P80

Amazon & D2C glossary · Harpy Media

P80 is a demand forecast confidence level: the unit volume that weekly demand is expected to stay at or below 80% of the time. Where the P50 forecast is the balanced midpoint and P70 adds a moderate cushion, P80 is a deliberately conservative plan — one that accepts carrying slightly more stock in exchange for covering most of the upside.

What is P80?

P80 is a demand forecast confidence level: the unit volume that weekly demand is expected to stay at or below 80% of the time. Where the P50 forecast is the balanced midpoint and P70 adds a moderate cushion, P80 is a deliberately conservative plan — one that accepts carrying slightly more stock in exchange for covering most of the upside.

It exists for products where being short is worse than being heavy. Volatile categories, high-growth SKUs, and seasonal peaks all punish understocking more severely than they punish a little extra inventory — so planning at the 80th percentile is a rational choice rather than an over-cautious one.

When the conservative tier earns its keep

Three situations justify P80 planning. Volatility: when demand swings widely, the cost of being caught short rises because restocking cannot react quickly enough. Growth: a rising trend means trailing averages understate the future, and a buffered forecast corrects for it. Peaks: when a predictable spike is coming, the pipeline has to be filled before it arrives, and the penalty for arriving empty is rank damage rather than a delayed sale.

The trade-off is explicit: more capital committed to inventory and more storage exposure. That cost is why the tier should be chosen per SKU rather than applied across a catalogue. A slow, steady consumable does not need an 80th-percentile plan; a trending gadget heading into its first fourth quarter very often does.

Reading and using the number

An P80 forecast of 1,000 units per week means the platform assesses an 80% probability that demand will not exceed 1,000 in a given week — and a 20% chance it will be higher. That is the planning number for production and replenishment: the figure the supply chain should be able to serve, not the number expected on average.

The forecasts appear in vendor reporting — demand and forecast reports, flow-planning and buy-plan views, and the replenishment models behind Amazon’s own ordering. The vendor-side discipline is to reconcile production capacity and lead times against the chosen tier, and to keep the choice stable. Switching tiers reactively — chasing P50 when cash is tight, jumping to P90 when the platform complains — produces the volatility you were trying to manage.

In practice

For a trending fitness tracker, the platform sets a P80 weekly forecast of 5,000 units as the fourth quarter approaches. The vendor plans production and inbound freight to serve that figure rather than the mean: the pipeline is filled before the peak, purchase orders are met in full, and the product holds its ranking through the highest-traffic weeks of the year without carrying enough excess to create a January storage problem.

How Harpy Media helps

Forecast-tier planning is part of how we run vendor supply conversations: volatility and peak exposure assessed per SKU, buffers chosen against the real cost of being wrong, and production plans reconciled with the platform’s own forecast reporting.

P80 FAQ

What does P80 forecast mean?

It means the platform assesses an 80% likelihood that demand will stay at or below that weekly unit figure — a conservative planning level that builds a buffer against demand running higher.

Should I always plan at P80 or P90?

No — the level should match the product’s economics and volatility. Higher tiers protect against stockouts at the cost of carrying more inventory; for steady, predictable products, the extra buffer is often money sitting idle.

Where do P80 forecasts appear?

In Vendor Central reporting — demand and forecast reports, flow-planning and buy-plan dashboards — and in the automated replenishment models Amazon Retail uses to size purchase orders.