Conversions API gateway common mistakes frequently asked questions

In this guide

The eight common mistakes A quick diagnosis path Frequently asked in practice A worked example Terms you will see How to keep it healthy Frequently asked questions

Check these first

The most frequent mistake is a missing or mismatched event ID, which causes double counting.
The second is thin customer data: events without hashed e-mail, phone, fbp and fbc match poorly.
Use Test Events and the Events Manager diagnostics before you change anything else.

You set up the Conversions API (CAPI), Events Manager shows events arriving, and then the numbers are strange: two purchases for every order, a low match quality score, or fewer events than expected. This page lists the mistakes that cause most of these and gives the fix for each. If you are still choosing a route, read Meta's gateways and Meta Conversions API vs Pixel first.

The eight common mistakes

SymptomPurchases are counted twice

Likely causeThe Pixel and the server send the same purchase with different or missing event IDs.

FixSend the same event_id from both and keep the event name identical. Check "Deduplicated" in Events Manager. See Facebook event deduplication.

SymptomLow event match quality

Likely causeThe server event has no hashed e-mail, phone, fbp or fbc.

FixPass the user data you have, and add stored values for returning visitors. With Open Tracked, the Data Store and Enricher add them for you. See event match quality.

SymptomEvents show only in Test Events, not in reports

Likely causeThe test_event_code is still in the tag.

FixRemove it before you go live.

SymptomEvents are too old

Likely causeEvents are sent after the allowed window, or with a wrong timestamp in milliseconds instead of seconds.

FixUse a Unix timestamp in seconds and send events promptly.

SymptomCurrency or value missing on purchases

Likely causeThe data layer lacks value or currency.

FixSend both on every purchase, with a valid three-letter currency code.

SymptomThe wrong dataset receives the events

Likely causeSeveral pixels, and the server tag uses a different ID.

FixUse one dataset ID in both the browser and the server tag.

SymptomThe source URL is missing or wrong

Likely causeevent_source_url and action_source are empty, or point to your tagging domain.

FixSend the page URL and action_source = website.

SymptomEvents sent without consent

Likely causeThe server tag fires regardless of the visitor's choice.

FixGate the tag on consent and respect the signals your banner passes. See Consent Mode v2.

Set up in minutes

Connected in minutes

Add your domain, install the pixel, connect your ad platforms and watch the first events land.

A quick diagnosis path

Open Events Manager, Test Events

Use the test code and your own browser.

Send a purchase

Check both the browser and the server copy arrive.

Read the event details

Does each carry the same event_id? Which user data fields are present?

Open Diagnostics

Meta lists warnings such as missing parameters and low match quality.

Fix one thing at a time

Then wait for a few hundred events before you judge the effect.

Frequently asked in practice

Event ID: The shared identifier that lets Meta count a browser and a server event once.
fbp and fbc: The browser ID cookie and the click ID cookie Meta uses for matching.
Match quality: Meta's 0 to 10 score of how well an event can be matched to a person.
Dataset: The place events arrive in Events Manager, formerly the pixel.
Action source: Where the event happened: website, phone call, physical store and others.

0 of 6 done

Browser and server events share an event IDEvery purchase has value and currencyThe test event code is removedUser data fields are passed, hashed where requiredConsent controls the tagPurchases in Meta are close to store orders for a week

If you use Open Tracked, the Data Store, Enricher, Cookie Keeper and Visitor Location power-ups on every plan cover most of the matching-data gaps above.

Live view

Watch events arrive, live

A sample feed. In Open Tracked you see the same stream for your own store.

A worked example

A store sees 100 orders in Shopify, 190 purchases in Meta and a match quality of 4.8. Two things are wrong: the browser and server purchases have different event IDs, so Meta counts both, and the server event carries no e-mail or phone. The fix is to send the order number as the event ID from both sides and to add hashed e-mail and phone. A week later Meta shows 104 purchases, marked as deduplicated, and match quality of 7.9. The numbers here are an illustration, not a promise: your own result depends on your traffic and data.

Terms you will see

CAPI: Short for the Conversions API.
Access token: The secret that lets your server send events to a dataset.
Hashed data: E-mail or phone passed through SHA-256 after trimming and lowercasing.
Event source URL: The page where the event happened.
Diagnostics: The Events Manager tab that lists warnings about your events.

How to keep it healthy

Look at the Overview and Diagnostics tabs each week for the first month.
Keep a note of the event ID rule you used, so the next developer does not change it.
Re-check after every site update that touches the checkout or the data layer.
Compare Purchase counts with store orders on the same dates.

Works worldwide

Runs on your domain, anywhere

Your tracking runs on your own domain, so a visit from any country is handled the same way.

Frequently asked questions

Why is my event match quality low?

The events carry too little customer data. Add hashed e-mail and phone, and keep fbp and fbc alive. Stored values for returning visitors help.

Why do I see duplicate purchases?

The browser and server events do not share the same event ID and name.

How long can I wait before sending an event?

Send it as soon as it happens. Meta accepts events only within a limited time window, so late events can be rejected.

Do I need the Conversions API if I have the Pixel?

Most advertisers use both. The server copy recovers events the browser loses.

Is the Conversions API affected by consent?

Yes. Where consent is required, do not send personal data without it.

Do I need spreadsheets to read my results?

No. Once tracking is live, the Open Tracked analytics shows spend, revenue, ROAS and profit, and you can ask Ari, the built-in AI assistant, questions about your own numbers in plain words. Ari's allowance depends on your analytics plan.

Is Open Tracked different from Open Track or Open Space Track?

No. Open Tracked is the official name of our server-side tracking platform, and the website is opentracked.io. Many people search for it as Open Track, OpenTrack or Open Space Track. All of them lead to the same product: server-side Google Tag Manager hosting, first-party analytics and a CRM pipeline.

Fix your tracking at the source

Open Tracked hosts your server-side Tag Manager, keeps first-party cookies alive and shows what each ad is really worth.

Start free More in Troubleshooting →

Written by

The Open Tracked team

We build Open Tracked: server-side tracking hosting, analytics and the tools around them, for stores that spend on ads. Questions about this guide, or an idea for the next one?

Write to us Read the docs

Was this article helpful?

Comments

Loading comments...