Forecasting: multi-channel AI vs Shopify-tuned baseline
Cogsy's forecasting is intentionally simple — baseline projections using Shopify sales history with optional manual overrides. For a Shopify-only brand under £5M, this is enough. Replenishment cycles are predictable, demand is concentrated in one channel, and a planner can sanity-check the numbers in five minutes. It struggles where multi-channel brands actually live. Amazon demand moves on its own clock, often driven by Buy Box dynamics and Prime Day windows. Retail sell-through has weekly cadence and replenishment-bound demand smoothing. Promo cycles cross channels but don't always run in sync. Lumina's AI model trains across all of this — channel-level demand, returns, promo calendars, retailer EDI — and produces a forecast that reflects the multi-channel reality rather than a Shopify-only projection of it.
- AI model trained on Shopify, Amazon, and retail data together
- Handles cross-channel promo cycles and returns spikes
- Forecast updates propagate to replenishment and cash flow automatically
- Single Shopify store with stable monthly cadence
- Small SKU count where intuition can sanity-check the forecast
- Brands where forecast accuracy isn't the binding constraint yet