whispr LabswhisprLabs

Blog & research

Notes from our data team on the creator economy, benchmarks and how the product works.

Research· 6 min read

Content ideas from a brain scanner: reading Meta’s TRIBE v2

Meta trained a foundation model on more than 1,000 hours of fMRI to predict how people respond to video, audio and language. It is not an engagement model — but five testable content ideas fall out of it.

Read more

Research

How we score audience authenticity

A look under the hood at the signals — mass followers, inactive accounts, suspicious patterns — that power the audience-quality score.

In the works

Benchmarks

Engagement benchmarks by niche in India

What a “good” engagement rate actually looks like across fashion, tech, finance and lifestyle, by follower tier.

In the works

Playbook

Pricing a creator: from CPM to negotiation anchors

How to turn reach and engagement into a defensible rate card — and the counters to have ready when a brand lowballs.

In the works

Product

Reading a brand-fit score

Fit is decision support, not a guarantee. Here’s how the components are weighted and how to use them.

In the works

More research is on the way — check back soon.

    Blog & research · whispr Labs