How do I forecast demand for a brand-new SKU with zero sales history?
With zero sales history you don't forecast, you triangulate, use category benchmarks from similar SKUs you already sell, watch competitor sell-through signals like review velocity and stockout patterns, and start with a small pilot batch (200-500 units) instead of a full production run. Tools like Cogsy can extrapolate reasonably well once you have even 4-6 weeks of real sales data feeding the model. For a true zero-history launch, treat your first order as a test-and-learn batch you're financially fine losing money on if it flops.
Go deeper
4 resources, 1 India-specific, 2 link-checked.
📄 Article
✓ Link checkedIndiaFreeIntermediate
From an AI-driven inventory/merchandising platform that reports 20-25% forecast accuracy gains for brands using it, useful for understanding what a serious optimization layer actually improves versus a basic OMS.
Frames inventory as a working-capital and cash problem first, tooling second, the right mental model for a founder who's about to tie up money in stock. Includes tool picks by revenue band, which is more useful than a generic 'best practices' list.
A practical walkthrough of how demand forecasting actually works for a DTC brand replacing spreadsheets, useful whether or not you end up buying the tool, because it explains what inputs (sales velocity, lead times, seasonality) any forecasting method needs.
The category-defining tool for founder-led DTC brands who are still running inventory off spreadsheets, it predicts stockouts, recommends PO quantities against supplier lead times, and quantifies revenue lost to being out of stock, which is the number that actually gets budget approved.
