US TIKTOK SIGNAL LAB · PUBLIC BETA

What we can observe.
What we cannot claim.

An evolving research log built from permissioned, anonymized NativePort operations—not recycled platform folklore.

PUBLIC COHORTS0

We will not publish premature statistics.

MINIMUM BATCH10 posts

One upload is not a benchmark.

PRIMARY REPORT30 posts

Enough structure for learning sprints.

01 · RESEARCH QUESTION

Which signals can an operator control?

We can control workflow consistency, publishing context, account preparation, content instructions and measurement. We cannot control recommendation eligibility, moderation, reach or viewer response.

02 · MEASUREMENT

What each cohort records

  • Physical-device and native-app publishing context
  • Account age and preparation window
  • Niche and intended audience
  • Comparable creative batches
  • 2-second hold, watch time and completion
  • Audience geography when available
  • Profile visits, qualified actions and conversion
  • Moderation or operational incidents
03 · EVIDENCE LADDER

How claims earn confidence

  1. Documented factPrimary platform documentation
  2. Repeated observationComparable permissioned cohorts
  3. Working hypothesisUseful but not causal
  4. UnknownNot enough evidence
04 · RELEASE POLICY

What happens next

The first report will be released only when a full cohort is complete, privacy review is finished and limitations are documented. Until then, zero is the honest number.

Read the full methodology →
Chat on WhatsApp