Sign in

Blog · Forecast and pipeline · FMCG and CPG brands

Sell-in against sell-out per retailer: the inventory building in the channel

How a CPG brand compares what it shipped to each retailer with what the retailer sold through, from the shipment ledger and the sell-out data: the channel inventory implied by the difference, weeks of cover per retailer and per SKU, the retailers where sell-in has run ahead of sell-out for a quarter, the promotions that loaded the channel, and why a strong shipment quarter with rising channel inventory is next quarter's problem.

The short answerSell-in is what the brand shipped to a retailer; sell-out is what the retailer's stores sold. Cumulative sell-in less cumulative sell-out, per retailer and per SKU, is the inventory in the channel, and divided by the sell-out rate it is weeks of cover. A retailer whose weeks of cover has risen for a quarter is holding stock the brand already booked as revenue, and the next quarter's shipments will fall by that amount whatever the sales team does. The identity is that the implied inventory never goes negative by more than a stated tolerance; where it does, the sell-out data is incomplete.

A brand's shipment quarter is strong. Its sell-out data says the retailers sold less than they received, and the difference is sitting in distribution centres and backrooms. Next quarter's shipments will be short by that much, and the sales team will be asked why. This guide sets out sell-in against sell-out per retailer, weeks of cover, the rising-cover list, and the identity.

The measures

Per retailer, per SKU, per week:

Implied channel inventory = Σ sell-in to date − Σ sell-out to date, from a stated starting inventory Weeks of cover = implied inventory ÷ average weekly sell-out, trailing eight weeks Trend = weeks of cover now against the trailing four quarters' average

Per retailer:

Share of SKUs with rising cover; the sell-in value ahead of sell-out this quarter

The rows you need

  • Shipment ledger: retailer, SKU, week, units, value.
  • Sell-out: retailer, SKU, week, units.
  • Promotion calendar: retailer, SKU, weeks.
  • Starting inventory: retailer, SKU, date, where known or stated.

Retailer identifiers only.

The identity

implied inventory ≥ −tolerance, per retailer per SKU per week

Inventory cannot be negative. Where the implied figure falls below zero by more than the tolerance, sell-out exceeds sell-in, and the sell-in data is missing shipments or the sell-out data double-counts. The row is listed, not floored.

A worked view

Retailer SKU Sell-in this qtr Sell-out this qtr Implied inventory Weeks of cover Trailing avg Reading
R-2207 K-104 48,000 31,000 41,000 11.4 5.2 Rising: loaded
R-2207 K-109 12,000 12,500 4,800 3.3 3.5 Fine
R-4471 K-104 30,000 29,000 9,200 2.8 3.0 Fine
R-9034 K-104 21,000 8,000 38,000 41 6.1 Loaded; promo not in calendar

Retailer R-2207 holds eleven weeks of K-104 against five normally. Seventeen thousand units of this quarter's revenue are next quarter's shortfall. Retailer R-9034 has forty-one weeks and no promotion in the calendar to explain it, which is a question for the account manager before it is a question from the CFO.

Per retailer

Retailer SKUs rising Sell-in ahead of sell-out Next-quarter shipment effect
R-2207 14 of 40 $410,000 −$410,000 at stated sell-out
R-9034 9 of 22 $290,000 −$290,000

Where it goes wrong

Sell-in reported alone. A strong quarter that is next quarter's miss.

Sell-out coverage unstated. Weeks of cover computed on half the SKUs, presented as all.

Starting inventory unknown. The level is wrong; the trend still holds. State it as trend only.

Promotions not on the page. A legitimate promotional load looks like a loading, or the reverse.

Every week, cover per retailer per SKU

Mapped once, the shipment ledger, the sell-out files and the promotion calendar produce implied inventory, weeks of cover, the trend and the identity every week. Covirage builds this from the exports as they are. The CPG brands page describes the setup, and the promo lift guide covers the promotional shipments that this measure separates from loading.

Questions people ask

Where does sell-out come from?

Retailer portals, distributor sell-out files or syndicated data, weekly per retailer and SKU. Coverage is rarely complete; the report shows the share of sell-in that has matching sell-out data and computes weeks of cover only where it does.

What is a normal weeks of cover?

It depends on the retailer's replenishment cycle and the SKU's velocity. The useful comparison is each retailer-SKU against its own history: weeks of cover this quarter against the trailing four. Rising is the signal; the level is context.

How is a promotion separated from a loading?

A promotion pulls sell-out up and sell-in up together, and weeks of cover holds. A loading pushes sell-in up with sell-out flat, and weeks of cover rises. The promotion calendar on the same page says which shipments were promotional.