SPINS API vs. portal — short memo (LMNT)
To: Shivani · From: Brainforge · Date: 2026-04-01
Recommendation
Add the SPINS Sales API (~$15k/year, per recent discussion). Keep using the portal (Satori) for spot checks. Use the API so SPINS lands in Snowflake on a schedule—same definitions every run—next to Emerson and the rest of your commercial stack. Without the API, SPINS stays a manual side channel; with it, it can support OKRs, leadership cuts, and competitive/category views without re-building exports each month.
What is SPINS? (one paragraph)
SPINS is syndicated retail POS data plus product intelligence—market-wide sell-through and attributes across natural, specialty, and conventional retail (and related channels SPINS describes on their site). It answers “what’s selling in the market, vs. us and competitors” in a way first-party sell-in alone cannot.
More detail from SPINS: Point-of-Sale Data overview
What the API adds (vs. the portal only)
SPINS’ own docs describe the Sales API as GraphQL extracts into your warehouse / BI—large, automated pulls (e.g. Parquet / JSON / CSV), multi-year history, standard weekly/monthly/YTD-style periods, async jobs—instead of repeated manual downloads from Satori.
Full product detail (measures, attributes, auth, FAQ, release schedule): SPINS Sales API documentation
Worth one look in the docs: aggregation is time / product–oriented; geography aggregation in the API is limited today—retailer/geo views often land in BI from the right base grain.
Why this matches what LMNT is trying to do
- Less time on manual retail pulls and reconciliation; more repeatable numbers for OKRs and leadership.
- Splits and retailer/period cuts you need monthly are painful if every run is a new export from a UI.
- You’ve already pointed to SPINS for competitive and category questions; the API is how that stays joined to the single certified commercial view in the warehouse.
What we’ve seen elsewhere (brief)
On another CPG implementation: API → storage → warehouse, expect a short alignment phase on measures/windows, plan backfill + multiple timeframes up front, and keep one base grain in the mart, roll-ups in BI—same playbook we’d use for LMNT.