FCST (Forecast)

Amazon & D2C glossary · Harpy Media

FCST (Forecast) is the projected future — of sales, demand, or inventory — used to plan what to make, order, ship, and spend. On Amazon, forecasts exist on both sides of the table: Amazon’s own models forecast category demand and replenishment needs; sellers and vendors build forecasts to run their businesses against.

What is FCST?

FCST (Forecast) is the projected future — of sales, demand, or inventory — used to plan what to make, order, ship, and spend. On Amazon, forecasts exist on both sides of the table: Amazon’s own models forecast category demand and replenishment needs; sellers and vendors build forecasts to run their businesses against.

Amazon’s forecasts are a service and a signal: vendor teams share projections (and expect you to plan production around them), the platform’s replenishment engine orders against its own views, and demand-planning tools offer guidance. Treating the platform’s forecast as gospel or as noise are both errors; the professional posture is triangulation — Amazon’s view, your own model, and the market’s behaviour, reconciled regularly.

What Amazon’s forecast is good for

Three uses. Production planning: vendor forecasts indicate expected purchase volumes — useful for raw-material commitments and production scheduling, provided you remember they are projections, not contracts (Amazon is famously non-binding about them). Event preparation: Prime Day, Q4, and promotional periods arrive with demand-shape guidance (which ASINs, roughly how much), and the vendors who prepare stock accordingly capture the spike while the unprepared watch it. Triangulation: when Amazon’s model, your model, and recent actuals disagree, the disagreement itself is information — it points at seasonality you missed, promotions you forgot, or category shifts neither of you has priced yet.

Building your own model beside theirs

The seller-side discipline (see demand forecasting for the math): baseline velocity × seasonality × promo lift, tracked monthly against actuals, with error rates scored. The specific Amazon survival skills: never commit production capital against platform forecasts alone (blend them down into ranges), watch the platforms’ tools for signals rather than instructions (their incentives include network optimisation, not just your demand), and keep your forecast as the budget’s foundation — then use variance reviews to improve it. One rule of thumb from the vendor world: Amazon’s forecast tells you what they might buy; your sell-through data tells you what customers actually did. Both matter; only one can be audited.

In practice

A vendor receives Amazon’s Q4 forecast calling for a 40% lift on two ASINs. Their own model, built from sell-through and last year’s shape, says 25%. The plan splits the difference on production (with an option tranche for the upside), and the actual lands at 27%. No stockout, no warehouse of forecasts-coming-true-later. The gap between the two forecasts had been the most useful data in the meeting.

How Harpy Media helps

Monthly forecast scoring — Amazon’s view versus ours versus actuals — is a standing routine in our planning work; the gaps tell us more than any single model does.

FCST FAQ

What is FCST on Amazon?

Forecast — projections of sales, demand, or inventory used by Amazon and by sellers/vendors to plan production, ordering, and logistics.

Can I rely on Amazon’s forecast?

Use it as one input, not a contract: it’s non-binding, shaped by the platform’s needs too. Blend with your own model and score both against actuals.

Where do sellers find forecast tools?

Vendor Central and Seller Central offer demand and replenishment guidance; third-party tools add layers. Either way, keep your own scored forecast as the budget’s foundation.