Overview
When we say “Meta Health Assessment”, we don’t just mean “do you have a restriction or not?”. We scan four pillars with 40+ individual signals.
Here’s how we frame it internally so you know what we actually look for.
Pillar 1 – Asset trust & reputation (HiVA Tiers)
We infer how trusted each major asset is:
Core Business Managers
Active Ad Accounts
Main advertising Pages
Sometimes key domains/pixels
Example signals:
Low HiVA Tier / degraded trust tier
Security-score concerns
Trust drag caused by shared/legacy assets
How this shows up:
Higher CPM vs similar accounts
More friction, Learning Limited, and fragile scaling
Pillar 2 – Penalties, enforcement & risk
We look for patterns that match quiet enforcement:
Repeated generic rejections
Slow, sticky reviews
Invisible spend ceilings
Campaigns that “hit a wall” at certain budgets
Example signals:
ACE Warning (Asset Compliance Escalation)
DMCA/IP complaints
Engagement Abuse flags
Structural Invalidations and misuse patterns
These don’t always display as big red banners, but they change how Meta treats your spend.
Pillar 3 – Customer experience & reputation (Feedback & CX)
We analyse your reputation from the user’s point of view:
Facebook Feedback Score bands
Negative Feedback density (hides, reports, complaints)
Refund / chargeback behaviour
Compliance reputation history
Example signals:
Negative Feedback / low experience ratings
Feedback Score sitting near/under key thresholds
Reputation-driven disable events
These signals make you pay a CPM tax and reduce delivery priority even when ads are technically “good”.
Pillar 4 – Signals & delivery health (data & structure)
We check whether Meta has a clean, strong signal to optimise from:
Pixel + CAPI implementation
Event Match Quality and mapping
AEM configuration and change history
Budget vs audience sizing (Budget Mismatch)
Delivery health: reach, volatility, auction competitiveness
Example signals:
Payment Restriction or recurring billing issues
Weak or noisy conversion signals
Structural inefficiencies that cause poor learning
Why we show pillars, not just a list of problems
Your performance is almost never ruined by a single flag. It’s usually a stack:
Slightly low trust tier
Some ACE-style friction
Feedback pressure from a bad shipping season
A few signal/structure issues
By grouping signals into these four pillars, your report makes it clear:
What’s hurting you most
What you can safely ignore for now
What to fix first for CPM, stability and scale