Cannibalisation: measuring what the new offer genuinely displaces.
Key distinctions
Three conditions before calculating.
Origin of the volumes
Separate new customers, increased frequency, reactivation and transfer from each existing offer. A single average proportion hides different displaced margins.
Margin, not revenue
A transfer from a high-margin offer can destroy contribution even if the price of the new item is higher. The calculation must use the variable costs specific to each product.
Migration horizon
A launch can first attract the closest customers before widening demand. Measuring too early sometimes overestimates durable cannibalisation; measuring too late can hide an established erosion.
Method
Estimating cannibalisation without double counting
- 01Build the origins
Attribute each sale to a mutually exclusive origin — new demand, transfer A, transfer B, purchase brought forward or competitive volume won back.
- 02Value the transfers
Multiply the units displaced by the contribution that would have been generated without the launch, not by the list price of the old offer.
- 03Add the specific costs
Deduct listing, production, support, launch discounts and assortment complexity over the decision horizon.
- 04Test the tipping threshold
Compute the maximum transfer rate compatible with a positive net contribution and compare that threshold with a plausible range.
The contribution cascade
The revenue of the new item is broken down to the contribution that is genuinely additional. Transfers are valued with the margin of the offers they leave.
net contribution = new margin − displaced margins − specific costs12,000 sales that create only €26,600
The new offer sells 12,000 units with €18 of unit contribution. 35% come from offer A at €22 of contribution, 15% from offer B at €15, and the launch costs €70,000.
Gross contribution: 12,000 × €18 = €216,000. Displaced margin: 4,200 × €22 + 1,800 × €15 = €119,400. After €70,000 of launch, the net contribution is €26,600.
The launch stays positive but fragile. A slide of eight additional points from offer A would almost cancel the gain; the rollout must therefore stay progressive.
Acceptance conditions
What must be true to act.
- 01
The origins of volume are exclusive and total 100%.
- 02
Every transfer is valued with the contribution of the offer given up.
- 03
The maximum acceptable cannibalisation rate is published.
- 04
The test also measures new demand and retention, not only the product mix.
Limits
What this analysis does not prove.
- Intention surveys identify imperfectly what would have been bought without the new item.
- Stockouts, promotions and distribution changes can artificially alter the transfers.
- Long-term portfolio value can justify a deliberate cannibalisation that the short-term calculation penalises.
References
Works cited.
- Copulsky, W. (1976), “Cannibalism in the Marketplace”, Journal of Marketing.
- Srinivasan, S. R., Ramakrishnan, S. & Grasman, S. E. (2005), Marketing Intelligence & Planning.
- Moorthy, K. S. & Png, I. P. L. (1992), Marketing Science.