Open Events Manager, send a test event from your server and confirm it appears in the test events tool, then check the server event details tabs for coverage, deduplication, match quality and diagnostics. If the test event shows up and the tabs show no warnings, the setup is working.
TL;DR
- Start with the test events tool: it shows whether Meta received a well-formed server event.
- Check that the same events arrive from both the browser and the server.
- Meta says to aim for a 75% event coverage ratio between server and browser events.
- Check that one order shows as one purchase, not two.
- Open the Diagnostics tab and read any warning before changing anything.
- Recheck after a theme, checkout or platform update.
What you need
- Access to the dataset (pixel) in Events Manager.
- The test event code from the Test events tab.
- A way to send a test event: your platform's connection, your developer, or Meta's Graph API Explorer.
- Five minutes with a real or test order.
The seven checks at a glance
| Check | Where in Events Manager | What it tells you |
|---|---|---|
| Test event arrives | Test events tab | The payload is well formed and Meta received it |
| Both sources send events | Event overview tab | The same events come from the browser and the server |
| Server share of events | Event coverage tab | How many browser events also reach Meta from the server |
| Duplicates removed | Event deduplication tab | Which deduplication keys you share |
| Customer matching | Event matching tab | How well events match to people |
| Delay | Data freshness tab | Time between the event and Meta receiving it |
| Warnings | Diagnostics tab | Issues Meta has found, with recommendations |
Meta recommends monitoring these tabs regularly after you set up website events, not only once.
Step 1: Send a test event and see it arrive
Go to Events Manager, open the Data sources tab, select your dataset and click the Test events tab. Find the server events section and use the test code with your setup. Meta says the tool checks that your server event payload is formed correctly and that Meta received it.
How to check: the test event appears in the tool within moments. Test information stays there for 24 hours or until you clear it. If nothing appears, check the payload with Meta's payload helper tool before anything else.
Let Madgicx's AI do the work
Its AI marketer finds what's working and prepares the next move. You approve.
Try for Free ($0 Trial)Step 2: Confirm both sources are sending
Open your dataset and the server event details, then the Event overview tab. Meta's best practice is a redundant setup: the Conversions API alongside the Meta Pixel, sharing the same events with both. Your key events should show a browser source and a server source.
How to check: Purchase, Add to cart and Lead each list both sources.
Step 3: Read the Event coverage tab
The Event coverage tab compares events received from the Conversions API with unique browser events from the Pixel. Meta says to aim for a 75% event coverage ratio for accurate reporting. A low ratio usually means events are missing on the server side, not in the browser.
How to check: the coverage for your main conversion event sits near or above 75%.
Step 4: Check that duplicates are removed
Meta's deduplication documentation says the Pixel's eventID must match the Conversions API event_id, and the event names must match. Open the Event deduplication tab to see which keys you send. Then place a test order.
How to check: one order shows as one purchase in Events Manager.
Step 5: Look at match quality
The Event matching tab shows your event match quality and tips for raising it. Meta says better match quality can lead to more conversions and lower cost per result. Meta scores it from 0 to 10 for website events, so a rising score after you add customer details is a good sign.
How to check: the score for your purchase event does not fall after a site change, and the tab's tips list shrinks.
Step 6: Check data freshness
The Data freshness tab shows the delay between an event happening and Meta receiving it. Meta says to share events in real time, or as close to it as possible. One firm limit sits underneath: an event can be up to 7 days old when you send it, and if any event in a request is older, the whole request fails.
How to check: delays are small, and no batch of old events is waiting to be sent.
Step 7: Read Diagnostics
Open the Diagnostics tab. If you see an ! icon, hover over it: Meta explains the issue and gives recommendations. Fix the warnings in order, then send another test event.
How to check: no open warnings for your key events.
What does each symptom mean?
| What you see | Likely cause | What to do |
|---|---|---|
| Test event never appears | Payload not formed correctly, or the wrong dataset or test code | Check the payload with the payload helper and the dataset ID |
| Only a browser source is listed | The server is not sending, or events fail before sending | Check the connection and recheck Step 1 |
| Purchases look doubled | Event ID or event name differs between browser and server | Send the same event_id and event_name from both |
| Coverage is well below 75% | Server events missing for some pages or actions | Compare which events the browser sends with those the server sends |
| Match quality is low | Few customer details sent with events | Send more of the customer information parameters Meta lists |
| Events arrive late | Events are batched or queued | Send them closer to real time, and within 7 days at most |
Before you rely on this table, remember that a symptom can have more than one cause. Change one thing at a time and retest.
What is the most common mistake?
Checking only that events arrive. Events can arrive and still be counted twice, or arrive without enough customer details to match. Always check deduplication and match quality as well as arrival.
What should I do next?
Put a monthly reminder in your calendar to open these tabs. Recheck straight after you change your theme, checkout, tag manager or platform settings. If the setup itself needs work, our step-by-step Conversions API setup answer walks through the methods.
Where Madgicx fits
Madgicx can do the server side for you. Madgicx Cloud Tracking hosts Meta's Conversions API Gateway and Signals Gateway and sends conversions to Meta from the server, so Meta sees and attributes more of your conversions. You still run the checks above in Events Manager to confirm what Meta receives.
FAQ
How do I know if my Conversions API events are being received?
Events Manager shows them. Open the Test events tab for a test event, and the Event overview tab for live events with a server source.
How long does it take for Conversions API events to show up?
Test events normally appear shortly after you send them. Live events should arrive in real time or as close to it as possible, which the Data freshness tab confirms.
What is a good event coverage ratio?
Meta advises aiming for 75% between Conversions API and Meta Pixel events. It is Meta's own target for accurate reporting, not a guarantee of results.
Why do I see double purchases after adding the Conversions API?
Your browser and server events are not being matched. The event_id and event name must be the same from both sources, or Meta counts them as two events.
Do I need a developer to test the Conversions API?
Not always. The test events tool is in Events Manager. A developer helps if the payload needs fixing, since a direct build sends it from your own code.
Related
- Answers: How do you set up the Conversions API?
- Answers: What is the Conversions API?
- Answers: Why Ads Manager results differ from revenue
- Glossary: Event match quality explained
Go deeper
For a longer guide to the options and what each sends, read our guide to the Facebook Conversions API.
Sources
- Monitor and improve your Conversions API setup in Meta Events Manager, Meta Business Help Center
- Test your server events using the test events tool, Meta Business Help Center
- Handling duplicate Pixel and Conversions API events, Meta for Developers
- Using the Conversions API, Meta for Developers
