

Compliance Risk is driven primarily by the strongest detected risk signals. Higher scores mean more detected compliance risk; the score is not a probability that an ad will be rejected. Genuinely prohibited offers can reach maximum risk — rewording alone will not make them acceptable. Platform enforcement decisions are made independently and may change without notice.

Sources

Meta Advertising Standards
Google Ads Healthcare and medicines policy
FTC Advertising and Marketing Guidance
FDA: Dietary Supplements

Ruleset 2026-07-24-enforcement-risk-v1 · 65 enabled detectors · Risk patterns update when confirmed changes are identified.

Want to see how Optimus Pass scores itself?

View the visibility report