How to Improve Facebook Event Match Quality (EMQ) Score

What Is Event Match Quality and Why It Matters

Meta compares the hashed identifiers in your event payload against its account database. If the payload carries only a vague combination (say, country plus gender), Meta marks the event as unmatched. Unmatched events do not count toward attribution, so your ad account reports fewer conversions than actually happened.

The score appears per event name inside Events Manager under the dataset quality tab. It currently applies to web events only.

A low EMQ directly inflates your cost per result because the delivery algorithm receives weaker optimization signals. Media buyers who raised EMQ from 4 to 8+ routinely report 10-20% CPA improvement on the same budget.

For a broader look at why server events complement the browser pixel, see server-side tracking vs browser pixel.

Which Customer Parameters Affect EMQ

Meta scores events based on the quantity and quality of identifiers in the user_data object. The official hierarchy from the customer information parameters docs:

Events that rely only on geo and gender are considered invalid and score near zero. Aim to send at least email or phone plus two additional identifiers on every event.

How to Check Your EMQ Score

In Events Manager

Navigate to Events Manager, pick the relevant data source, and open the overview or quality tab. Each event name shows a composite EMQ score out of 10 plus a breakdown of which identifiers contributed.

Via the Dataset Quality API

The official Dataset Quality API returns per-event scores programmatically:

curl -G \
  graph.facebook.com/v25.0/{dataset_id} \
  -d "fields=web{event_name,event_match_quality}" \
  -d "access_token={access_token}"

The response includes a composite_score and match_key_feedback showing coverage percentages for each identifier type. Use this in a scheduled script to alert when scores drop below your threshold.

If you are just getting started with sending server events, follow the Conversions API setup guide first.

Step-by-Step Fixes for Low EMQ

1. Fix normalization before hashing

The most common cause of low EMQ is hashing dirty data. A single trailing space or uppercase letter produces a completely different hash.

import hashlib

def hash_email(raw: str) -> str:
    return hashlib.sha256(raw.strip().lower().encode("utf-8")).hexdigest()

def hash_phone(raw: str) -> str:
    digits = "".join(c for c in raw if c.isdigit())
    digits = digits.lstrip("0")
    return hashlib.sha256(digits.encode("utf-8")).hexdigest()

2. Add missing identifiers

Audit your payload. If you send only email, add phone, name, and location fields. Every additional matched key raises the composite score.

3. Pass click and browser cookies

The fbc and fbp cookies tie the server event to a specific browser session. Without them, Meta loses the link between the ad click and the conversion. Read the cookie values server-side from the request headers or store them in a first-party cookie on the landing page.

4. Always include IP and user-agent

These two fields are never hashed and require no normalization. Forward them from the incoming HTTP request. They alone can lift match rates by several points.

5. Deduplicate properly

If you fire both pixel and CAPI for the same event, include a shared event_id so Meta deduplicates rather than double-counting. See Facebook pixel and CAPI event deduplication for the exact pattern.

Common Mistakes That Tank Your Score

If you manage multiple ad accounts or pixel destinations manually, Pixel Activator lets you activate and audit pixel events without writing code. For teams that need automated routing across trackers and platforms, MOST handles normalization, hashing, and deduplication in one pipeline.

Sources

value optimization with purchase value.