How TagFrog works.
TagFrog is a five-stage tracking pipeline. Events enter at the source, get shaped by a module, move through a transport, land at one or more destinations, and the verification layer watches the whole thing.
-
Sources
Where the data starts. The storefront emits the user-facing events (page views, product views, add-to-cart, checkout). The server emits the truth events (order created, refund, fulfillment). Mobile and embedded apps emit the same shape via a thin SDK so the rest of the pipeline doesn't have to care which surface fired.
-
Modules
How we tag and structure events. Each storefront platform has its own quirks (Shopify checkout extensibility, WooCommerce hooks, BigCommerce script manager). Modules normalize that surface into one consistent event schema so downstream destinations see the same shape regardless of where the order was placed.
-
Transports
How it moves. Browser events go through a thin client that respects consent and dedups against the server stream. Server events go through our custom Go event router, which buffers, retries, and signs the payloads for each destination's API contract. One transport stack, every platform.
-
Destinations
Where it lands. GA4 (browser and Measurement Protocol), Meta Conversions API, TikTok Events API, Pinterest, Google Ads, Microsoft Ads, Snapchat, your warehouse. Each destination formatter is maintained against the live spec, so when Meta moves to a new CAPI version we ship the update before your reports go quiet.
-
Verification
How we prove it's actually working. Continuous monitoring on every account: event volumes, parity between browser and server, dedup keys, destination ack rates. When something drifts we flag it before the ads team notices the report is wrong. That's the part that makes the whole thing worth installing.