Meta scores event match quality out of ten in Events Manager, and it has never published what a good score is. That absence is the most useful thing about this metric, because every specific target circulating online was made up by whoever needed a number for their article.
The score itself is real and it matters. The benchmark you have been given for it does not exist.
This is for you if you have been told to get this figure above some particular threshold and want to know where that threshold came from.
What event match quality is measuring
It scores how well Meta can connect the events you send to actual accounts.
When somebody converts, your pixel or server sends the event along with customer parameters: a hashed email address, a phone number, a name, a city. Meta uses those to work out which of its users this was.
More clean parameters means a better match. A better match means the conversion gets attributed properly and the algorithm learns from an identified person rather than an anonymous event.
An unmatched event is not lost. It is counted, and it teaches the system considerably less about who to look for next, which is where the cost shows up.
The benchmark that does not exist
Meta’s own documentation for the Dataset Quality API defines event match quality as a score out of ten. It gives no threshold, no target, and no description of what counts as good.
That is the whole published position. A scale, and silence about where on it you should be.
Search for a target and you will find plenty. Aim for 8. Above 6 is healthy. Below 7 needs work. None of those trace to Meta, and none of the pages carrying them cite anything, because there is nothing to cite.
This is worth stating plainly rather than politely. Absence of evidence, reported honestly, is more useful than a fabricated benchmark, and this metric is a clean example of a number that got invented because a gap felt uncomfortable.
What event match quality actually looks like in live accounts
Since no benchmark exists, here is a range from accounts I run rather than a target I made up.
Mine sit between roughly 4 and 8. That is a wide spread and it reflects real differences: what each site collects, whether Advanced Matching is on, how much traffic the account gets, and how long it has been running.
Two patterns hold across them. It improves over time, because matching gets better as data accumulates. And it improves faster with Advanced Matching enabled, which is the single change with the clearest effect.
Treat those as one practitioner’s observations from a small number of accounts. They are not a benchmark either, and I would rather give you a described range with its limits than a confident number with none.
How to improve event match quality
Send more, and send it clean.
Email address does the most work, followed by phone number. Name, city, postcode and country each add smaller amounts. The improvement comes from breadth of parameters more than from perfecting any single one.
Turn on Advanced Matching so the pixel picks up what is already present on the page, rather than requiring somebody to pass every field manually. This is usually the fastest available gain and it is a toggle rather than a project.
Then check the obvious failure: whether the parameters are actually populated on the pages that matter. Plenty of sites collect an email on the form and never pass it to the event, so the score reflects a plumbing gap rather than a data gap.
One terminology note, because it causes confusion in meetings. Advanced Matching is Meta’s feature. Enhanced Conversions is Google’s. They do comparable jobs and the names are not interchangeable.
When a low score is fine and when it is a warning
A low event match quality score is usually a configuration choice, not a fault.
If you send an email and nothing else, you will score low and everything is working exactly as built. That is a decision about what you collect, and there are legitimate reasons to collect little.
What matters is a drop. A score that sat at 7 for months and is now 4 means something stopped being collected, almost always after a site change. Nothing else in the interface will tell you that a form field stopped passing through.
So the useful habit is not chasing a target. It is knowing your normal figure and noticing when it moves.
Why chasing ten is the wrong goal
The returns flatten well before the top of the scale, and the cost of climbing rises.
Getting from 3 to 6 usually means turning on a feature and fixing a field. Getting from 8 to 10 means collecting and transmitting more personal data than many businesses should, for a smaller improvement in matching.
There is also a proportionality question worth taking seriously. Every additional parameter is more customer data leaving your site. Doing that to move a score whose meaning Meta declines to define is a poor trade.
Improve a bad score. Leave a decent one alone. Then go and fix the conversion event itself, which moves results considerably more than this ever will.
Why the missing benchmark is the story
It is worth asking why Meta has never published a target, because the likely answer is instructive.
A threshold would have to hold across wildly different businesses. Compare a high-volume ecommerce site collecting email, phone and address at checkout with a service business capturing one email on a short form. One pass mark would be wrong for most of them on day one.
There is also a reason not to invite gaming. Publish a number and every advertiser optimises to it, including by sending data they should not, and the score stops measuring what it was built to measure.
So the silence is probably deliberate and probably sensible. What is not sensible is the industry filling that silence with invented figures and presenting them as guidance.
The general habit is worth carrying beyond this metric. When a platform declines to publish a threshold and the internet supplies one anyway, the internet made it up. That pattern repeats across creative benchmarks, frequency caps and lead quality rates, and event match quality is the cleanest example of it.
What to change this week
Three steps.
Open Events Manager and write down your current event match quality on your money event. That number is your baseline and it only becomes useful once it exists.
Then check whether Advanced Matching is enabled. If it is not, that is the single highest-return change available here and it takes minutes.
Then stop treating any published threshold as real. If somebody quotes you a target, ask where it came from, and watch what happens.
The Meta Conversions API covers the delivery side of the same problem, what conversion numbers actually measure covers the counting underneath, and optimising on the wrong event is the mistake that no amount of matching quality will rescue. Most of the fixes here happen on the site rather than in the ad account, which is web development work. If you want yours read properly, book a teardown.