How standard events train delivery

The standard list#

Nine names carry most accounts: PageView, ViewContent, Search, AddToWishlist, AddToCart, InitiateCheckout, AddPaymentInfo, Purchase and Lead, with CompleteRegistration and Contact beside them for account and enquiry journeys. Fixed names matter because Meta models train on them across advertisers; custom names for standard behaviour throw that shared learning away. Meta own standard events reference lists each with its intent.

Map each funnel step to exactly one standard name, and fire it in exactly one place: one Purchase confirmation, one Lead success handler. Double coverage from overlapping triggers is the most common event fault we audit. Pixel setup covers placement.

Parameters that pay#

Names alone optimise for counts; parameters optimise for money. Value and currency turn Purchase into ROAS bidding fuel, content IDs and content type link behaviour to catalogue products, and order or lead IDs power dedup and offline matching. Events without value can only chase volume, which is how accounts full of cheap accidents happen.

Send net values after discounts, keep currency consistent with the ad account, and align content IDs exactly with catalogue IDs or matching fails silently. Parameter discipline is reviewed in catalogue setup, and the shop wiring checklist sits in Shopify integration.

Custom conversions versus custom events#

Custom conversions repackage counted events with URL or parameter rules inside Events Manager, with no code changes; custom events are new coded names for behaviour no standard covers. Prefer custom conversions for filters like high-value purchases, and reserve custom events for genuinely novel actions such as configurator completions or quiz qualifications.

Custom events train on your data alone, so they learn slower and optimise weaker than standards at equal volume. Exhaust the standard list plus custom conversions before inventing vocabulary. The hub places events in the measurement family.

Testing before trust#

Every event earns trust in Test Events mode: trigger it as a visitor would, confirm the name, parameters and value arrive intact, then confirm dedup where browser and server both fire. Test the unhappy paths too: failed payments must not fire Purchase, and bounced form submits must not fire Lead. Developer test tooling is documented at server event testing.

Re-test after every site release that touches templates, carts or forms. Event faults hide inside successful deploys, and a fortnight of misfires costs more than an hour of checks. Dedup checks belong in the same session.

Matching catalogue IDs#

Catalogue matching joins behaviour to products on content IDs: the ID in the event must equal the ID in the feed, character for character. Variant schemes, bundle SKUs and platform-specific prefixes are the usual suspects when match rate sags while events look healthy. Align Shopify, Pixel and catalogue on one ID scheme before scaling, and recheck the scheme after every platform migration.

Monitor match rate weekly as the vital sign of feed health: sudden drops mean an ID scheme changed, usually after a platform update nobody announced. New products should arrive matched from day one, with onboarding checks before spend follows. Feed health lists the flags, and product sets show how matched products become campaigns.

Priority order under AEM#

On opted-out traffic, only the top-ranked event per domain configuration counts, so order events by commercial value with Purchase first and page views last. The default slot order is unsafe for most shops, because it ranks informational events above revenue ones and starves optimisation of buyer signal. Our AEM guide sets the safe order.

Review priority quarterly against margin, not just revenue: the event that pays salaries outranks the event that flatters reports. Keep the ranked list to eight or fewer slots, documented beside domain verification notes. Ask our team for a priority review before peak.

Frequently asked questions#