Store cannibalization: what it is and how to measure it before you open

What is store cannibalization?

Cannibalization happens when a new store serves customers who used to buy at another one of your stores. The sales do not disappear; they move. The new store books them and the nearby store loses them.

That is why it helps to split any new store’s sales in two. Incremental sales are sales the chain did not have before: new customers, or customers who used to buy from competitors. Transferred sales are sales that simply changed stores.

For example, if a new store sells 100 and 30 of those sales used to happen at your nearest store, the chain grew by 70, not 100. Those 30 are the cannibalization. The numbers are only an illustration.

Why does it matter to a retail chain?

Every opening commits a large investment and a multi-year lease. If the new store lives off customers who were already yours, the problem does not show up in its own report. It shows up at the nearby store, which drops for no obvious reason.

The manager of that store gets the blame, targets end up wrong, and the chain pays two leases for roughly the same sales.

With franchises, it is even more sensitive. If an opening takes sales from a neighboring franchisee, the conflict can end up in the contract. That is why cannibalization should be measured with the same criteria for every location.

And your investment committee will ask. Bringing an estimate of transferred sales, not just the new store’s total forecast, makes the decision much easier to defend.

How do you estimate cannibalization before opening?

There is no single formula. The most reliable approach combines three views and checks whether they tell the same story. If they do not, that is the question to investigate before you sign.

• Trade area overlap. Map each store’s trade area by travel time, such as everything within a 10-minute drive, and measure how much of your existing store’s area the new one would cover. Use drive times, not radius circles: a river, a highway, or a missing U-turn changes who can get there.
• Visitor flows. Aggregated, anonymized mobility data shows which areas your current store’s visitors come from and what share of them come from the area the new store would cover. If that share is large, the risk goes up.
• Incremental versus transferred sales. A sales model estimates how much the new store would sell and how much each nearby store would lose. Gravity models, such as the Huff model, split the probability of a visit across every store in the area based on distance and each store’s attractiveness.

What data do you need to measure it?

Your best evidence is often your own history: if you have opened stores near others before, look at what happened to the nearby stores in the months that followed. This is the data that helps most:

• The exact location of your stores and their sales by store, ideally by month.
• Each store’s trade area, calculated by travel time.
• Population and socioeconomic profile by area, to know whether there is enough demand for both stores.
• Your competitors’ locations, because a new store also takes customers from them.
• Mobility data showing where each store’s visitors come from and how many they share.
• Housing, office, or retail projects that will change the area.

When is cannibalization worth accepting?

Not all cannibalization is bad. The right question is not whether the new store takes sales from another one, but whether the whole chain sells and earns more with both stores than with one.

It is often worth it when your current store is saturated and loses customers to long lines or full parking lots; when a competitor is looking for a site in the same area and, if you do not open, those sales leave anyway; when the area is growing with new housing or offices; or when the new format serves a different need, such as a small convenience store next to a large weekly-shopping store.

This table summarizes the most common scenarios:

Warning signs and common mistakes

These signs suggest an opening may be moving sales around rather than creating them:

• The new store’s forecast depends mostly on areas that currently feed another one of your stores.
• The new store’s trade area covers much of the area of a store that is already underperforming.
• Overlap was measured with fixed-radius circles instead of travel times.
• Nobody checked what happened to nearby stores after past openings.
• The area is not gaining residents, offices, or businesses that would bring new customers.

How do you check cannibalization after opening?

The pre-opening estimate is a hypothesis. After you open, test it, both to correct course in time and to forecast the next opening better.

Compare sales at the nearby stores before and after the opening with sales at similar stores far from it. That separates the effect of the new store from what happened to the whole market, such as seasonality or a price change.

Also track shared visitors and visitor origins for each store over several months. The first few weeks are often misleading, with opening-day curiosity and launch promotions.