The distributor’s purchase order says 2,000 cases. Their warehouse shipped 240 cases to stores last week. Both numbers are real. Only one of them is demand.
The difference between sell-in and sell-through, why production planned off purchase orders backfires, how to get depletion data, and the per-store math that turns it into a production number.
What is the difference between sell-in and sell-through?
Sell-in is what you sold to the distributor: their purchase orders. Sell-through is what moved from their warehouse out to stores: their depletion reports. Sell-in is their forecast plus their buffer. Sell-through is demand. Only the second tells you what consumers actually did, so plan production on it.
The two run apart by design. A distributor orders to cover their own service levels, promo bets, and storage math, and none of that obliges a consumer to buy anything. The gap between the numbers is inventory someone will eventually own, and contractually that someone is usually you.
Why does restocking against purchase orders backfire?
Because a purchase order can outrun demand for months before anything looks wrong. Build to the orders and the excess eventually comes back: older, closer to its date, and already paid for. The gap between sell-in and sell-through is exactly your risk.
The pattern has a shape: two or three strong orders, a production run scaled to them, then a quiet quarter while the distributor works down their buffer, and a return authorization for the stock that aged out in their warehouse. Every step looked like growth on the sell-in line.
How do you get depletion data?
Ask for it in the distribution agreement, not after signing. Most distributors can export depletions by account, weekly or monthly; the workable minimum is units shipped to stores per week, by product. If they won’t share any of it, weight their orders accordingly.
The clause costs nothing to ask for at signing and a negotiation afterward. Where a portal exists, take the export cadence you can actually sustain: a monthly number you read beats a weekly feed you don’t.
How do you turn depletions into a production number?
Units per store per week × active doors = real weekly demand. Size production to that run-rate. Treat any purchase order far above it as a question worth asking rather than a signal to build against, because the gap is stock somebody has to own.
| Line | Value | Where it comes from |
|---|---|---|
| Active doors | 200 | distributor account list |
| Velocity | 1.2 units/store/week | depletion report |
| Real weekly demand | 240 units | 200 × 1.2 |
| Purchase order on the desk | 2,000 units | sell-in |
| Weeks of supply in that order | ≈ 8.3 | 2,000 ÷ 240 |
Eight weeks of supply might be fine ahead of a reset, and reckless in a normal quarter. The point of the math is not to refuse the order; it is to know which conversation you’re in before the run is scheduled.
What is GMROI, and what does it tell you that sell-through does not?
GMROI is gross margin return on inventory investment. The formula is one line: gross margin divided by the average cost of the stock you held to earn it. Sell-through says a line moves. GMROI says whether the cash tied up moving it came back with a profit attached.
The arithmetic is one line, and it settles arguments velocity alone cannot. Take the gross margin a product earned over a period and divide it by the average inventory value at cost you carried across the same period. Above one, the line returned more margin than the cash it consumed. Below one, it did not.
| Line | Gross margin, 12 months | Average stock at cost | GMROI |
|---|---|---|---|
| Core hot sauce | 96,000 | 40,000 | 2.4 |
| Seasonal gift set | 30,000 | 50,000 | 0.6 |
Both lines sold through. Only one paid for the cash it used. The gift set needed far more stock on hand to achieve the same movement, and what that stock cost to carry never appeared on the sell-through report. GMROI is the same instinct as inventory turnover, with margin attached rather than units alone.
What is fill rate, and what does it add?
Fill rate is the share of ordered units a distributor actually shipped to stores. Read against sell-through it separates two different failures: a high fill rate with weak sell-through is a demand problem, while a low fill rate means demand existed and simply went unserved.
Sell-through alone cannot tell those two apart, and the difference decides whether a bigger production run helps or does harm. Ask for fill rate in the same clause as depletions. It costs nothing at signing and is a negotiation afterwards.
What do you do when the numbers disagree?
Say so, with the number: “depletions show 240 a week and this order is eight weeks of supply, what’s driving it?” Sometimes it’s a promo you didn’t know about. Sometimes their buffer is about to become your problem. Either answer is worth having before the run.
The same per-account math settles broker renewals: which accounts they actually opened, the margin those accounts produce net of promo support, and whether it clears the commission. The operator’s playbook carries that version of the arithmetic.
What should you do next?
Four moves, one spreadsheet, this quarter.
Get the data
- Put a depletion-data clause in every distribution agreement; ask existing distributors for a monthly export by account.
- Track units per store per week, per product, from the exports.
Use it
- Size every production run to run-rate demand, not to the purchase order.
- Challenge any order above ~6 weeks of supply with the number, before scheduling the run.
Frequently asked questions
What is a good sell-through rate?
There is no universal number: velocity per store per week is the usable metric, and its trend beats any benchmark. A “good” rate is one that is stable or climbing while doors grow; a falling rate with growing purchase orders is the pattern to catch early.
Do distributors share depletion data with small brands?
Usually yes when it’s asked for in the distribution agreement, and grudgingly afterward. Weekly or monthly depletion exports by account are a normal ask; a distributor who refuses any visibility is telling you how the relationship will run.