How to improve your Meta match rate: what to check and fix
Meta match rate is the percentage of conversion events you send that Meta successfully connects to a real user profile. When that number drops, Meta's algorithm loses signal. It can't find buyers who look like your customers. ROAS gets volatile, targeting drifts, and most teams only notice weeks later when the numbers stop making sense.
Here's what to check and what to fix.
What a good match rate looks like
Anything above 80% is solid. Below 60%, Meta is discarding a meaningful share of your events. Below 40%, the algorithm is working with a genuinely incomplete picture of who converted.
The number lives in Events Manager under event match quality. Meta also gives you an event match quality score, or EMQ, from 0 to 10. Below 7 means you're leaving match potential on the table. Below 5 means there's a structural problem in the implementation.
Why match rate drops
Match rate falls when the identifiers attached to your events are missing, incomplete, or hashed incorrectly.
Meta matches each event to a user profile using a stack of signals: email, phone number, name, date of birth, city, country, browser IP, and user agent. The more of these you send, the higher your match probability. Most implementations send two or three. The strongest ones send six or more.
Email is the most important single identifier. If your checkout collects email but your Conversions API isn't forwarding it, or is sending it unhashed, match rate drops immediately. Meta requires SHA-256 hashing on all personal data. That's not optional and it's a surprisingly common gap.
The second most common cause is checkout events firing without user data attached. Purchase events that fire at the confirmation page lose session context, especially on guest checkouts. If the email and phone aren't being pulled from your order data server-side and attached to the event payload, those events go out nearly empty.
There's also a deduplication issue that hits teams shortly after going live with a Conversions API. When you run a pixel and a server-side event together, Meta expects them to share an event_id so it can merge them rather than count them twice. If the IDs don't match, Meta drops the events instead of deduplicating them. Match rate falls and nobody knows why because the setup looks correct on paper.
The fixes that actually move the number
Start with email coverage. Pull it from your order confirmation data on the server side, not from the browser session. Server-side order data doesn't disappear when a user has an ad blocker or a slow connection. Hash it with SHA-256 before it leaves your system.
Then extend the identifier stack. First name, last name, city, and country each add match probability. Adding all four typically moves an implementation from the 60s into the 80s. Phone number, if you collect it, is the second most powerful identifier after email. Most brands have it and aren't sending it.
If deduplication is the problem, fix the event_id. Generate a single ID at the moment of purchase and pass the same value to both the pixel and the Conversions API. If the two sides are generating IDs independently, they will never match.
The highest-impact change for most brands is moving Purchase event capture to the server entirely. When a purchase fires from the browser, it's subject to iOS restrictions, ad blockers, and whatever the user's browser decides to limit. When it fires from your backend or payment system, none of that applies. The event goes out with full order data attached, every time.
Match rate isn't a one-time fix
It's worth checking monthly. A drop usually means a checkout flow update, a new browser restriction, or a code change broke something quietly. Teams that monitor it catch problems early. Teams that don't find out through unexplained ROAS drops.
You can read more about what match rate is and how it's calculated in the match rate glossary.
Signal quality problems are fixable. Match rate is usually the fastest place to see the impact.
If you want to see where your signal stands today, run a free signal audit with Journify.
