Reference Documentation

Data methodology

How the core fields in the influencers.club platform are computed, validated, and kept honest across all supported platforms.

5 features
13 data points
6 audience data points

Data points

13 data points
Type of Profile Creator Profile
Determines whether a profile is a brand or a creator.
Avg F1-score93%
Gender Creator Profile
Determines a creator's gender from their public profile.
Avg accuracy92%
First Name Creator Profile
Determines a creator's first name from their public profile.
Accuracy96%
Location Creator Profile
Pinpoints where a creator is based down to country, region, and city.
Avg accuracy95%
Language Creator Profile
Identifies which language a creator publishes their content in.
Accuracy93%
Engagement Rate Content & Engagement
How much interaction a creator's content gets relative to their audience size.
Coverage100%
Email Creator Profile
Finds the best email to reach a creator at.
Personal Website Creator Profile
Analyzes all URLs used by a creator to find their personal website.
Follower Growth Growth & Trends
Audience size across platforms, how fast it's growing, and how much of it is real.
Coverage100%
Posting Frequency Content & Engagement
How often a creator publishes new content, measured as posts per month.
Coverage~95%
Hashtags & Mentions Content & Engagement
Hashtag usage patterns, brand mentions, and sponsorship detection from captions.
Coverage~90%
Links in Bio Creator Profile
Detects URLs in creator bios and identifies link-in-bio services like Linktree.
Coverage~72%
Views Growth Growth & Trends
How a creator's view counts are trending over time across multiple periods.
Coverage~80%

Audience data points

6 data points

All audience data points share one approach: they're estimates extrapolated from public engagement signals on a creator's content. The reasoning is explained on the Audience Analytics page. Live across the major platforms today, with more coverage landing every month.

How we validate every data point

Two-track process

Every accuracy number on this site comes out of a two-track validation process. We don't trust a single source: AI by itself can be biased by training data, and human reviewers alone can't cover the scale we need. So we run both, in parallel, and compare them.

Multi-AI Cross-Validation
  • At least 2 independent AI sources label the same data
  • Tens of thousands of records per validation cycle
  • Outputs are compared against each other to surface disagreements
  • Two AI sources agreeing is a much stronger signal than one
Stratified Manual Review
  • At least 1,000 records reviewed per subset
  • Subsets defined across country, language, follower bucket, account type, niche
  • Catches edge cases random sampling would miss
  • Reviewers blind to the pipeline's prediction