Skip to content

Raise your EMQ score

Event Match Quality is Meta’s rating of how well the customer information you send with an event can be matched to a Meta account. Meta scores it from 0 to 10 and recalculates it continuously from the last 48 hours of data.

It is not a measure of whether your integration works. It measures how much identifying information each event carries, which means almost everything that moves it happens in your form or on your site rather than in Studio. That is the useful part.

In Events Manager, open your data source, find the event, click View details and open the Event matching tab.

Meta calculates Event Match Quality for website events sent through the Conversions API. It is not available for offline events, app events or conversion leads.

Offline conversions have their own measure instead, the offline data quality score, which is also out of 10 but is built differently. It combines freshness, frequency, attribution and accuracy over the last 28 days rather than the last 48 hours. Meta publishes bands for that one: 8 to 10 is high, 4 to 7 is medium, 0 to 3 is low, and 8.5 or above is the point at which Meta describes omnichannel use as confident. Uploading daily, and at least 12 times in any 14 day period, is what keeps freshness and frequency up.

Meta publishes a priority ranking rather than numeric weights. Email address and the click ID are the only two it ranks as high.

PriorityParameters
HighEmail address, and the Meta click ID.
MediumPhone number, country, date of birth, external ID, browser ID, and Facebook login ID.
LowFirst name, last name, city, postcode, and lead.

Note what is absent. Value, currency and the product fields describe the conversion rather than the person and do not affect the score. The IP address and user agent are not in the ranking either, though Meta requires the user agent on web events and says sending both may help matching.

These are the changes that have made a difference across the connections we review, ordered by how often they turn out to be the problem.

What to changeWhy it moves the score
Collect an email address if you are not alreadyEmail is one of only two parameters Meta ranks as high priority, so this is the largest single lever available.
Capture the Meta click IDThe other high-priority parameter. It costs nothing to store against the event once the tag is capturing it.
Make first name and last name required on the formIndividually low priority, and the most common thing we find missing. An optional name field means the parameter is offered and rarely arrives.
Implement a dataLayer where the tag cannot read the fieldsWhere personal data is not exposed in the page in a way the script can read, nothing reaches Meta however well the tag is configured.
Finish the subdomain setupUntil the DNS records are in place, events are not collected first-party and identifiers are lost to browser restrictions.
Send value on purchase eventsNot a match parameter, and one of the most frequent omissions we see. Worth fixing in the same pass.

Check the formatting before blaming the coverage

Section titled “Check the formatting before blaming the coverage”

A parameter that arrives in the wrong shape counts as one that did not arrive. An email address with a capital letter, a phone number without a country code, or a value hashed before it was normalized all produce a hash that matches nothing. See Normalize email and phone before hashing.

  • A form that only collects a phone number. Phone is medium priority and email is high, so where there is no email address to collect the ceiling is lower.
  • A checkout that captures less than the rest of the site does, which is common and shows up as a low score on purchase events specifically.
  • Very low event volume. The score is calculated from the last 48 hours, so an event with no instances in that window has nothing to score.

Meta does not publish a target or a minimum. Events Manager shows a qualitative label alongside the number, and Meta does not publish the ranges behind those labels either, so treat any specific band you see quoted elsewhere as someone else’s interpretation.

The practical approach is to compare against yourself. Note the score, make one change, and look again after 48 hours.

Reviewing the tag configuration needs access to your tag manager container. If you would like Datahash to look at it, granting that access is the step that unblocks it.