Docs

TikTok Events API's New customer_type Parameter

TikTok's Events API now accepts customer_type (new vs. returning) — track it server-side so attribution reporting stops blending both together.

Revenue

TikTok’s Events API added a customer_type parameter this year, letting advertisers mark each server-side conversion event as coming from a new or returning customer, directly in the payload TikTok uses for its own attribution reporting. Before this, separating first-time acquisition revenue from repeat-customer revenue meant exporting TikTok’s numbers and joining them against a CRM or data warehouse after the fact — a process slow enough that campaign decisions were often made on blended numbers that overstated how much of a campaign’s return was genuinely new business. With customer_type sent at event time, that split now happens inside TikTok’s own reporting, in near real time.

The catch is that this parameter only helps if it’s populated correctly and consistently. A server-side event pipeline that guesses at customer type — or worse, defaults everything to new because that’s what the integration shipped with — doesn’t just fail to add value, it actively corrupts TikTok’s attribution model with confidently wrong labels. Teams running server-side tracking across multiple ad platforms also now need to keep TikTok’s customer_type values consistent with however they’re distinguishing new versus returning customers elsewhere, since TikTok separately ships a payload converter that auto-maps Meta Conversions API events into its own format — and a mismatch between platforms undermines exactly the cross-channel comparison teams are trying to build.

Data Points to Track

  • customer_type population rate, measuring what share of server-side conversion events actually carry a new or returning value versus arriving unset, since an unset field degrades attribution the same way a wrong one does
  • Customer-type determination source, logging whether the label came from your own first-party purchase history, a CRM lookup, or a fallback default, so a data-quality issue can be traced to its origin
  • Cross-platform label consistency, comparing how a given customer is classified in TikTok’s Events API versus other ad platforms’ equivalent fields, to catch drift before it distorts blended attribution reports
  • New vs. returning revenue split by campaign, tracked independently of TikTok’s own dashboard so you can validate the platform-reported split against your own source-of-truth numbers
  • Payload converter mapping accuracy, spot-checking events passed through TikTok’s Meta-to-TikTok converter to confirm customer_type and other fields translate correctly rather than dropping silently

Setup Steps

  1. Determine customer type from first-party purchase history at the point of event capture, not from a heuristic guess, so the label reflects an actual prior-purchase record.
  2. Add customer_type to your server-side event payload schema for every conversion event sent through TikTok’s Events API, with a required (not optional) field to prevent silent omission.
  3. Reject or flag events missing a resolvable customer type rather than defaulting to new, and route them to a review queue instead of sending a guess to TikTok.
  4. Align customer-type logic across every ad platform integration, using one shared first-party definition of “returning” so TikTok, Meta, and any other channel agree on the same customers.
  5. Build a validation report comparing your own new-vs-returning revenue split against TikTok’s reported numbers, run weekly, to catch drift before it skews campaign budget decisions.

Actionable Insights

The value of customer_type is entirely dependent on how it’s populated — treat it as a data-quality problem before it’s a reporting feature. Once population rate is high and the source is first-party purchase history rather than a guess, this parameter finally lets you see, inside TikTok’s own attribution, how much of a campaign’s return is genuine new-customer acquisition versus repeat spend that would likely have happened anyway. That’s the number that should drive budget allocation between acquisition and retention campaigns, not the blended figure TikTok reported before this field existed.

Expert help

Need help tracking this in your app?

Our team sets up analytics pipelines for mobile and web teams every day. Talk to us and get your first events flowing in under an hour.

Talk to an expert