Promotions: Measuring Incremental Sales, not the volume displaced
The volume sold during a promotion is a mix of new sales and displaced sales: purchases brought forward, customers coming from another store, and sales taken from comparable products. Black Week 2025 demonstrated this on a national scale, with volumes down 3% while sales value remained virtually unchanged.
An honest assessment is based on the entire category, over a period extending beyond the end of the promotion, and compared to a group of control stores whenever possible. The result is a range: its value comes from the ranking of the promotions, all of which are measured in the same way.
The incremental impact of a promotion is the portion of sales that would not have occurred without it: neither on that date, nor in that store, nor for that product. Everything else is shifted volume.
This article discusses post-campaign measurement: what the evaluation should focus on, over what time frame, and against what benchmark. Pre-launch cost estimates are covered in a separate article on promotional simulation.
Incremental Sales: What a Promotion's Volume Doesn't Tell You
In many committees, a promotional campaign is still judged by the volume sold during the period. The higher the sales, the more successful the campaign is deemed to be.
However, this figure combines two things: sales generated by the promotion, and sales that would have occurred anyway—earlier, later, at another store, or for a similar product.
```
A customer who would not have purchased this product at that time without the offer:new net revenue.
```
A purchase that would have taken place anyway, but was simply brought forward or postponed.
Only the first category justifies the cost of the discount. The second category was paid for twice: once by the customer and once through the margin.
of consumer goods sales came from promotions in 2025, a 20-year record. 45% of households now say they try to buy items on sale.
(NielsenIQ, Consumer Goods Market Outlook 2025, February 2, 2026)
At this level, a few overestimated incremental points, repeated every week across an entire network, add up to a margin.
Leaks that inflate volume without adding value
Non-incremental volume consists of three mechanisms, which should be considered separately.
| Leak | Mechanism | How to limit it |
|---|---|---|
| Deferred Purchase | Regular customers make their purchases or stock up when a sale is on, and then don't buy anything for several weeks. | Track sales over a broader time frame after the promotion, not just during it. |
| Store Report | A customer switches to a different store within the same chain to take advantage of the offer, without making any additional net purchases. | Compare sales at the store running the promotion with those at neighboring stores of the same chain. |
| Intra-category cannibalization | The promoted item captures sales that would otherwise have gone to a similar variant or format in the same aisle. | Measure sales trends for the entire category, not just the promoted item. |
A fourth type of customer defection works in your favor: the customer poached from a competitor. This is the only type that creates value for you without creating value for the market.
The statistical mechanism behind the third leakage is detailed in our article on cross-elasticity, cannibalization, and the halo effect.
Black Friday 2025: A Nationwide Delay in Purchases
Deferred purchasing is not just a theoretical concept. Black Week 2025 demonstrated this across the entire French market.
in volume purchased during Black Week 2025 in France, representing a 0.4% increase in value. 62% of the week’s purchases were Christmas gifts.
(Joko, anonymized banking transactions from one million French people, reported by Ecommerce Nation on 12/22/2025)
Reading: fewer items, slightly more expensive, purchased earlier. A large portion of this demand was reportedly met in December, often at full price.
A Black Week that “saw a 30% increase” proves nothing until December’s results are in.
Our position: A promotional campaign's results should be evaluated after six weeks, not on the very last day
Most promotional dashboards stop on the last day of the campaign. This is the most costly bias we encounter, because it is not neutral.
He always promotes deep discounts on non-perishable items: these are the ones that cause the biggest dip in sales in the following weeks, and that dip never shows up on the balance sheet.
A product sells 1,000 units per week. During a promotional week, it sells 2,000: +1,000 on the dashboard.
Over the next three weeks, its sales fell below normal levels: a total of 350 fewer units. Its two substitutes on the shelf lost 250 sales during the promotion.
Actual increment: 1,000 - 350 - 250 = 400 units. The other 600 already existed and were sold at a discount.
Our rule: Read the entire category, focusing on a timeframe that covers at least the weeks it takes to return to normal. For fresh produce, one week is often enough; for laundry detergent or beverages, it takes several weeks.
The same rule applies to days with sales: they must be excluded from any calculation, because a sale distorts the measurement of a promotion by underreporting its actual impact.
Showrooms: The simplest approach is often the most honest
To measure incremental impact, there’s no need to wait for a research firm. A group of stores that does not receive the offer—and is comparable in size, customer base, and competition—provides a solid benchmark.
Our belief: A well-chosen case study is better than a sophisticated model that no one can explain to a committee. The model takes over when a case study isn't feasible—for example, for a national brochure.
A Southeast Asian food retail group has made comparisons with control stores its standard for evaluation.
Each price change is measured against a control group.
The new prices were first rolled out in a few pilot stores and compared to control stores. The results are reviewed weekly and supplemented by an in-store price perception survey.
- Issue
- Move away from a robust but manual method using Excel across multiple retail chains.
- What Has Changed
- Pilot program with a control group, benchmark indices, and capped price fluctuations.
customers surveyed in-store to gauge their perception of the pilot's price.
Pricing Department · Food Retail Group. Case study written by Booper based on an actual client project; the client has been anonymized.
A common pitfall: a benchmark that is too close causes customers to switch to the alternative offering, thereby widening the gap.
If there are no sales figures, the baseline must be calculated by reference, by store, and by week, excluding past promotions. Otherwise, it will be based on sales that are already inflated.
The True ROI of a Promotion: A Range and a Ranking
The cost of a promotion isn't just the discount on sales generated. It also includes the discount given to everyone who would have made a purchase anyway, and that is often the biggest expense.
ROI is therefore calculated based on the category's incremental margin over the entire time frame, minus that cost, plus any contribution from the supplier.
A useful counterintuitive insight: trying to find an exact number is a dead end. An incremental measure is a range, and it will remain so.
Value comes from evaluating all operations in the same way, ranking them, and correcting those at the bottom of the ranking.
To estimate costs for scenarios in advance rather than after the fact, see how to simulate a promotion before launching it.
Booper calculates a baseline sales figure by SKU and by store, adjusted for past promotions and out-of-stock periods. This figure serves as the benchmark for the financial statement.
Before the operation, the simulation quantifies the expected impact on sales, margin, and inventory. Afterward, the actual results are compared to this scenario across the entire category. See the Booper modular pricing platform.
See also in this feature
- Promotional Pricing: 7 Strategies to Stop Eroding Your Margins
- Pricing, promotions, and markdowns: Bringing them all together in a single retail tool
- Black Friday: Real Discounts, No Fake Strikethrough Prices Since 2022
- Back-to-School Pricing: Focus on Granularity, Not the Average
FAQ
This is the portion of sales generated during a promotion that would not have occurred without it: neither on that date, nor in that store, nor for that product.
The rest of the sales volume is attributed to: advance purchases, customers coming from another store in the chain, and sales taken from similar products.
We compare actual sales to a baseline—what would have been sold without the promotion—and then subtract the drop in sales over the following weeks and the sales lost to substitutes in the same department.
The calculation is performed on the entire category and over a window that extends beyond the end of the operation.
Starting with the category's incremental margin over the entire measurement window, and subtracting the discount granted to customers who would have purchased the product anyway, as well as the costs of the transaction.
The supplier's contribution is deducted. An ROI calculated based solely on the volume for the period is almost always overestimated.
This is a comparison between a group exposed to the offer and a control group that is not: stores, a geographic area, or customers excluded from the campaign.
The difference between the two, over the same period, isolates the promotion's own effect from seasonal or cyclical variations.
There are three categories: tests with a control group, baselines modeled by reference and by store, and marketing mix models for media investments.
For in-store promotions, the first two are often sufficient. The criteria for choosing a tool are detailed in the section on what a promotion management tool must demonstrate.
At least until sales return to normal. For fresh produce, a week is often enough; for non-perishable goods, the slump can last several weeks.
A tally as of the last day of the promotion ignores this dip and systematically overestimates the deepest discounts.
Cannibalization is one of the causes of non-incrementality: sales volume taken from a neighboring SKU in the same department.
Purchase deferral and store deferral are two others. Incremental revenue is what remains after these three leaks are removed.
Sources: NielsenIQ, 2025 Consumer Goods Market Outlook, press release dated February 2, 2026 · Joko, 2025 Black Week Report, published by Ecommerce Nation on December 22, 2025 · Example with figures: illustrative case study created for this article, without customer data.
Further Reading
Management doesn't approve a price; it approves a trade-off between margin, volume, and price image. Presenting a single recommendation means asking management to trust you. Presenting the cost of each option gives management the information it needs to make a decision.
A sound projection provides a range rather than a single figure, specifies what is being excluded, and anticipates the gap between the projected increase and the actual increase. When compared with actual results, it establishes the credibility of the next projection.
A franchisor cannot impose a minimum resale price: doing so is punishable by law; the franchisor may only recommend a price or set a maximum. A franchisee’s margin is therefore governed by a documented target, defined by network segment and applied uniformly across each segment.
The best indicator of this governance is not adherence to the suggested retail price, but the rate of compliance with it by segment. A suggested retail price that franchisees in a given area do not follow often signals a poorly set price, not a lack of discipline.
Optimix merged with Maxxing in July 2025, and PricingHUB was acquired by Akeneo in June 2026. For their customers, the question of the independence of the pricing roadmap has become a legitimate concern.
What sets a query apart in France isn't the processing speed: it's the native support for French rules, pricing by store or by zone, the validation process, and testing with your own data.