Event Match Quality Explained: How to Get Above 7 (and Why It Pays)
Event Match Quality is one of those numbers that sits in Events Manager for years without anyone looking at it, quietly costing money.
It’s Meta’s 0-to-10 score for how confidently it can connect the conversions you report to actual people on its platform. That’s it. It’s not a quality score for your ads, or your business, or your creative. It’s a measure of how much useful identifying information you’re handing over with each event.
Low score, weak connection. And a weak connection costs you in ways that never announce themselves.
What a low score actually does to you
Nothing breaks. That’s what makes it so easy to ignore.
What happens instead is that your retargeting pools quietly shrink, because Meta can’t confidently tell that the person who filled your form is the person scrolling Instagram. Your lookalikes get built from a fuzzy seed, so they’re a fuzzy audience. And the optimization itself gets worse, because the algorithm is learning from conversions it can only half-attribute to real profiles.
Your dashboard looks unremarkable the whole time. Costs just drift.
What it costs you
Smaller retargeting audiences, weaker lookalikes, and a higher cost per acquisition, none of which show up as an error anywhere.
The identifiers that actually move it
Meta scores you on two things: which customer information you send, and what percentage of your events include it. Both matter. Sending email on 40% of events is not the same as sending it on all of them.
The ones that carry the most weight are the ones that identify a person rather than a device: email, phone number, first and last name, city, state, ZIP, country. Then the platform-specific ones, the fbp and fbc cookies, which tie the event back to the click. IP address and user agent help, but on their own they’re thin. If IP is most of what you’re sending, that’s usually your whole problem.
Here’s the part that makes this worth your afternoon, and it’s specific to lead generation: you already have the data. The person just typed their email and phone number into your form. You are not missing it. You’re just not sending it.
Format first, or none of it matters
This is where most low scores actually come from, and it’s unglamorous.
Normalize before you hash. Email gets trimmed and lowercased. Phone gets stripped to digits with the country code. Names get lowercased. Then you SHA-256 it. Meta normalizes and hashes its own copy the same way and compares the two.
If you hash Keith@Example.com with a capital K and a trailing space, it will not match keith@example.com. The hash isn’t “close.” It’s a completely different string. Nothing warns you. Your score just sits there at 4 and you assume Meta is being difficult.
What I’d do: Send email, phone, and name server-side through CAPI on every conversion event, along with fbp and fbc. Normalize before hashing. Reuse what you captured on that person’s later events instead of sending a thin event because the form was three steps ago. Then wait 48 hours before judging it, because that’s the window Meta scores on.
What good looks like
Above 7 is where you want to be to scale. Nine and up means you’re sending rich, correctly formatted information on essentially every event, and it’s very achievable in lead gen precisely because your forms collect real contact details.
Most accounts I touch jump several points from formatting and coverage alone, before anything clever happens. It’s not a hard win. It’s just one nobody gets around to.
Your free score checks this along with five other signals, in about three minutes.
Common questions
What is a good Event Match Quality score?+
Meta scores it 0 to 10. Under 6 means it is struggling to match your events to real people. Six to eight is decent. Nine or above means you are sending rich, correctly formatted customer information on nearly every event. For accounts scaling real budget, aim for 7 and up.
How do I check my Event Match Quality?+
Open Meta Events Manager, select your dataset, click the event, and look at the Event Match Quality score. It is calculated from the last 48 hours, so give any fix a couple of days before you judge it.
Does hashing my customer data hurt matching?+
No, as long as you normalize before you hash. Hashing is how the data stays private in transit: you hash it, Meta hashes its own copy, and the two hashes are compared. Send the raw value and you have a privacy problem. Hash a badly formatted value and it simply will not match.
Why is my EMQ low even though I send email?+
Usually formatting or coverage. Email that is not lowercased and trimmed before hashing will not match, and sending it on only some events drags the score down. Percent of events sending matters as much as which identifiers you send.
Find out which of these are true for your account.
Get your Tracking 100 score in about three minutes. No account access.