Sales cannibalization¶
Sales gained by one offering or outlet that displace sales of another unit in the same commercial system.
Core Idea¶
Sales cannibalization is the part of one offering's or outlet's sales that comes at the expense of another unit in the same commercial portfolio, brand, or chain. Those sales count at the focal unit, but are not new sales for the system as a whole. The effect must be attributed against what would otherwise have happened; an old unit's decline alone does not prove where its customers went.[ref-a18a89f45c89][ref-b86a118bc113]
The same relation can occur among products or locations. A detergent line extension may take purchases from its parent brand, while one fast-food chain outlet may take purchases from nearby chain outlets. The first setting was studied with consumer panel data; the second with a fitted spatial retail model.[ref-a18a89f45c89][ref-b86a118bc113]
Scope of Application¶
A common commercial system has to be named. For a product line this may be one parent brand. For retail outlets it may be a chain, including franchise arrangements; same-chain affiliation does not prove that every location has the same legal owner. The focal unit may be newly introduced or already operating. What matters is the within-system shift of purchases.[ref-a18a89f45c89][ref-b86a118bc113]
Lomax's original abstract reports parent-brand cannibalization for all three studied detergent extensions. Pancras, Sriram and Kumar estimate a cannibalized component of individual-store sales within one fast-food chain. Neither source establishes a universal rate or a universal profit outcome.[ref-a18a89f45c89][ref-b86a118bc113]
Clarity¶
A new detergent variant's gross sales can come from three sources: new demand, sales won from an independent rival, and purchases the parent brand would have made. Only the last source is cannibalization inside that brand. Lomax used panel-data methods to examine that component; the accessible abstract does not name the individual detergent brands or give each one's diversion rate.[^ref-a18a89f45c89]
In the fast-food study, the fitted average composition of sales at individual stores was 86.7% incremental, with the remainder attributed to nearby stores of the same chain. It is not a claim about every store or only about newly opened outlets. The estimate illustrates why an outlet's gross sales should not simply be added to system growth.[^ref-b86a118bc113]
Manages Complexity¶
The term keeps unit growth separate from system growth. It asks how much of one unit's sales would have occurred at another affiliated unit without the focal sales. This avoids treating every added sale at one unit as wholly new to the portfolio or chain.[ref-a18a89f45c89][ref-b86a118bc113]
A counterfactual is necessary. Consumer-panel histories and retail-demand models offer different ways to estimate diversion. Time changes, chain growth, and the choice of outlet location can complicate attribution; parallel sales movements by themselves are not a complete causal test.[ref-a18a89f45c89][ref-b86a118bc113]
Abstract Reasoning¶
First draw the commercial boundary. Next identify the focal unit and the unit whose sales may have been displaced. Estimate the purchases the latter would otherwise have received. Separate that within-system component from genuinely new demand and purchases drawn from independent systems. Only after that can a manager assess revenue, costs, margins, and the whole system's benefit.[ref-a18a89f45c89][ref-b86a118bc113]
The live Competition Prime is a strict parent at the unit level: two offerings or outlets contest some purchases, so a diverted sale gains one unit what the other would have received. This entry adds the shared portfolio, brand, or chain boundary and causal sales attribution. The commercial system can still gain overall.[ref-a18a89f45c89][ref-b86a118bc113]
Knowledge Transfer¶
The same roles appear in unlike carriers: detergent extensions and restaurant outlets. Each has a shared commercial system, two units that can serve overlapping buyers, a focal sale, and a possible displaced sale. The method of measurement changes, but the within-system diversion question survives.[ref-a18a89f45c89][ref-b86a118bc113]
The label does not automatically apply to two independent firms, any falling sales series, or two affiliated units that merely coexist. Outside the cited studies, the system boundary and causal diversion must be established anew.[ref-a18a89f45c89][ref-b86a118bc113]
Example¶
Detergent product lines. Lomax reports that three studied line extensions cannibalized their parent brands. Mapped roles: common system → the parent brand and each extension; focal unit → an extension; displaced unit → its parent brand; diverted sales → purchases the parent would otherwise have made; evidence → consumer panel techniques described in the publisher abstract. The accessible abstract supports the finding without individual brand names or rates.[^ref-a18a89f45c89]
Fast-food chain outlets. Pancras and colleagues fitted sales composition at individual outlets of one chain. Mapped roles: common system → chain affiliation; focal unit → an outlet whose sales are modeled; displaced units → nearby chain outlets; diverted sales → the modeled share drawn from those outlets; incremental component → sales the model attributes as new to the chain rather than diverted from its other outlets. The reported 86.7% average belongs to that fitted study, not every outlet.[^ref-b86a118bc113]
Relationships to Other Abstractions¶
Current abstraction Sales cannibalization Domain-specific
Parents (1) — more general patterns this builds on
-
Sales cannibalization is a kind of Competition Prime
Same-system sales diversion instantiates unit-level rivalry for overlapping customer purchases.
Hierarchy path (1) — routes to 1 parentless root
- Sales cannibalization → Competition
Neighborhood in Abstraction Space¶
Sales cannibalization sits in a sparse region of the domain-specific corpus (96th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Unclustered & Miscellaneous (2551 abstractions)
Nearest neighbors
- Material Flow Analysis — 0.78
- Kinetic Exchange Models of Markets — 0.77
- Dependency Grammar — 0.77
- Name your own price — 0.77
- Cornering the market — 0.77
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
Sales cannibalization is not all competition, because rivalry between independent systems lacks the common commercial boundary. It is not gross focal sales, an unexplained decline elsewhere, or a guaranteed loss for the entire business. A sale can move between affiliated units while total system sales grow. A particular profit decision requires more than the fact of diversion.[ref-a18a89f45c89][ref-b86a118bc113]
References¶
[^ref-a18a89f45c89]: Wendy Lomax, “The measurement of cannibalization”, Marketing Intelligence & Planning 14(7), 1996, 20–28, original publisher abstract. It describes consumer panel data for three detergent line extensions in UK and German markets and reports cannibalization of their parent brands by all three. Full article was not accessible here; no individual brand names or rates are asserted.
[^ref-b86a118bc113]: Joseph Pancras, S. Sriram, and V. Kumar, “Empirical Investigation of Retail Expansion and Cannibalization in a Dynamic Environment”, Management Science 58(11), 2012, 2001–2018, original publisher abstract. It reports a model-based composition of sales at individual fast-food chain stores, discusses growth and location endogeneity, and notes franchisee concerns. The 86.7% figure is the fitted study average, not an all-store or new-store rule.