Inventory margins: know what you make on every product

Ask a business owner their overall margin and most can answer. Ask what they make on one specific product and the room goes quiet. That silence is expensive: a blended margin is an average, and averages hide exactly the thing you need to see — which products carry the business and which ones quietly ride along for free.

COGS in one sentence

Cost of goods sold is what it directly cost to deliver the thing you sold — and it hits your books when the item sells, not when you bought the stock. Buy $8,000 of materials in June and your cash drops in June, but your P&L records cost bag by bag, as each one goes out the door. That’s what makes a true per-product margin possible: every sale carries its own cost with it.

The roaster: what a $16 bag actually costs

Take a small coffee roaster with two products: a 12 oz retail bag and a 5 lb wholesale bag, same coffee inside. Green beans cost $4.20 a pound. Here’s the walk from raw cost to true cost:

• Roast loss. Roasting drives off moisture — about 15% of the weight. A pound of roasted coffee therefore costs $4.20 ÷ 0.85 = $4.94, before anything else happens to it.
• The coffee in the bag. A 12 oz bag holds 0.75 lb of roasted coffee: $4.94 × 0.75 = $3.71.
• Packaging. Bag, label, and one-way valve: $0.62.
• Freight in. The shipping you paid to get green beans to the roastery is part of the cost of the beans. Spread across the order it adds $0.17 a bag.

True cost per retail bag: $4.50. Sell it at $16 and the gross profit is $11.50 — a 72% margin. Now the wholesale bag: 5 lb of roasted coffee is $24.71, plus $1.10 of packaging and $0.55 of freight, for a cost of $26.36. It sells to cafés at $52. Gross profit: $25.64 — a 49% margin.

What about the hour spent roasting? Reasonable people disagree. Larger operations put direct labor into COGS; most small businesses leave it in operating expenses because splitting one person’s day across products costs more precision than it buys. Either answer works — what breaks the numbers is switching between them, because the moment the rule changes, this month’s margin stops being comparable to last month’s. Pick one and hold it.

Neither number is wrong, and neither is the whole story alone. But the P&L’s blended margin — somewhere around 60%, depending on the mix — told the owner nothing about the actual decision on the table: whether to chase more wholesale accounts. Per-product margins turn that from a feeling into arithmetic. Wholesale at 49% can still be a great business — the volume is steadier and the selling cost per bag is lower — but now it’s a choice made with open eyes.

When costs move: the average does the remembering

Green coffee doesn’t stay at $4.20. Buy the next lot at $4.65 and the honest cost of “a bag of coffee” shifts. Average costing handles this without heroics: your remaining stock’s cost blends with each new purchase, and every sale carries the average that’s true at that moment. What matters isn’t the method’s name — it’s that the cost travels with the sale automatically, so your February margin reflects February’s bean prices without anyone rebuilding a spreadsheet.

What to do with per-product margins

• Reprice the outliers. A product 20 points below its siblings either earns a higher price, a cheaper input, or a reason — strategic loss-leaders are fine, accidental ones aren’t.
• Chase freight, not just unit price. Freight-in is part of COGS and it’s invisible until you allocate it. Consolidating two shipments into one is a margin improvement no customer ever notices.
• Watch the drift, not just the level. A margin sliding from 72% to 66% over three months is a supplier increase or a discounting habit — and it’s the earliest warning you’ll get.
• Kill or bundle the losers. A product that can’t earn its margin after repricing can still earn its keep in a bundle — or it can stop consuming shelf space and attention.

None of this requires a quarterly spreadsheet project. Tracked inventory posts cost of goods at every sale, which means the per-product view is a report you open, not an analysis you commission. The roaster’s $16 bag was always making $11.50. The only change is that now somebody knows.