Signal: Seasonality
Seasonality is the demand calendar overlay that models predictable, recurring variation in consumer demand by category and geography. It is not real-time — it is a structured forward-looking curve that the Forecast, Pricing, and Commercial agents use to shape plans before the season arrives.
Purpose
SMB retailers on Counterpoint operate without the demand planning infrastructure that tier-1 retailers take for granted. The seasonality signal gives the local market agent the ability to inject a calibrated demand curve into the Forecast and Pricing modules — holiday windows, back-to-school, weather-driven seasonal transitions, local event calendars. Without it, the Forecast module plans from historical transaction data alone and misses the forward signal.
Signal
The seasonality signal is a structured calendar of demand multipliers and event windows:
| Component | Description | Granularity |
|---|---|---|
| Demand multipliers | Index (1.0 = baseline) by category and week | CAT_CATEGORY × GEO_DISTRICT × week |
| Event windows | Named periods (Holiday, Back-to-School, Harvest, etc.) with start/end dates | GEO_REGION × named event |
| Transition windows | Seasonal transition periods where category mix shifts (Summer → Fall) | CAT_DEPT × GEO_REGION × date range |
| Local event overlays | Non-national events: regional festivals, school schedules, sports seasons | GEO_DISTRICT × event name × date range |
Signal frequency: updated weekly by the local market agent. Major event windows are seeded 90 days forward; multipliers are revised as the window approaches.
Sources
- National retail calendar (NRF, major holiday schedules)
- Regional school district calendars (local market agent scrapes or subscribes)
- Local event listings (Chamber of Commerce feed — see signal-community-intel)
- Historical transaction patterns from Module T (used to calibrate multipliers for the geography)
Consumers
| Consumer | Action |
|---|---|
| Module O (Orders) | Applies demand multipliers to forecast buckets; triggers early replenishment orders ahead of peak windows |
| Module P (Pricing & Promotion) | Schedules promotional windows aligned to seasonal peaks; suppresses discount promotions during high-demand periods |
| Module M (Merchandising) | Commercial strategy adjusts OTB allocation by season; adjusts range emphasis |
Invariants
- Seasonality multipliers are overlays, not overrides. Module O applies them to the base forecast; they do not replace the statistical model.
- Local event overlays are GEO_DISTRICT scoped. They are not propagated to other districts.
- The signal is forward-looking only. Historical seasonality patterns live in Module T's transaction history — the agent does not re-model the past.
Related
- local-market-agent — the agent that sources and delivers this signal
- signal-weather-seo — weather is a related real-time signal that interacts with seasonal demand
- category-hierarchy — demand multipliers are applied at CAT_CATEGORY level
- geography-hierarchy — multipliers are scoped to GEO_DISTRICT or GEO_REGION