Trust & methodology
Numbers you can defend in the room.
Every figure shows where it came from, how confident we are and what it's compared with.
Every number shows
- Source · public, connected account or import
- Confidence · low <10, medium 10–29, high 30+
- Comparison · same channel, same days
Methodology
Six rules behind every number.
- Identity
- One confirmed client brand per workspace. Sister brands are labelled, never pooled.
- Attribution
- Brand-owned, creator/earned and paid are always split.
- Data quality
- Source and confidence on every metric; small samples say “not enough data”.
- Fair comparisons
- Same channel, same days, posts of comparable age. Outliers shown separately.
- Grounded AI
- AI answers cite the posts behind each claim and never invent figures.
- Proof
- Lagged commerce links are early signals. Proof comes from holdout tests.
Commitments
What's in place, and what isn't yet.
- IdentityEach brand has one confirmed client identity; sister brands are labelled, never pooled.In place
- AttributionBrand-owned, creator/earned and paid content are split in every metric.In place
- Data qualityEvery metric shows its source and a confidence level based on sample size.In place
- Fair comparisonsBrands are compared on the same channel over periods where both have data.In place
- Equal-age viewsViews compared at day 7 and day 30 once enough snapshots exist.Rolling out
- Grounded AIAI answers cite the posts behind each claim and never invent numbers.In place
- ProofCommerce links are early signals; proof comes from holdout tests.In place
- Connected accountsRead-only official connections for reach and watch time.Planned
Security & privacy
How we protect your data.
See where you're winning in two weeks — get a teaser.
Two weeks: your brand, your competitors, your next three moves.