Working prototype · public seasonality + lead-time context · peak-season commit decision

The order you place months early, against a forecast you cannot trust yet.

Toys sell in a short holiday window but source months ahead from Asia, so the buy quantity is committed long before demand is known. Peak Commit takes one seasonal SKU, a long lead time, and a forecast-error band, and shows the real tradeoff a planner lives in: order light and stock out, order heavy and mark down.

This is the planner's problem, not the CFO's: forecast accuracy, lead time, and the commit quantity are your levers, set in the S&OP room quarters ahead of the print. Pick a SKU profile: Peak Commit runs a synthetic Q4 demand curve through the lead time, scores the commit risk 1 to 99, and lands the recommended order so the verdict is on screen before you touch a slider. The one slider that matters, the commit quantity, sits right under the verdict.

01 · Commit a seasonal SKU

One click lands the recommended order.

Each preset loads a seasonal toy SKU: a Q4 demand peak, the Asia lead time that forces an early commit, and how wide the forecast could be wrong. The verdict shows the recommended commit and the expected cost of being wrong. The commit slider is the one control that matters; the rest are optional.

One click loads a SKU and lands the recommended commit. Drag the commit slider to feel the tradeoff.

Seasonal SKU · commit risk

...

Planner moves at this band

Commit quantity (units ordered now, against the long lead time)

The cost of being wrong

At this commit, where the expected cost lands

Demand vs commit

Synthetic Q4 demand band and the units you committed

demand band (forecast error) sold from commit excess to mark down

Tune the SKU optional · sliders

Forecast error band (how wide you could be wrong)

Gross margin on a sold unit

Markdown depth on unsold peak stock

What this is and is not. The simulator reports the expected cost of the commit decision, split into lost-margin stockouts and markdown on excess, scored 1 to 99. It uses a classic newsvendor balance (the optimal commit sets the chance of selling the last unit equal to the cost ratio) so the recommended order is a known, defensible result, not a guess. The demand curve, the lead time, and the cost inputs are synthetic and tuned to the structure of a seasonal toy SKU, never to Mattel's real demand. The production extension reads the real demand history, lead times, and unit economics from the planning system.

02 · Sources & method

The structure is real, the SKU is synthetic.

SeasonalityToys concentrate sales into the November to December window while production builds through the first three quarters ahead of the holiday peak; ocean capacity and rates are committed in advance. Source: Mattel Form 10-K (seasonality & supply-chain risk factors): MAT 10-K filings (EDGAR)
Forecast accuracyThere is no single "good" MAPE; accuracy gains translate directly into inventory and service, and unrealistic targets in volatile segments cause overcorrection that raises inventory without improving service. Source: ToolsGroup (2026-03-05)
S&OP cadenceThe S&OP cycle is almost always monthly, with an executive meeting where leadership commits to one plan; decision follow-through is the metric most predictive of high performers. Source: Onepint.ai (2026-05)
China sourcingRoughly 80% of toys sold in the US are made in China; granular trade data puts toy-import China share at 73% to 78%, which is what makes the lead time long and the early commit unavoidable. Source: Al Jazeera (2025-05-11)
MethodThe optimal commit is the newsvendor critical-ratio result: order to the demand quantile where margin / (margin + markdown) is met. The expected stockout and markdown costs integrate a normal demand band whose width is the forecast-error input. The math is in the page source between marker comments.
SyntheticThe demand curve, the lead time, the forecast-error band, and the unit economics are synthetic, tuned to the structure of a seasonal toy SKU and labelled wherever shown. No Mattel demand or cost figure is used.