BRANDBUSTERS SIGNAL / 05 · SALES AND MARGIN
Discounts drive revenue. But do they drive growth? How to promote without giving away margin
A promotion can increase sales, accelerate stock rotation and acquire new customers. It can also discount purchases that would have happened anyway, pull demand forward from the following week and teach customers that the regular price is not real. The difference only becomes visible when the business measures incremental effect and contribution, rather than revenue during the campaign.
- 13 MIN READ
- PROMOTIONS · MARGIN · RETAIL
- UPDATED: OCTOBER 2026
A good promotion is not a date in the calendar. It is an investment with a specific job.
In many businesses, the promotional calendar is drawn up before anyone has answered which business problem each campaign is meant to solve.
Black Friday, the start of the season, the brand’s anniversary, a members’ weekend, a competitor’s campaign or pressure to hit the plan become reason enough. The team sets the discount, prepares the campaign and watches sales rise.
None of this answers the most important question:
How much of those sales happened because of the promotion, and how much would have happened without it?
If the customer was ready to buy at the regular price, the discount did not create a sale. It merely reduced its value. If the customer brought forward a purchase planned for next week, the campaign improved one period at the expense of the next. If they chose the discounted product instead of another product from the same company, part of the result is cannibalisation, not growth.
That is why a good promotion does not start with “what percentage?”. It starts by defining the job, the audience, the economics and how the result will be measured.
The most expensive discount is the one given to a customer who was ready to pay full price.
DISCOUNT ECONOMICS · ILLUSTRATIVE EXAMPLE
Why can a 20% discount require twice the sales?
- Regular price
- PLN 100
- Variable cost
- PLN 60
- Contribution before discount
- PLN 40
- New price
- PLN 80
- Variable cost
- PLN 60
- Contribution after discount
- PLN 20
Required volume increase
+100%
In this simplified example, the business must sell twice as many units to maintain the same total contribution margin. That is before accounting for campaign costs, handling higher volume, returns, cannibalisation and sales pulled forward from another period.
This is an illustrative example. The actual result depends on the cost structure, product mix, customer behaviour and incremental effect.
Revenue answers “how much did we sell?”. It does not answer “what did the promotion create?”.
Judging a promotion by comparing sales with the previous week or year can be misleading. The result is shaped by seasonality, weather, price changes, availability, competitor activity, media campaigns, the number of stores, product mix and the general state of demand.
The decision should rest on a credible no-promotion baseline: an estimate of what would have happened had the campaign not run.
Only the difference between the actual result and that counterfactual shows the incremental effect. From it, you then deduct the cost of the discount, media, display, additional handling and the campaign’s other costs.
A promotion makes sense if it creates additional value the business could not have obtained another way — or if it deliberately serves another important purpose, such as reducing stock, acquiring the right customers or accelerating the launch of a new product.
THE BRANDBUSTERS MODEL
Seven decisions before launching a promotion
- 01
01 / JOB
Which specific business problem should the campaign solve?
- 02
02 / AUDIENCE
Does the discount reach the person whose behaviour we actually want to change?
- 03
03 / ECONOMICS
How much margin remains after the discount, costs and change in mix?
- 04
04 / MECHANIC
Which type of benefit best supports the behaviour we want?
- 05
05 / INCREMENTALITY
How will we estimate the result that would not have happened without the promotion?
- 06
06 / EXECUTION
Are stock, channels, stores, systems and service ready?
- 07
07 / AFTER-EFFECT
What will the customer do after the promotion, and will they return without another discount?
If the team cannot answer these questions before launch, it will probably not be able to account honestly for the campaign once it ends.
The seven decisions in detail
- 01
Job: a promotion cannot do everything at once.
Acquiring a new customer needs a different mechanic from increasing basket value, reactivation, reducing stock, launching a product or increasing purchase frequency.
General goals such as “increase sales” are not enough. The business must decide whether volume, margin, new customers, stock rotation, product trial or behaviour after the first purchase matters most.
Question
What single change in customer behaviour should the promotion trigger?
- 02
Audience: a broad discount often pays for behaviour that already existed.
The wider the group, the greater the risk that part of the budget goes to customers who were ready to buy without an incentive. It is therefore worth distinguishing new customers, active customers, price-sensitive buyers, those at risk of leaving, first-time buyers and high-value customers.
Not every segmentation requires complex personalisation. Sometimes it is enough to exclude the most obvious group that needs no incentive.
Question
Who should not receive this discount?
- 03
Economics: the discount percentage does not show the full cost.
The cost of a promotion includes lost margin on baseline sales, media, production of materials, paid display, additional logistics, handling returns, channel commissions and the effect of a change in mix.
It is also worth checking whether a rise in transactions has been bought with a lower basket value, sales of lower-margin products or a higher share of customers who buy only on promotion.
Question
How much incremental contribution — not additional revenue — must the campaign create to cover its full cost?
- 04
Mechanic: a percentage discount is only one of the tools.
A business can use a basket-value threshold, a product bundle, a benefit on the next purchase, an additional service, a gift, early access, points, a member benefit or a promotion built around a shared idea.
Research published in the Journal of Consumer Research shows that threshold promotions and capped promotions can be perceived differently in terms of fairness, despite a similar maximum benefit. The effect depends, among other things, on the level of the threshold.
Another study, covering more than 800 promotions in loyalty programmes, found that bonus rewards could achieve an effect comparable to a significant price discount, but combining the two mechanics weakened the result.
Question
Can the same goal be achieved without lowering the regular price?
- 05
Incrementality: growth during a promotion is not yet the effect of the promotion.
The best solution is a credible control group: a share of customers, stores, regions or time in which the campaign did not run. Where a full experiment is not possible, you can use matched stores, cohort analysis, a demand model or a carefully constructed baseline.
Comparing only with last year does not separate the effect of the promotion from other changes.
Question
What result would we consider most likely had the promotion not launched?
- 06
Execution: a good mechanic can fail on a bad detail.
An out-of-stock product, inconsistent prices across channels, poor display, unprepared staff, a code that does not work, overly complex terms or a benefit calculated incorrectly can destroy both the result and customer trust.
Before launch, you need a single checklist covering sales, marketing, e-commerce, operations, finance, customer service and legal compliance.
Question
At which point in execution might the customer feel that the promised benefit is not real?
- 07
After-effect: the best result may only show later.
A promotion that acquires customers has value only if a sufficient share of them return, buy products with healthy margins and do not need another discount. Conversely, a campaign that increases basket size may merely bring forward a purchase that would otherwise have happened later.
It is therefore worth tracking customer and category behaviour over the following weeks: repeat purchase, cohort margin, returns, the move to regular price, frequency and the drop in sales after the campaign ends.
Question
What behaviour do we expect from the customer 30, 60 and 90 days after the promotion?
Not every promotion has to cut the price.
| # | Mechanic | When it can make sense | Main risk |
|---|---|---|---|
| 01 | Direct price promotion | When it can make senseRapid stock reduction, a strong price barrier or a controlled test. | Main riskGiving margin away to customers who would have bought without a discount. |
| 02 | Threshold promotion | When it can make senseIncreasing transaction value or adding a logical product. | Main riskArtificial purchases to reach the threshold, followed by returns. |
| 03 | Bundle | When it can make senseThe products together deliver a fuller outcome for the customer. | Main riskShifting the discount onto a product that sold well on its own. |
| 04 | Benefit on the next purchase | When it can make senseBuilding a second transaction and a return visit. | Main riskA reward that is too distant or hard to redeem. |
| 05 | Loyalty programme | When it can make senseA long-term relationship, consented data and customer recognition. | Main riskComplex rules and rewarding spending for its own sake. |
| 06 | Added benefit or service | When it can make senseHigh perceived value at a lower marginal cost. | Main riskA benefit unrelated to the core need. |
| 07 | Early or exclusive access | When it can make senseCustomers value novelty, status or limited availability. | Main riskArtificial exclusivity that customers do not take seriously. |
| 08 | Symbolic promotion | When it can make senseThe brand can link the purchase to a community, place or idea. | Main riskNo authentic connection between the brand and the value it claims. |
LAW AND TRUST
A promotion must not only be profitable. It must also be true.
Under EU consumer law, when a trader announces a price reduction, the prior price shown should be the lowest price applied during at least the 30 days before the reduction. This reference-price principle applies to in-store sales, online sales and advertising.
How it is implemented — and which exceptions apply, for example to products recently introduced, progressive reductions or genuinely personalised offers — depends on the market and the type of offer.
In Poland, for example, the consumer protection authority UOKiK took action in 2026 against large retailers and platforms over the way discounts were presented. It shows that how a promotion is calculated and communicated is not a detail to be left to marketing alone.
This section is business information, not legal advice. The mechanics of any specific promotion should be reviewed by the person responsible for legal compliance in the relevant market.
Seven traps that create revenue instead of growth
- 01The promotion went into the calendar but was never given a specific business job.
- 02The entire customer base receives the same benefit, regardless of readiness to buy.
- 03Success is measured in revenue without accounting for margin, cost and cannibalisation.
- 04The business compares the result only with the previous period and calls the whole difference the effect of the promotion.
- 05Campaigns follow one another so often that the regular price loses credibility.
- 06The promotion ends with the campaign and nobody analyses the behaviour of the customers it acquired.
- 07The team reports the number of transactions but cannot say which decision will be taken on the basis of the result.
A promotion that cannot be accounted for honestly is a cost dressed up as a sales strategy.
A minimum promotion scorecard
| # | Measure | Question |
|---|---|---|
| 01 | Incremental sales | QuestionHow much additional sales happened because of the campaign? |
| 02 | Incremental contribution | QuestionWhat value remained after the discount and all costs? |
| 03 | Cost of acquisition or behaviour change | QuestionWhat did a new customer, a return visit or a larger basket cost? |
| 04 | Cannibalisation | QuestionWhat share of sales was shifted from another product, channel or period? |
| 05 | Mix change | QuestionDid customers buy products with better or worse margins? |
| 06 | New customers | QuestionHow many customers were genuinely new, and how many simply unrecognised? |
| 07 | Post-promotion behaviour | QuestionDid customers return without another discount? |
| 08 | Returns and cancellations | QuestionWas the rise in transactions reversed after the campaign ended? |
| 09 | Effect on the reference price | QuestionDo customers still accept the regular price? |
Putting promotions in order in 30 days
WEEK 1
Calculate the true cost.
- Select the three largest recent promotions.
- Reconstruct price, variable cost, discount, media and execution costs.
- Check margin, returns, mix change and post-campaign sales.
- Record the gaps in the data.
WEEK 2
Separate jobs and audiences.
- Assign each mechanic one main job.
- Define the group whose behaviour the promotion should change.
- Identify customers who do not need a discount.
- Set the minimum required economic result.
WEEK 3
Design a test.
- Choose one mechanic to improve.
- Create a control group or a credible baseline.
- Set the primary measure, guardrail measures and a stop condition.
- Check execution and how the price is communicated.
WEEK 4
Take the decision.
- Account for the incremental effect and contribution.
- Separate the campaign result from the business result.
- Decide: scale, improve or stop.
- Record the lessons in a single repository of promotion knowledge.
Ten questions before approving the next promotion
- 01Which single customer behaviour do we want to change?
- 02Why would the customer not take this action without a promotion?
- 03Who should not receive the discount?
- 04How much margin will remain after all costs?
- 05What volume increase is needed to offset the discount?
- 06What is our credible no-promotion baseline?
- 07How will we check for cannibalisation and demand pulled forward?
- 08Can the goal be achieved with a mechanic other than a price cut?
- 09What should the customer do after the campaign ends?
- 10Which decision will we take once we see the result?
Do not abandon promotions. Stop treating them as an automatic route to growth.
Promotions can be a very effective tool. They help trigger product trial, increase transaction value, accelerate rotation, win back customers and build the next purchase.
The problem starts when the size of the discount replaces the idea, rising revenue replaces the economic result and the campaign calendar replaces strategy.
Mature promotion management brings marketing, sales, finance, CRM, e-commerce and operations together around one question: what additional value did this decision create?
First decide which behaviour must change. Only then choose the benefit intended to change it.
DOES YOUR BUSINESS SELL MORE AND MORE ON PROMOTION?
Let us identify which campaigns create growth and which merely shift demand and give away margin.
We can combine an analysis of promotion economics, customer segments, mechanics and measurement with a practical plan to improve margin and incremental sales.