Promotions, Coupons and Trade Funding

Introduction

Temporary price reductions and promotional incentives are central to retail revenue management. They can accelerate sales, attract shoppers, encourage trial, clear inventory, increase basket size, and strengthen retailer–supplier relationships. Yet a promotion that produces a large sales uplift may still destroy value if the uplift consists mainly of purchases shifted forward in time, purchases diverted from higher-margin products, or purchases funded by the retailer rather than the supplier.

The central managerial question is therefore not “How much did sales increase during the promotion?” but “How much additional profitable demand did the promotion create after accounting for baseline demand, cannibalisation, stockpiling, post-promotion dips, execution costs and funding?” This chapter develops a framework for answering that question.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Distinguish temporary price reductions, multibuy offers, coupons, loyalty rewards and trade-funded promotions.
2. Estimate promotional incrementality and evaluate cannibalisation, stockpiling and post-promotion effects.
3. Calculate promotional profit, break-even volume and supplier-funded contribution.
4. Design suitable tests, metrics and governance controls for promotions.
5. Evaluate promotions from retailer, supplier and consumer perspectives.
6. Recommend promotion strategies using customer, category and channel evidence.

1. The Economics of a Promotion

A promotion changes the normal purchase environment for a limited period. It may alter price, quantity, visibility, assortment, reward eligibility or the perceived value of an offer. A promotion can affect:

• Purchase incidence: whether a shopper buys the category or brand at all.
• Purchase quantity: how many units are purchased.
• Purchase timing: whether a planned future purchase is brought forward.
• Brand choice: whether the shopper switches from a competing brand.
• Store choice: whether the shopper visits the promoting retailer.
• Basket composition: whether the promoted item changes purchases of complementary or substitute items.

A useful decomposition of promotional sales is:

where Q_p is observed promotional-period quantity, Q_b is expected baseline quantity without the promotion, I is incremental quantity created by the promotion, C is quantity cannibalised from other products or pack sizes, S is quantity pulled forward through stockpiling, and P is the post-promotion reduction caused by forward buying or delayed replenishment. In practice, these components are estimated rather than directly observed, and their definitions must be kept consistent. If cannibalised units are already included in total category quantity, they should not be subtracted twice.

Baseline demand and incrementality

The baseline is the quantity that would probably have been sold without the promotion. It can be estimated from comparable stores, non-promoted periods, control products, or a demand model. A simple percentage uplift is:

where U is observed unit uplift. This measure is descriptive, not necessarily causal. If the promotion occurred during a holiday, media campaign or unusually strong weather period, the uplift may be incorrectly attributed to the offer.

Incremental quantity is more valuable than observed promotional quantity. Incremental sales should ideally be measured against a credible counterfactual: what would have happened in the same stores, to similar customers, during the same period, without the treatment.

Elasticity and promotional response

Price elasticity of demand describes the percentage response in quantity associated with a percentage change in price:

where ε is price elasticity, %∆Q is percentage change in quantity, and %∆P is percentage change in price. Demand is commonly called elastic when the absolute value of elasticity exceeds 1. Elasticity is not constant across customers, brands, price ranges or promotional mechanisms. A coupon may affect only redeeming shoppers, whereas a shelf price reduction is visible to almost everyone.

Elasticity alone does not establish profitability. A highly elastic product may generate volume but still lose contribution if the discount exceeds the contribution earned on incremental units.

2. Major Promotion Mechanisms

Temporary price reductions

A temporary price reduction lowers the shelf price for a defined period. It is easy for shoppers to understand and often produces high participation. Its disadvantages include margin dilution, price expectation effects, retailer funding exposure and possible price wars. The offer should specify the normal price, promotional price, duration, stock availability and funding source.

Multibuy offers

Multibuy offers include “2 for £5,” “buy one, get one free” and “buy three, save 20%.” They can increase units per transaction and encourage stockpiling. They are most appropriate when the product is storable, consumption is repeatable and the quantity threshold is credible. They are less appropriate for perishable products or products with low purchase frequency.

For a two-unit offer priced at F, the effective unit price is:

This is valid when the customer buys exactly two units and both units have the same economic value. “Buy one, get one free” is not equivalent to a 50% reduction on every transaction if some customers would have purchased only one unit and others would have purchased more than two.

Coupons

Coupons reduce the price for an eligible transaction or customer. They may be paper, digital, personalised, manufacturer-funded or retailer-funded. Coupons support targeting and measurement, but redemption may be low and coupon users may already have intended to buy. A coupon can therefore produce high redemption without high incrementality.

Important coupon measures include distribution rate, activation rate, redemption rate, incremental redemption, average basket value, cost per incremental buyer and fraud or misuse rate.

Loyalty rewards

Loyalty rewards include points, member prices, cashback, threshold rewards and personalised offers. Their objectives may include retention, share of wallet, data acquisition and frequency rather than immediate unit uplift. The economic cost of points must include expected redemption, breakage, administration and any liability under applicable accounting rules.

A reward should be evaluated over the customer relationship, not only the transaction in which it is earned. However, long-term claims require evidence; a customer who receives a reward is not necessarily retained because of it.

Supplier-funded promotions and trade funding

Suppliers may provide off-invoice allowances, bill-backs, scan-back payments, lump-sum payments, display fees, media funding or performance rebates. Trade funding changes the distribution of promotional cost but does not automatically create consumer value or incremental demand.

The retailer should distinguish:

• Front margin: invoice or purchase price relative to selling price.
• Back margin: rebates, allowances and other supplier payments.
• Net promotional margin: margin after discount, execution cost and funding conditions.

Funding agreements should define eligible units, proof of performance, timing, audit rights, payment terms, inventory ownership and treatment of unsold stock.

3. A Framework for Promotional Profitability

The fundamental analysis should be conducted at an economically consistent level, such as product, brand, category, customer or campaign. Let:

• p_n = normal retail price per unit;
• p_p = promotional retail price per unit;
• c = retailer variable product cost per unit before funding;
• f = supplier funding per eligible unit;
• Q_p = promotional-period quantity;
• Q_b = baseline quantity;
• K = fixed promotional cost, including media, labour and display;
• V = variable execution cost per unit.

Promotional contribution is:

Normal contribution is:

The incremental promotional contribution is:

These equations assume that prices, costs and funding are measured on the same tax basis and that f is actually collectible. They exclude wider category effects unless those effects are included explicitly. If the promotion changes sales of complementary or substitute products, their incremental contribution should be added to the analysis.

The break-even promotional quantity, when the promotional unit contribution is positive, is:

This expression identifies the promotional quantity required to match the normal contribution benchmark. It is valid only when the denominator is greater than zero and when the benchmark is defined consistently. If promotional unit contribution is zero or negative, additional units cannot recover fixed promotional cost under the stated assumptions.

Worked example

A retailer normally sells a product for £10 and buys it for £6. During a promotion the price is £8. The supplier provides £0.50 per promoted unit. Variable execution cost is £0.10 per unit and fixed campaign cost is £1,200. Baseline demand is 1,000 units and promotional demand is 1,700 units.

Normal contribution:

π_b = (10 - 6) × 1,000 = £4,000.

Promotional contribution:

π_p = (8 - 6 + 0.50 - 0.10) × 1,700 - 1,200

π_p = 2.40 × 1,700 - 1,200 = £2,880.

Thus, short-run incremental contribution is £2,880 - £4,000 = -£1,120. The promotion generated 700 additional observed units, but it did not match the normal contribution benchmark. If the supplier funding were £1.20 rather than £0.50, promotional unit contribution would be £3.10 and promotional contribution would be £4,070; the campaign would then exceed the benchmark by £70, before any post-promotion effects.

4. Cannibalisation, Stockpiling and Post-Promotion Effects

Cannibalisation

Cannibalisation occurs when a promoted item takes sales from another product that would otherwise have contributed profit. It may occur across:

• sizes and pack formats;
• premium and economy tiers;
• brands within a category;
• physical and digital channels;
• products consumed together or as substitutes.

Category incrementality is therefore often more meaningful than brand incrementality. A brand promotion that raises brand units by 30% but raises category units by only 3% may mainly redistribute demand.

A simple cannibalisation ratio is:

where C is estimated cannibalised quantity and I_g is gross incremental quantity attributed to the promoted item. The ratio should be interpreted cautiously because the numerator and denominator depend on the baseline model. Profit-based cannibalisation is usually superior to unit-based cannibalisation because products have different margins.

Stockpiling and forward buying

Stockpiling occurs when shoppers buy more during the promotion and consume or use the product later. It can be beneficial for the shopper and supplier, but it may not represent new consumption. It also creates a post-promotion dip.

A post-promotion adjustment can be represented as:

where I_p is incremental promotional-period quantity and P_post is the estimated quantity lost after the promotion relative to baseline. The time window should reflect product durability and purchase cycle. A two-week post-promotion window may be inadequate for detergent but excessive for fresh bakery products.

Reference price and promotion dependency

Frequent promotions can lower the shopper’s reference price and train customers to delay purchase. This creates intertemporal substitution and may weaken the normal price. Retailers should track the proportion of sales sold on deal, average time between promotions, price realisation and baseline sales in non-promotional weeks.

5. Designing a Reliable Evaluation

A strong evaluation follows five stages.

Stage 1: Define the objective

The objective may be trial, penetration, stock clearance, basket growth, retention, share gain, competitor response or supplier compliance. A campaign cannot be evaluated properly if success is undefined.

Stage 2: Specify the counterfactual

Possible designs include:

• matched control stores;
• geographic holdouts;
• customer-level randomised offers;
• product-level control groups;
• pre–post models adjusted for seasonality;
• difference-in-differences models.

In a difference-in-differences design, the estimated treatment effect is:

where Y_T is the outcome for treated units and Y_C is the outcome for control units. The method assumes that, absent the promotion, treatment and control groups would have followed parallel trends. That assumption should be examined rather than taken for granted.

Stage 3: Measure the complete economic effect

At minimum, measure units, revenue, gross margin, net margin, funding received, redemption, category sales, basket effects, inventory, stock-outs and post-period sales. Include media, labour, markdown, logistics, data and payment-processing costs where material.

Stage 4: Segment the result

Results should be segmented by customer value, region, store format, channel, product, price tier and fulfilment method. An average positive result can conceal loss-making segments.

Stage 5: Decide and learn

The result should lead to a decision: scale, redesign, restrict, renegotiate funding or stop. A promotion calendar should record hypotheses and outcomes so that learning accumulates rather than being lost in campaign-by-campaign reporting.

6. A Decision Framework for Offer Design

The following sequence supports managerial decisions:

1. Diagnose the commercial problem. Is the issue awareness, trial, availability, price, frequency or excess inventory?
2. Select the mechanism. Use a coupon when targeting matters; a multibuy when quantity expansion is plausible; a loyalty reward when retention or data is central; a price reduction when broad visibility is required.
3. Estimate the economic threshold. Calculate the required uplift and maximum affordable discount.
4. Predict substitution. Identify likely cannibalised products and competitor responses.
5. Set guardrails. Define duration, eligibility, stock limits, minimum basket, redemption rules and funding conditions.
6. Pilot and test. Use a holdout or controlled rollout wherever feasible.
7. Evaluate total contribution. Include post-promotion and category effects.
8. Institutionalise learning. Update demand, customer and funding assumptions.

7. Illustrative Case: FreshWay’s Coffee Promotion

This is a fictional illustrative case.

Background

FreshWay, a supermarket chain, promoted a 500-gram coffee brand from £8 to £6.50 for two weeks. The supplier offered £0.75 per promoted unit and paid for half of the in-store display. Management expected the promotion to recruit shoppers and increase total coffee-category sales.

Problem

Promotional brand sales rose from an expected 12,000 units to 18,000 units. However, premium coffee sales declined by 2,500 units, the category experienced stock-outs in 8% of stores, and sales of the promoted brand fell below baseline for the following three weeks.

Analysis

The apparent uplift was 6,000 units. Store-level controls estimated that 1,800 units were attributable to general seasonal demand, leaving 4,200 gross incremental units. Analysis of category sales suggested that 2,100 units were diverted from other coffee products. The post-period decline was estimated at 1,000 units.

Therefore, estimated net category incrementality was approximately:

4,200 - 2,100 - 1,000 = 1,100 units.

The result was positive at category level but much smaller than the brand-level uplift. Stock-outs also limited the observed effect and weakened the reliability of simple store averages.

Approach and solution

FreshWay redesigned the next promotion as a targeted loyalty offer. Premium regular buyers received a smaller discount, while lapsed category buyers received a stronger first-purchase coupon. The retailer reduced the duration, improved replenishment rules and required the supplier to fund a portion of incremental media and unsold promotional inventory risk. A matched holdout group was retained.

Outcome

The redesigned campaign produced a lower brand uplift but a higher category contribution per promotional pound. Premium-brand cannibalisation fell, redemption became more concentrated among lapsed buyers, and the post-promotion dip was reduced. The key managerial conclusion was that maximising promoted-brand units had been the wrong objective.

Learning points

1. Brand uplift and category incrementality are different outcomes.
2. Stock-outs create both lost sales and biased evaluation.
3. Targeted incentives can be more efficient than broad discounts.
4. Supplier funding should be linked to measurable performance.
5. Post-promotion demand must be included in the decision.

8. Governance, Ethics and Regulation

Promotions create governance risks as well as commercial opportunities. Retailers should ensure that advertised prices are genuine, conditions are prominent, stock is reasonably available, and claims such as “was/now” or “best value” are substantiated under applicable consumer-protection law. Rules differ across jurisdictions, so legal review is essential.

Personalised coupons require responsible use of customer data. Organisations should establish a lawful basis for processing, purpose limitation, security, retention controls, transparency and appropriate safeguards against discriminatory outcomes. Vulnerable customers should not be exploited through excessive urgency, opaque pricing or offers that encourage harmful overconsumption.

Supplier funding can create conflicts of interest. Category managers may favour funded products even when the offer harms shoppers or category profitability. Governance should separate commercial negotiation from measurement, document assumptions, audit claims and prohibit retroactive reclassification of ordinary sales as incremental promotional sales.

9. Technology, AI and Contemporary Practice

Retailers increasingly use electronic shelf labels, real-time inventory, loyalty data, digital wallets and retail media to deliver and measure promotions. Machine-learning models can estimate propensities to redeem, predict stockpiling and identify treatment heterogeneity. However, prediction is not the same as causal inference. A model that identifies likely coupon users may simply target people who would have purchased anyway.

AI-supported promotion systems should therefore combine predictive models with experimental controls, monitoring for drift, explainable decision rules and human approval of material price changes. Important controls include minimum and maximum price boundaries, inventory feasibility, supplier-funding validation, anomaly detection and post-campaign review.

Challenges and Limitations

Promotional evaluation is difficult because demand is affected by seasonality, weather, competitor prices, media, assortment, availability and macroeconomic conditions. Baselines are estimates, not observed facts. Customer-level randomisation may be constrained by fairness, operational complexity or contamination between treatment groups. Supplier funding may be uncertain or delayed. Long-term brand effects are difficult to identify within short reporting cycles.

Managers should report confidence ranges or sensitivity scenarios where assumptions are uncertain. For example, if estimated cannibalisation ranges from 30% to 50% of gross uplift, a campaign should not be labelled successful solely on the point estimate. Robust decisions recognise uncertainty and protect against optimistic attribution.

Best-Practice Checklist

• Define the commercial objective before selecting the offer.
• Measure incremental contribution, not only sales uplift.
• Use a credible control or counterfactual.
• Evaluate brand, category and total-basket effects.
• Model stockpiling and post-promotion demand.
• Include stock-outs and execution costs.
• Treat supplier funding as conditional economic value, not guaranteed profit.
• Set clear legal, data and pricing controls.
• Segment results by customer, store, product and channel.
• Record hypotheses, results and lessons in a promotion knowledge base.

Chapter Summary

Promotions are temporary interventions whose value depends on incremental profitable demand rather than headline sales. Temporary price reductions, multibuys, coupons, loyalty rewards and supplier-funded offers have different targeting, behavioural and financial consequences. Evaluation requires a credible baseline and attention to incrementality, cannibalisation, stockpiling, post-promotion dips, execution cost and funding reliability. Contribution analysis provides a common financial language, while controlled experiments and difference-in-differences designs improve causal interpretation. Effective promotion management aligns the offer mechanism with the commercial objective, protects category economics, governs supplier relationships and uses customer data responsibly.

Key Terms

Baseline demand: Estimated demand that would occur without the promotion.

Cannibalisation: Loss of sales or contribution from related products caused by the promoted offer.

Coupon: A voucher or digital entitlement that reduces an eligible transaction price.

Difference-in-differences: A method comparing changes over time in treatment and control groups.

Incrementality: Demand or contribution caused by the promotion beyond the counterfactual.

Loyalty reward: An incentive linked to membership, repeat behaviour or customer identification.

Multibuy: An offer that changes price or value according to purchased quantity.

Net promotional contribution: Promotional contribution after discounts, funding and relevant costs.

Post-promotion dip: Sales reduction after an offer, often caused by forward buying.

Price elasticity: Percentage change in quantity divided by percentage change in price.

Promotional funding: Financial support provided by a supplier for a retail activity.

Reference price: Price level used by a shopper to judge whether an offer is attractive.

Redemption rate: Proportion of distributed or activated coupons that are used.

Stockpiling: Buying additional quantity during a promotion for later consumption.

Supplier allowance: Payment or discount provided by a supplier under a commercial agreement.

Trade funding: Financial support supplied to retailers for distribution, display, price or promotional activity.

Treatment group: Stores, customers or products exposed to the promotion.

Control group: Comparable units not exposed to the promotion.

Review Questions

Short Answer Questions

1. Why is promotional-period sales uplift not equivalent to incrementality?
2. Define cannibalisation and give one retail example.
3. What is the purpose of a control group in promotion evaluation?
4. Explain how stockpiling can create a post-promotion dip.
5. Identify two risks associated with supplier-funded promotions.

Long Answer Questions

1. Compare temporary price reductions, multibuy offers, coupons and loyalty rewards in terms of objectives, benefits and risks.
2. Explain how a retailer should calculate promotional contribution and break-even quantity.
3. Discuss why category-level analysis may be more useful than brand-level analysis.
4. Evaluate the assumptions and limitations of a difference-in-differences promotion study.
5. Discuss the ethical and governance issues involved in personalised promotional pricing.

Analytical / Application Questions

1. A product normally sells for £12 with a unit cost of £7. It is promoted at £9, receives £0.40 supplier funding per unit, and incurs £0.10 variable execution cost. Baseline demand is 2,000 units, promotional demand is 3,000 units, and fixed campaign cost is £800. Calculate normal contribution, promotional contribution and incremental contribution. State whether the campaign beats the normal contribution benchmark.
2. A promotion creates 5,000 gross incremental brand units. Evidence suggests that 35% are cannibalised from other category products and 900 units are pulled forward from the following month. Estimate net category incrementality and explain two limitations of the estimate.
3. Design a controlled test for a targeted loyalty coupon. Specify the treatment, control, primary outcome, financial measures, possible sources of bias and the decision rule for scaling the campaign.

Multiple Choice Questions with Answers

Question 1

Which measure most directly captures demand caused by a promotion beyond the counterfactual?

A. Redemption rate

B. Incrementality

C. Display compliance

D. Distribution breadth

Correct Answer: B. Incrementality

Question 2

A product normally sells for £10, has a unit cost of £6, and is promoted at £8 with £0.50 supplier funding and £0.10 variable execution cost. What is promotional unit contribution?

A. £1.40

B. £2.00

C. £2.40

D. £2.50

Correct Answer: C. £2.40

Question 3

Which situation is the clearest example of cannibalisation?

A. A promotion attracts a new category buyer

B. A shopper buys an additional complementary product

C. A promoted economy pack replaces a premium pack the shopper would have bought

D. A supplier pays for an end-cap display

Correct Answer: C. A promoted economy pack replaces a premium pack the shopper would have bought

Question 4

What is the principal purpose of a control group in promotion evaluation?

A. To increase coupon redemption

B. To estimate what would have happened without the promotion

C. To guarantee supplier funding

D. To eliminate all stock-outs

Correct Answer: B. To estimate what would have happened without the promotion

Question 5

Which promotion is generally most suitable when targeted customer-level incentives are important?

A. Personalised digital coupon

B. Unrestricted national price cut

C. Unadvertised shelf relocation

D. Uniform product delisting

Correct Answer: A. Personalised digital coupon

Question 6

A promotion sells 18,000 units against a baseline of 12,000 units. What is the observed unit uplift percentage?

A. 33.3%

B. 40.0%

C. 50.0%

D. 66.7%

Correct Answer: C. 50.0%

Question 7

Which effect occurs when shoppers buy during a deal and reduce purchases immediately afterward?

A. Reference-price formation

B. Post-promotion dip

C. Trade allowance

D. Display compliance

Correct Answer: B. Post-promotion dip

Question 8

Which assumption is required for a basic difference-in-differences interpretation?

A. Promotional demand must be zero

B. Treatment and control groups must have parallel pre-treatment trends

C. All customers must redeem a coupon

D. Supplier funding must equal the retail discount

Correct Answer: B. Treatment and control groups must have parallel pre-treatment trends

Question 9

If promotional unit contribution is negative, what follows under the simple break-even model?

A. More units automatically recover fixed cost

B. The promotion is profitable if redemption is high

C. Additional units cannot recover fixed cost through unit contribution

D. Cannibalisation becomes irrelevant

Correct Answer: C. Additional units cannot recover fixed cost through unit contribution

Question 10

Why should supplier funding be treated cautiously in profitability analysis?

A. It is always illegal

B. It may be conditional, delayed or dependent on proof of performance

C. It never affects retailer margin

D. It eliminates the need for a baseline

Correct Answer: B. It may be conditional, delayed or dependent on proof of performance