We tested whether Meta credits Append Attribution passbacks

appendattribution passback (final).mp4

🧪 Posting the findings here because nobody could give us a straight answer on this one.

The question: does Meta actually credit AppendAttribution passbacks? The documentation says one thing, practitioners say another, and nobody we asked had tested it properly. So we sent real passbacks and watched what happened.

Meta accepted every single one. Which is the trap, because accepted is not the same as credited. Five things decided whether the credit actually landed, and every one of them fails silently.

Test events are not tests. Meta processes them live against your real data. There is no sandbox, so if you have been firing test events to see what sticks, you have been writing to production.

The call matters. AppendAttribution gets credited, AppendClickAttribution does not. Two parameters, nearly identical names, completely different outcomes.

The dataset matters more than people expect. Your passback has to reach the same dataset your ad sets are running against. Send it to a different one and it is accepted, processed, stored and entirely invisible to the campaign you were trying to inform. Nothing errors, the number simply never appears.

Use the standard Purchase event. Custom events do not carry here.

Build fbc from each click, not just the most recent. Two Meta clicks per order is normal rather than an edge case, and passing only the latest hands Meta an incomplete version of a journey it already saw.

The common thread is that none of these produce an error. The passback goes through, the dashboard looks unchanged, and you conclude the method does not work. It did work, it just landed somewhere nobody was looking.

If you are running passbacks today, do the dataset check first.

See how this is configured: https://www.customerlabs.com/request-a-demo/?utm_source=circle&utm_medium=community&utm_campaign=appendattribution

Question for the group: has anyone else tested passback crediting and found something different? Particularly interested if anyone has results on view-through passbacks.

#AppendAttribution #MetaAds #CAPI #Attribution #1PDOps