Discount Cannibalization: Price the Lost Full Sales

Discount cannibalization occurs in this planning model when a customer who would have bought at the regular price receives the promotional price instead. Those purchases still generate revenue, but the concession reduces contribution without creating an additional order. Separate them from newly induced purchases and from purchases moved between products or dates. A high count of discounted orders does not reveal how much of the order base is genuinely additional.

Divide the promoted orders by origin

Assume a hypothetical promotion generates 300 orders at $45 instead of the regular $50. Variable cost remains $30 per order. Regular contribution is $20 and promotional contribution is $15. Suppose the scenario estimates that 180 of the 300 buyers would have purchased at full price, while 120 are additional within the defined window. The 60% natural-buyer share uses all 300 promoted orders as its denominator and is an assumption requiring evidence.

Calculate the incremental result

The additional buyers contribute 120 times $15, or $1,800. The 180 natural buyers receive a $5 concession, reducing contribution by $900. A further $300 campaign expense leaves an incremental result of $600. The promoted orders themselves contribute $4,500 before campaign expense, but that entire amount is not additional: the no-promotion baseline already contains $3,600 from the 180 regular-price purchases.

Solve the maximum natural-buyer share

Let the natural-buyer fraction be c while total promoted orders stay 300. Incremental contribution after campaign expense is 300 times [(1 minus c) times $15 minus c times $5], minus $300. Setting this expression to zero gives c = 70%. At that boundary, 90 additional buyers contribute $1,350, 210 natural buyers receive $1,050 in concessions, and setup costs $300. Above 70%, the modeled promotion reduces contribution under unchanged assumptions.

Look beyond the promoted product

A shopper might switch from another product that would have contributed more, or buy this month instead of next month. Those outcomes require a broader baseline than simply no purchase of the advertised item today. Compare total customer or category contribution over a suitable horizon, including the displaced product and later purchases. Avoid counting an order both as a natural purchase and a separately displaced sale when they describe the same underlying transaction.

Estimate behavior with a defensible comparison

Use comparable customers who did not receive the offer when a practical test supports that design. Examine preexisting purchase frequency and customer mix, and retain uncertainty around the estimated natural-buyer share. Redemption records identify who used an offer, not what they would have done without it. If reliable causal evidence is unavailable, report several cannibalization scenarios instead of labeling one share as observed fact.

Use the threshold to review an offer

The 70% boundary applies only to the stated price, cost, campaign expense, and order count. Larger baskets, changed delivery costs, returns, and capacity constraints can move it. A negative computed threshold means the modeled campaign already fails even with no natural buyers; a threshold above 100% may reflect unusually favorable assumptions that need explanation. Keep the full signed result and use it to identify the assumptions worth testing before committing a larger promotion budget.

Method references: OpenStax: relevant information for decisions; Google Ads: conversion lift. Accessed October 22, 2026. All dollar figures are hypothetical worked examples.