Unrestrict your
Health & WellnessHealth & Wellness brand.

Meta classifies domains that sell health-adjacent products under Health & Wellness, then restricts what conversion data those domains can share back through the pixel and CAPI. Ads keep running. The data feeding Meta's optimization stops.

Meta's classifier flagged your domain. From that point Purchase, AddToCart, and Lead events are filtered or suppressed depending on the restriction level. Notifications go to the Business Manager email, not the ad account, so most brands discover it weeks after the fact.

Meta has classified your domain under a restricted category. Depending on severity your events are filtered (Level 1), partially blocked (Level 2), or fully suppressed (Level 3). Check Data Sources, then Categories to see your level.

Level 1 (Core Setup) strips metadata but events still fire. Level 2 (Standard Event Restrictions) blocks Purchase, AddToCart, and Lead. Level 3 (Full Domain Restriction) blocks everything from the domain. Each level needs a different fix.

Yes. THC is prohibited from sale entirely. CBD is restricted to 18+ and cannot make health claims. Brands selling CBD with condition-adjacent positioning get classified under Health & Wellness on top of the CBD restriction.

Yes. GLP-1, semaglutide, and any quantified weight-loss product or claim is treated as health-adjacent. Even compliant supplement brands get classified when their positioning suggests body-change outcomes.

Yes. Cosmetic procedures fall under Meta's restricted categories. Before-and-after photos are explicitly prohibited and the classification compounds with provider-relationship signals like booking flows or consultation copy.

Yes. Meta restricts non-surgical genital enhancement, ED treatment, libido supplements, and testosterone boosters under both Health & Wellness and Adult Sexual Products policies.

Yes when claims tie to a condition: blood sugar, sleep, hormones, cognitive function. Generic vitamin SKUs without condition-adjacent positioning often escape classification.

No. Meta's classifier is inferential, not keyword-based. Brands selling ergonomic pillows, fitness apparel, breast pumps for lactating mothers, immigration-law services flagged under "nationality," and supplement brands without overt health claims have all been classified. If buying your product implies the buyer has a health condition, life situation, or protected attribute, you are at risk regardless of whether you use clinical language.

Yes. Meta explicitly prohibits before-and-after body-change comparison in the context of weight loss and any imagery that implies negative self-perception. Transformation photos common in fitness marketing trigger restriction immediately.

The personal-attributes policy filters anything that could make a viewer feel bad about their body, age, weight, skin, or appearance. Before-and-after framing is the canonical example.

No. Cosmetic surgery, injectables, body contouring, skin whitening, and laser treatments cannot use before-and-after framing. Even non-clinical "transformation" results trigger the same filter.

Not as before/after framing. You can use customer testimonials in writing or video that describe feelings ("more energy", "less bloated") without showing body changes.

Avoid direct health claims and transformation imagery. Show product, ingredients, and lifestyle context. Frame copy around the desire ("Ready for a change?") rather than the outcome ("Lose 14 lbs").

Yes. Any body-change comparison, including weight, fitness, posture, or skin appearance, falls under personal-attributes and is rejected.

Lifestyle photography that does not compare two states is allowed. A single image of a person feeling good in their body, doing an activity, is fine. The comparison is what triggers the policy.

Google and TikTok enforce similar but not identical policies. TikTok's weight-loss policy is stricter on transformation framing. Google flags health/medical before-and-after with platform-specific exemptions for licensed clinical advertisers.

The personal-attributes filter catches creative that implies the viewer has a flaw (weight, skin, age, hair) and that the product fixes it. Reframe the hook around what the customer wants, not what they want to fix.

Customer voiceover testimonials, ingredient stories, lifestyle imagery without comparison, ASMR-style product shots, and benefit copy that does not reference outcomes. Most compliant DTC brands moved to UGC-style talking heads.

Pixel is a browser-side script that sends events from the user's browser. CAPI (Conversions API) is a server-side endpoint that sends events from your server. They are different delivery methods for the same Meta event system.

No. Website events sent through CAPI remain subject to Meta's Business Tools terms, privacy controls, and data-sharing restrictions. CAPI improves connectivity; it is not a policy bypass.

Where Meta permits the event and the underlying data, Zappush can map outcomes such as profit value, new-customer purchases, and AOV tiers into a server-side event strategy. We confirm the permitted fields and optimization options during the audit.

CAPI is Meta's server-to-server event endpoint. You send events from your backend with hashed user data and event metadata. Meta deduplicates against pixel events using event_id. It bypasses ad blockers and iOS ITP, not domain restrictions.

Zappush can support multiple authorized datasets or Pixels when the business has a legitimate operational need. We review ownership, consent, destination, and event-routing requirements before recommending the architecture.

Not always. We first audit the existing Pixel or dataset, domain classification, event status, and permitted data. A new Pixel does not remove a policy restriction by itself, so we only recommend one when the implementation and account history justify it.

Server-side tracking can improve reliable delivery of permitted events when it is paired with consent controls, payload cleansing, accurate event mapping, and a compliant destination. Server-side delivery alone does not remove a policy restriction.

EMQ is Meta's diagnostic for how effectively permitted customer information matches events to Meta accounts. It can fall when identifiers are missing, malformed, duplicated, or restricted. Improvement depends on consent, implementation quality, event volume, and the fields Meta permits you to send.

Zappush evaluates each event before delivery, removes prohibited health context from URLs and metadata, maps permitted business outcomes to accurate neutral values, and sends the resulting payload through CAPI. Cleansing does not fabricate an outcome or change what the customer did.

We remove prohibited health context from permitted event payloads, apply consent rules, and hash allowed identifiers in Meta's required format before delivery. Hashing is a security measure, not permission to send sensitive health information.

Meta controls category classification and appeal outcomes, so no provider can guarantee removal. Zappush focuses on preventing prohibited health context from entering permitted events and building a compliant signal path around the classification that applies.

Usually. We audit the existing site and recommend the smallest compliant change. That may include consent updates, landing-page revisions, payload cleansing, or server-side routing. The ad destination must accurately represent the offer and the experience customers receive.

No provider can guarantee Meta's enforcement decisions. Meta reviews ad destinations and may restrict businesses that attempt to evade review. When a separate landing experience is appropriate, Zappush requires it to represent the offer accurately, remain consistent for reviewers and customers, and avoid cloaking or crawler-specific content.

Technical onboarding can often be completed quickly, but full deployment typically takes several days to two weeks. Timing depends on the landing experience, consent setup, domain and Pixel history, approvals, and existing infrastructure.

Not automatically. A new domain does not cure prohibited data or an inaccurate destination, and it may be classified again. Zappush audits the restriction and recommends a domain change only when it supports a legitimate, consistently presented, policy-compliant customer journey.

Any additional step can affect conversion rate. When a separate landing experience is required, we keep it accurate, fast, and visually consistent, then measure click-through and downstream conversion rather than promising a universal impact.

Eligible implementations include a 30-day money-back guarantee subject to the conditions in the service agreement. We document the agreed scope, verify delivery, and investigate implementation failures during that window. Meta's independent enforcement decisions cannot be guaranteed.

Signal diagnostics may improve after clean, permitted events begin flowing, but there is no universal recovery window. Timing varies with event volume, account history, spend, consent rates, implementation quality, and Meta's learning systems.

We can audit an already-restricted setup, but the recommendation depends on the restriction, account history, destination, consent state, and permitted event data. We do not promise a fixed recovery period or a specific enforcement outcome.

Open Events Manager, then Data Sources, then Categories. If your domain shows a restriction notice (yellow or red), you are flagged. Or run our free audit at /audit, which scans the same surfaces Meta's classifier scans and tells you what is likely triggering classification.