A stock plan begins with two different questions: what demand might occur, and what buffer you are willing to finance if reality differs. Demand Forecast & Safety Stock Lab keeps those questions apart and connected — the forecast, its historical test and the buffer economics sit side by side for every product and location. It is built for retail, distribution and production teams in businesses.

For retail, distribution and production teams: review a few products or a larger catalogue across locations. Import your sales history and distinguish genuine zero demand from periods when the product was unavailable. A week that sold out is a gap, not zero sales; a missing date is a missing period, not a zero — and lost demand is never reconstructed silently.

Choose with evidence: compare eligible methods on the same rolling historical test. Inspect absolute error, bias and the number of observations instead of accepting a forecast because its line looks smooth. Each test forecast is fitted only on the history before its period, and a method is ranked only when it can be used at every test period. The tool explains when history is too short for a method — and labels a manual choice as manual.

Plan the buffer: enter delivery-time assumptions, review interval and unit cost. See the target-stock gap and the capital tied up in the safety buffer, with a ladder of 90%, 95%, 97.5% and 99% cycle-service targets. A cycle-service estimate is labelled as a model approximation, not a promise that stockouts cannot happen.

Keep uncertainty visible: historical residual ranges are separate from the point forecast. Missing history, intermittency and seasonal changes remain visible. The tool does not silently treat unavailable stock as zero customer demand.

Worked example from the fictional demo: a filter sells a mean of 10 units a week with a standard deviation of 2; the lead time is 3 weeks with a standard deviation of 1 and stock is reviewed continuously. The variance over the protection period is 3 × 2² + 10² × 1² = 112, so σ ≈ 10.58. At a 95% cycle-service target the whole-unit safety buffer is 18 and, with a flat forecast of 10 a week, the target stock is 30 + 17.4 → 48. With 20 on hand, 10 on order and 5 backordered, the inventory position is 25 and the gap to target is 23 units. At $4 a unit the buffer alone ties up $72.

For medium-sized catalogues the same engine runs every product and location in one calculation — in a background worker for large histories, so the page stays usable. Filter by location, department or result, see which method each product received, and compare buffer capital per currency. Currencies are never added together without an exchange rate you enter. No AI guesses: only the documented statistical methods are used, and every result shows its working.

Try the two fixed fictional demos — a small shop with 12 products and a distributor with 240 products across 4 locations and 104 weeks — then buy once for the full online tool and the offline file, or open it with an active PRO subscription.