Event Match Quality (EMQ): The Matching Mechanics in Meta Ads
How the Meta Ads algorithm calculates the match quality score, and the technical parameters needed to sustain high scores in large accounts.
Operating scenario: Advertisers with an EMQ score below 6.0 in Events Manager suffer from broken attribution and a costlier algorithmic auction.
Technical root cause: Sending incomplete payloads with only basic data, no advanced parameters such as fbp, fbc and IP, and without sanitizing strings before the SHA-256 hash.
Engineering guideline: Building a pipeline in sGTM to capture and normalize all 10 identifiers Meta recommends, sending client and server parameters at the same time.
What Event Match Quality (EMQ) Is and How It Is Calculated
Event Match Quality (EMQ) is a technical score from 0 to 10 assigned by Meta Ads. It measures how effectively the customer data sent with a conversion event can be matched to an active account in the Meta ecosystem (Facebook or Instagram).
When a user completes a transaction on the company's site, the system has to answer a basic question: which social network user made this purchase? If the event carries only the order ID and the monetary value, the match rate is zero. The more reliable parameters you send in a structured way, the better the algorithm can close a deterministic match.
The Weight Table for User Identifiers
Meta's engineering documentation assigns different relevance to each class of identifier sent in the user_data object:
Primary Identifiers (Highest Weight)
- Email (em): The most stable deterministic identifier, historically tied to the user's profile.
- Phone (ph): High match precision across mobile and WhatsApp.
- fbp (Browser Cookie): Created by the domain to identify the visiting device.
- fbc (Click Identifier): Captures the click code of the specific ad.
Secondary Identifiers (Probabilistic Support)
- IP Address and User-Agent: Used together to correlate networks and devices.
- First and Last Name (fn, ln): Strengthen cross-validation when paired with location.
- City, State and ZIP Code (ct, st, zp): Demographic data that reduces collisions on common names.
- Date of Birth and Gender: Optional profile refinement parameters.
Example Code: Sanitizing and Hashing Parameters
Sending identifiers with formatting errors produces SHA-256 hashes that differ from the records stored at Meta. Below is the technical routine in JavaScript (Node.js) to normalize and hash before sending:
import crypto from 'crypto';
function hashData(value) {
if (!value) return null;
const clean = value.toString().trim().toLowerCase();
return crypto.createHash('sha256').update(clean).digest('hex');
}
function normalizePhone(phone) {
if (!phone) return null;
// Remove all non-numeric characters
let digits = phone.replace(/\D/g, '');
// Add country code 55 if missing for Brazilian numbers
if (digits.length === 10 || digits.length === 11) {
digits = '55' + digits;
}
return crypto.createHash('sha256').update(digits).digest('hex');
}
// Object ready for the CAPI payload
const userData = {
em: [hashData("[email protected] ")], // Hashes the sanitized version
ph: [normalizePhone("(11) 98888-7777")], // Normalizes to 5511988887777
client_ip_address: "177.18.240.12",
client_user_agent: "Mozilla/5.0 (iPhone; CPU iPhone OS 17_5 like Mac OS X)...",
fbp: "fb.1.1718000000.123456789",
fbc: "fb.1.1718000000.PAZnRzaAUY-c9wZG9m"
};The Impact of a High EMQ on Acquisition Cost
The EMQ score directly affects the account's operating efficiency in three central ways:
- Accurate Attribution: Conversions that the browser would lose are counted in the dashboard, showing the real return on ads.
- High-Fidelity Lookalike Audiences: The buyer base used to generate lookalikes is better qualified when Meta knows with certainty who bought.
- Bid Stability: The delivery algorithm needs consistent signals to calibrate each user's likelihood to click and buy in subsequent auctions.
Frequently Asked Questions about Event Match Quality
What EMQ score is considered satisfactory for the Purchase event?
For e-commerce and companies with a digital checkout, scores between 8.0 and 9.5 are the market standard of excellence. Scores below 6.0 point to structural problems in parameter collection.
Why doesn't the fbc parameter show up on every event?
The _fbc cookie is only generated when the user reaches the site by clicking directly on a Meta Ads ad that carries the fbclid parameter in the URL. If the visitor arrived organically or directly, fbc is naturally absent.
Can I send data as plain text without SHA-256?
Network identifiers such as IP, User-Agent, fbp and fbc must be sent as plain text. All personally identifiable data (email, phone, name, address) must be sent hashed with SHA-256.
Want to raise your ad account's EMQ score?
Random Marketing audits containers and implements advanced CAPI parameterization for large advertisers.
I want to hire Random Marketing for my company ↗Official documentation & engineering references
Architecture guidelines, API specifications and official technical documentation consulted to support this dossier: