One engine. One analyst. One answer.
Signal is ingested and analyzed inside our engine. Customers receive scored reads and sourced answers, refreshed on a weekly cadence.
How signal becomes a read.
Three signal layers feed a shared metrics engine. The analyst turns governed functions into paragraph answers, module dashboards, or API reads.
Signal layers
- Visit analytics · 142M devices
- Spend analytics · 14M households
- Household graph · 92M
Fusion & metrics
- Cross-signal joins
- Trade-area & cohort scoring
- Nowcast & incrementality functions
Answer delivery
- Analyst interface
- Module dashboards
- API reads with provenance
Trade-area quality, visit trends, dwell, and competitive pull computed at ZIP and trade-area grain.
Category baskets, MCC-tagged spend, and merchant-attributed nowcasts for commerce intelligence.
Privacy-preserving joins that power cross-shop overlap, incrementality, and share-of-wallet products.
Ask a question. Get a sourced paragraph.
Store visits soft in tier-2 malls; basket size holding.
US visit index is -2.1% QoQ vs consensus +1.4%. Flagship formats flat. Spend-per-visit +3.2%. Cross-shopping with Vuori +27% YoY.
Sephora overlap at 28%, highest in trailing year.
Household overlap with Sephora up 4 pp YoY in Boston and NYC DMAs. Athleta overlap down 11%. White-space strongest in suburban ZIPs with N≥50 coverage.
Defensible by design.
Every ingested record carries its consent chain. Queries that would violate it are blocked at the engine.
Sensitive locations and protected use cases are filtered at ingest and refused at query time.
When consent is revoked, deletion propagates through derived analytics within 24 hours.