Product Discounts: Profit and Money Tied Up in Stock
Compare discounts on bestsellers, seasonal stock and slow sellers: what each sale leaves you, and when recovering cash matters more.
A single discount across the whole catalogue looks convenient. But the products in that catalogue serve different purposes. One already sells well and brings in money. Another is left over from last season. A third is taking up shelf space, while the money tied up in it is needed to buy products customers actually order.
So I would start with a question, rather than a discount percentage: what do we want this product to achieve? Earn more from sales, or recover some of the money already tied up in stock? The answer affects both the right price and how much we can spend on making the sale.
Let’s see what an ordinary order leaves us
Take a hypothetical online lighting shop. All amounts below are illustrative and exclude VAT. Each order contains one lamp. It sells for €100, costs €60 to buy, and incurs another €20 in variable costs. That €20 already includes advertising, the payment fee, packaging and the share of delivery paid by the shop. This leaves €20 per order before fixed costs.
It is the same logic we explore in ecommerce unit economics: revenue alone does not tell us how much money a sale leaves the business. And buying the product is not the only cost.
Now let’s offer a 10% discount. The customer pays €90. If the purchase cost and other costs per order remain unchanged for now, the shop has €10 left. The price has fallen by a tenth, but the amount left from the sale has halved.
Suppose we would have sold 100 of these lamps over the chosen period without the promotion. They would have left €2 000 before fixed costs. With the discount, we need 200 orders to retain the same amount: €2 000 ÷ €10.
That is what the promotion needs to achieve, not a forecast of demand. We do not yet know whether the discount will bring twice as many orders. If attracting them requires more expensive advertising, the calculation gets worse. If the payment fee depends on the amount paid, we need to recalculate it at the new price instead. We also need to check that we have enough stock and capacity to process the orders.
What if the lamp hardly sells?
On another shelf, we have an old model. It has been in stock for a long time and its product page is available, but there is hardly any demand. I would first make sure this is not a stock-record error, a new product we have not observed for long enough, or a seasonal item whose time has yet to come.
If it really is slow-moving stock, comparing the discount only with the old price is no longer enough. A sale at full price may exist only in our hopes. We need realistic alternatives: return it to the supplier, sell the stock wholesale, offer a useful bundle, reduce the retail price or wait for the right season.
Suppose we bought this old lamp for €60. We are now prepared to sell it for €50, and the new costs of selling it will be €8. The sale recovers €42. Against the purchase cost, we lose €18 before fixed costs. It is not a profitable sale, even if money has finally reached our account.
But suppose the supplier is only willing to refund €38 for it, after return shipping costs. The retail option could then recover €4 more. To get that difference, we still have to find a buyer and wait for payment. Returning it to the supplier may make more sense if that option is confirmed and retail demand remains only a hope.
Our separate article on selling slow-moving stock compares these options in more detail. The distinction here is that we protect earnings from an ordinary lamp sale, while for confirmed dead stock we choose an acceptable way to recover money. Those are two different objectives for a discount.
Seasonal stock raises another question
Now imagine a set of outdoor string lights. If its season is still ahead, a lack of orders now is not, by itself, a reason to clear the stock. I would look at sales in the comparable previous season, whether the model is still relevant, and the cost of storing it until demand returns.
If the season is ending, we need to compare money now with a possible sale later. Will the model still be relevant? What will storage cost? What else could we do with the money recovered? There is no universal age at which every product becomes dead stock.
A bundle needs its own calculation too. If we add a popular light bulb to the slow-moving lamp, we are giving away another product with its own purchase cost. Calling it a “gift” does not remove that cost. Let’s compare what is left from the whole bundle, without assuming that the benefit from one item applies to the entire order.
Let’s record a decision for each product
Instead of a blanket instruction to “take 10% off everything”, we now have a short working list. For each product, we record the objective, the price calculation, a limit on additional costs and a date to review the result.
For the old lamp, we can set a specific condition: I want to recover at least €42 per unit at a price of €50. That leaves a maximum of €8 for all the new selling costs combined. If packaging, the payment fee and delivery already take €6, no more than €2 per unit sold is available for advertising. This is a calculated limit, not a promise that we can find a buyer at that cost. A trial sale also needs an overall spending limit: the advertising money could run out without a single order.
We cover the technical side — submitting a new price and displaying a discount — in our Shopping pricing guide. But first, we need to choose the products and calculate our options. After the promotion, we can then assess what it was meant to achieve: whether ordinary sales earned enough and how much money we recovered from stock.
Bought for $50. Sell it for $40?
How to price slow-moving stock: calculate cash left after costs, compare ways to sell it and keep the clearance from consuming the money you want back.
Product Bundling Around a Bestseller to Raise Order Value
Earn more per order from the shoppers your bestseller brings: a worked example of how big a bundle discount can be, and when to add the bundle to Shopping ads.
Ecommerce Pricing Strategy for Google Shopping Ads
See whether to change price, shipping or discounts first in Shopping ads. In our data on 213,913 products, a smaller share sold if priced far above the market.
Price Testing on Google Shopping: Raise, Cut or Hold?
Find which products can take a higher price and which need a cut, then change prices about 5% at a time so Shopping ads and Smart Bidding keep working.
50% Markup: Working Out the Budget for Ads
We work through one order to separate markup from margin, account for delivery and payment fees, and calculate how much we can spend on ads.
ROAS Formula: How to Calculate It and When Not to Trust It
Get the ROAS formula, a worked example and checks that show if your Google Ads figure is real: a fixed value on every order can turn a true 4.5 into 3.75.