In short

Measure discount promotion results by how much money the promotion added after the cost of goods and the costs tied to it. Before launch, set the goal and write it down: extra profit, clearing inventory, or attracting customers. If the goal is profit, compare the promotion result with a well-founded estimate of what you would have sold without it.

A discount reduces what you earn on every unit sold, including the units customers would have bought at the regular price anyway. The extra volume must cover that loss, the advertising and the other costs. The worked example below shows how sales can grow while the result gets worse.

Calculate by category and for the whole chain, then check the period after the promotion. Some customers may have switched from a similar product or stocked up on their usual product in advance. Decide on a repeat after these checks. A good-looking sales chart for the promoted item is not enough.

Why H&M's sales and margin moved in different directions

In the first quarter of its 2025 financial year (1 December 2024 to 28 February 2025), H&M Group reported sales growth of 2% in local currencies. Gross margin fell from 51.5% to 49.1%. In its quarterly report, the company named external factors, higher markdowns and investments in the customer offer among the reasons.

A year later the picture changed. In the first quarter of its 2026 financial year (1 December 2025 to 28 February 2026), the group's sales in local currencies fell by 1%, and gross margin rose to 50.7%. In the new report, H&M separately noted a small reduction in markdown costs. To explain the result, the company also pointed to purchasing, external factors and cost control.

Here a markdown means a price cut on a product. One clearance sale cannot explain the whole result of an international group: purchasing terms, currencies, assortment and other decisions all affect it. This is a group result, not a test of one promotion. But the example shows well why sales and the gross margin percentage must be read side by side. They can move in opposite directions.

For a chain of your own, I would take one management principle from this story. Next to the sales report for a promotion, you need a calculation of what is left from those sales. After that you can discuss a specific product, the supplier terms and the decision on the next launch.

Decide what result you want from the promotion

The phrase "we need to boost sales" leaves too many options open. One manager expects new customers. Another wants to clear leftover inventory. A third hopes to raise the profit of the category. After launch they look at the same report and judge it differently. It is better to agree on this before you change the price tags.

If the goal is extra profit, the main measure is the difference between the result with the promotion and the expected result without it. If you are clearing inventory, also look at the number of units cleared, the inventory left and the cash raised. Selling off an old collection that has not moved can make sense at a lower margin if the alternative is worse: more storage, a deeper markdown later, or a write-off.

If the goal is to attract customers, you need one more question. What did these customers do after buying at a discount? Did they come back, what did they buy at the regular price, what result did they leave for the chain? Until you have this data, the cost of attracting customers remains an investment with an expected return. Calling it profit already earned is too early.

On the promotion sheet, write one main goal, the measure, the observation period and the person responsible. Next to them, list the limits: which products take part, in which stores, how much inventory is available and what spending you are ready to accept. This lets you stop the discussion of "but a lot of people came" and return to the agreed goal.

The calculation below covers the extra result of the promotion. I keep gross profit, store operating profit and the result after head-office costs separate. What each level includes is explained in the article on store profitability. What matters here is the change: which income and costs appeared because of the chosen decision.

Baseline: what to compare promotion sales with

The hardest number in the calculation is often in the "without promotion" column. You see actual revenue at the checkout. But how much you would have sold at the regular price on the same days cannot be seen directly. You have to estimate this baseline and save it before the result is known.

Start with regular sales of the same product in the same stores on comparable days. Check the calendar, the season, the price, product availability and past promotions. A week when the product ran out in the middle of the day does not work as a normal baseline. Otherwise restored availability will look on paper like an outstanding result of the discount.

Comparison with last year also needs checking. Assortment, prices, neighboring tenants, opening hours and customer traffic may have changed. The previous year helps you see the context, but it does not remove all the changes from the calculation by itself. It is better to have several comparable observations and to understand the reasons for large deviations.

If the promotion runs in only part of the chain, keep similar stores without it for comparison. Look at how both groups changed before the launch: are the trend, the format and product availability close? During the promotion, check whether only one group received extra advertising or a well-timed delivery. Otherwise it is easy to credit the whole difference to the discount.

When there is no suitable group, you can calculate the result against a chosen baseline and test several versions of it. For example, a lower and a higher regular volume. If a small change in the baseline turns a plus into a minus, it is too early to roll the promotion out to the whole chain. One clear line is enough in the report: this is an estimate relative to a baseline, not a proven net effect of the promotion.

How many units you must sell after a price cut

First calculate what is left per unit at the regular price and at the promotion price. For a simple calculation, take the selling price minus the purchase cost. Revenue and costs must be on the same tax basis, for example both net of VAT or sales tax.

Let Q be the regular volume, P the regular price, C the purchase cost of one unit and A the promotion price. Then gross profit without the promotion is Q × (P - C). To keep the same amount after the price cut, you need to sell Q × (P - C) / (A - C) units. This does not yet include separate advertising, supplier support or changes in the sales of related products.

If every sale carries variable costs, subtract them from what is left per unit in both versions, regular and promotion. Use the costs that apply to each version. Then compare the totals after these costs, keeping C as the purchase cost. If the store pays for delivery in both cases, include this cost in the regular version too, not only in the promotion version.

Add launch costs that do not depend on the number of units sold to the amount in the numerator. For items sold by the piece, round a fractional threshold up: you cannot sell part of a pack to satisfy the formula. Keep both thresholds side by side: how many units you need for the same gross profit, and how many to also cover the launch costs.

Compare the resulting number with past demand, the inventory on the shelf and the capacity of the store. The required sales threshold is not a demand forecast. If a small price cut requires a volume that the store has not sold even at its peak, review the promotion terms before launch.

I would ask the commercial director to show where the required volume will come from: more customers or more units in each purchase. Then I would check whether the delivery plan supports it. If the answer comes down to "with a discount they will buy it", there is little basis to approve the budget. You can reduce the discount, choose another product or run a limited test.

With zero or negative profit per unit, a larger quantity will not restore the earlier profit in this model. You then need a different economic reason: confirmed supplier support, an extra result from the basket, or a deliberate decision to clear inventory. Calculate each of these reasons separately. Do not put them into the formula as the promise that "the customer will pick up something else."

Do not confuse a discount as a percentage of the price with the loss of a share of gross profit. The two percentages have different bases. For products with different purchase costs, the same discount leaves a different amount of money. So check a single discount percentage for a whole category against the products in it first.

Worked example: sales grew, the result got worse

Take a fully hypothetical example in currency units. These are not H&M figures and not a story of my work with a client. One product usually sells at a price of 100, with a purchase cost of 70. For a comparable period without a promotion, we expect to sell 100 units. For now we leave other sales and costs unchanged.

In the promotion we cut the price to 90, that is by 10%, and sell 140 units. Extra materials and advertising cost 200. There is no supplier support. For simplicity, all sales are final, and the example has no returns or other variable costs.

Measure Without promotion With promotion
Price per unit 100 90
Purchase cost per unit 70 70
Units sold 100 140
Revenue 10,000 12,600
Cost of goods sold 7,000 9,800
Gross profit 3,000 2,800
Extra promotion costs 0 200
Gross profit minus promotion costs 3,000 2,600

Sales in units grew by 40% and revenue by 26%. But gross profit fell by 200, and after the extra costs the shortfall against the baseline was 400. If all other income and costs stay the same, the promotion made the result worse by this amount.

Here is why. A regular unit left 30, and a promotion unit left 20. All buyers of the product got the discount, including those who would have bought the expected 100 units without it. The extra 40 sales did not make up for the loss on the original volume and the launch costs.

To keep a gross profit of 3,000 with 20 left per unit, you need to sell 150 units. To also cover the 200 of costs, you need 160. That is volume growth of 60% over the accepted baseline. Selling 140 units was above the regular volume, but below the volume needed for this goal.

The 60% growth threshold applies only to this example with a 10% discount. It depends on the starting price, the cost, the baseline and the costs. Replace these numbers with your own and calculate before launch. This is more useful than comparing the discount percentage with the percentage growth of revenue.

What else to include in the chain-wide result

The next step is to extend the calculation from one product to the category and related purchases. A promoted shampoo could have attracted a new purchase. It could also have replaced another shampoo that earns the store more. In the second case, part of the growth of the promoted item is paid for by a drop in the result of the other shampoo. You cannot add both figures as independent successes.

Look at the basket. If the promotion produced extra sales of related products, compare their result with what you expected without the promotion. It is wrong to credit the promotion with the full value of all the baskets of the promoted product's buyers. Some of these products people would have bought anyway.

Check the period after the end separately. A customer may have stocked up on the usual product in advance and temporarily stopped coming for it. Then a high peak is followed by a dip, and the result over a longer period is more modest. Choose the observation period by the usual purchase cycle of the category. The length of the advertising banner should not set it.

Count supplier support by confirmed terms. If the supplier compensates part of the discount, check the products, the volume, the documents and the moment the right to compensation arises. Do not add the money twice if the supplier discount has already lowered the cost you use. Keep the promised bonus and the received confirmation on separate lines until they are reconciled.

Add costs for advertising, printing, extra shifts, delivery and handling of goods, if they appeared because of the promotion. Regular rent does not change with the decision. It does not become a new cost just because it was allocated to the promotion. Keep calculating the full profit of the store separately. That way you see both the change from the promotion and the overall result of the store.

Reconcile inventory and write-offs. The promotion could have sold some goods faster but left surplus stock from an oversized order after it ended. To clear the warehouse, compare the options: sell now, keep, return to the supplier, transfer to another store, or write off. The decision on assortment and transfers is related to assortment management in a chain (in Russian). The promotion calculation should include only the specific changes from the chosen option.

How to check a promotion across several stores

For each store, collect the same fields: baseline and actual volume, price, cost, availability, extra costs and confirmed supplier support. Then add the category result and the observation after the end. If one store's data include returns and another's do not, the total will not be comparable.

Record the conditions of execution too. Was the product on the shelf, was the right price tag in place, did the advertising start on time, was there enough inventory until the end? If a store did not receive a delivery, deal with preparation first. Checking interest in the promotion by looking at an empty shelf tells you nothing.

First look at individual stores and groups of similar stores. Then add up the money result of the chain. An average growth percentage can hide that most of the extra profit came from one store while the others lost money. Such a result calls for a different decision than a promotion that worked equally well in all formats.

For the next test, change one clear condition at a time: the product, the depth of the discount, the duration or the set of stores. If you change everything at once, you get a new launch without a clear answer on what was worth keeping. In a small chain, a short protocol is especially useful: what decision was tested, what happened and why the next step was chosen.

When to repeat a promotion, change it or stop

The final report should end with an action. If everyone discusses revenue growth and the decision to repeat is made out of habit, the calculation stays a reference note. Below is a map you can start the discussion with. The conditions in it refer to your accepted baseline and checked costs.

What you got Decision to consider What to check before the next launch
Extra result is positive, execution is stable Repeat on comparable terms Whether inventory is enough and the supplier terms still apply
Volume grew, result is below the baseline Change the discount depth or the product, or stop The required sales threshold and the loss on regular purchases
The promoted product grew, the category fell Rebuild the set of offers Customers switching between products
A peak was followed by a drop after the end Extend the evaluation period The result over a full purchase cycle
The result depends on an unconfirmed bonus Postpone the conclusion on profit The right to compensation and no double counting
Some stores had no product Fix the preparation and repeat a limited test Delivery, shelf and actual execution

A positive calculation does not mean that you can expand any next promotion to the whole chain. Check whether purchase terms and costs change at a larger volume. Set a clear stopping criterion too: who makes the decision, on what data and when.

If the goal was to clear slow-moving inventory, decide by comparing the available options. But keep the name of the result. Cash recovered from inventory, gross profit and extra profit answer different questions. A good report lets you see each of them without renaming one as another.

What to ask your team to do before the next launch

I would start with the last completed promotion that has decent data. The commercial director restores the terms and the agreements with the supplier. The finance lead checks the cost, the expenses and the threshold calculation. The operations director collects the actual execution by store. Everyone fills in one sheet with a shared baseline and result.

If the figures differ, first find the reason: different periods, returns, the way the supplier discount is recorded, or a missing cost. The finance lead reconciles the calculation, and the manager in charge decides on a repeat within the agreed budget. This way the team has one answer to the question of how much the promotion added and what to do next.

Before the next launch, agree on the goal, the baseline volume, the required sales threshold, the budget, the supplier terms and the inventory. Set a date for the first check of execution and a date for the final money result. If customers buy the product at long intervals, add a follow-up check. Then the report for the last days of the promotion will be interim, not final.

After the end, save the calculation together with the decision: repeat, change or stop. Next season such a sheet will be more useful than the memory "last time it sold well." You will know at what price, cost and supplier terms the promotion actually helped the chain.

Frequently asked questions

How do you calculate the result of a discount promotion?

Compare the sales result with the promotion against a well-founded baseline without it. Count the cost of goods, the extra costs, the confirmed supplier support, and the changes in the sales of related products and in the period after. State exactly what you got: the change in gross profit, or the result after the costs listed. Total promotion revenue alone does not give this answer.

What discount can you give without losing profit?

There is no universal percentage. Calculate what is left per unit after the price cut and what volume you need to make up for it. Then add the launch costs and check whether the demand is realistic. Two products with the same retail price can have different costs, so the same discount will have different consequences.

Why not judge a promotion by sales growth alone?

The amount left from each unit goes down. The extra sales may not cover this loss and the costs. Buyers can also switch from another product in your chain or stock up in advance. Check the category, the basket and the later period together with the promoted item.

Do you need a control group of stores?

A comparable group helps to separate the general trend from the changes in the stores where the promotion runs. Check how close the stores were before the start and what differed during the test. If there is no suitable group, use several reasonable versions of the baseline and call the result an estimate relative to them. A single before-and-after chart does not replace such a comparison.

How do you account for supplier money?

Use confirmed terms and the right to compensation. Check the participating product, the sales and the documents. If the support has already lowered the cost, you cannot add it a second time. Also reflect the time gap between accrual and payment in your cash plan: profit and the cash available in the account can differ.

Can a promotion that loses money by the calculation still be justified?

Yes, if the owner deliberately accepts this cost for another measurable goal. For example, selling inventory that risks losing more value. Compare the alternatives and limit the spending in advance. Call such a result "meeting the inventory goal" or "customer acquisition cost", and keep a separate estimate of profit.

When should you calculate the final result?

After returns, costs and supplier terms are reconciled, with a post-promotion period long enough for the category. The timing depends on the purchase cycle and the goal of the launch. During the promotion, monitor execution and budget spending. After it, assess the full result. Do not wait for the final report to check for missing product.