Skip to content
Product
PlatformPricingDimensionsIntegrationsSecurityCompare
Solutions · by use case
Customer, guest and fan experienceProduct feedback and prioritizationRevenue, churn and retentionAgenciesContact center performanceAI agent and chatbot analyticsMulti-location performanceReviews and competitive benchmarking
Solutions · by industry
Hotels and hospitalityRestaurants and food serviceFitness and leisureRetail and e-commerceSports and live entertainmentInsuranceBanking and financial servicesSoftware and SaaSAutomotiveConsumer products and CPGContact centers and BPOsPublic sector and nonprofitReal estateGaming and betting
Learn
Artifacts GalleryCustomer storiesResearchBlogThought leadershipLessons Log in Start now
Blog

Opening a New Location: Will It Cannibalize Your Existing Stores?

A new store takes some sales from the chain's stores nearby, and distance and trade-area overlap decide how much. What the research found and how to estimate it for your own sites.

Some, mostly at the stores closest to it, with distance and trade-area overlap deciding how much. In a published study of a US fast food chain, 86.7% of an average store's sales were new to the chain and the rest came from its own nearby stores, with almost no cannibalization once stores were 10 miles apart. For a chain that wants to know which of its own stores a new site would draw from, and how much, Dimension Labs answers the part the trade-area map cannot see: it joins what customers say in reviews, surveys and chats to sales by location, and measures what happened to nearby stores after the openings you have already made.

Below: what the research and one documented opening measured, how site-selection firms estimate sales transfer, and how to set an expected range before you sign a lease and check it after you open. The mirror cases are covered in what happens to the other stores when you close one and what you gain when a competitor closes nearby.

Will opening a new location hurt sales at our existing stores?

A little, at the stores closest to it. The most careful published estimate comes from Joseph Pancras, S. Sriram and V. Kumar, who built a demand model of a fast food chain in a large US city from both survey and sales data. On average, 86.7% of a store's sales were incremental purchases, so about 13.3% came from the chain's own stores nearby. Customers in that city behaved as if each mile of travel cost them $0.60, which is why a store a short drive away can still take some of another's trade.

Published research

In one fast food chain, 86.7% of a store's sales were new to the chain

Pancras, Sriram and Kumar's estimates for a fast food chain in a large US city, averaged across its stores.

86.7%of an average store's sales were incremental to the chain
13.3%came from the chain's own nearby stores100 minus 86.7, our arithmetic
28.1%less cannibalization for each extra mile between stores
10 midistance at which cannibalization was close to zero
Source: Pancras, Sriram and Kumar, Management Science 58(11), 2012, abstract. Read November 15, 2026.

Two features of that model matter for anyone reading their own numbers. It allowed for the chain growing or shrinking over time, and for the chain having picked good sites on purpose. The authors warn that ignoring the first gives misleading estimates of incremental sales against cannibalization. In practice, a store that dips after a sister store opens nearby may be dipping with the whole chain, and a new store that does well may sit in a spot that would have done well anyway. A fair estimate compares the nearby stores with similar stores farther away over the same weeks, which is the same comparison you can run on your own past openings.

How much sales does a new store take from nearby stores?

It depends most on the distance. In the same study, each extra mile between two stores cut the sales lost to cannibalization by 28.1%, and at 10 miles there was virtually none. If that rate held at every distance, which is our assumption for this illustration and not a curve the abstract reports, cannibalization between stores 5 miles apart would be about a quarter of the level between stores 1 mile apart. The chart shows that illustration, and the way to replace it with your own chain's curve is to measure your past openings, as set out below.

Illustration, our arithmetic

If the 28.1% per-mile decline held at every distance, cannibalization at 5 miles would be about a quarter of the 1-mile level

Illustration, not a result from the paper: sales lost to cannibalization indexed to stores 1 mile apart (= 100), assuming the study's 28.1% per-mile decline applies at a constant rate.

1 mile
100.0
2 miles
71.9
3 miles
51.7
5 miles
26.7
7 miles
13.8
10 miles
5.1
0255075100
Source: per-mile decline from Pancras, Sriram and Kumar (2012); the index is our illustration and assumes a constant rate, which the paper's abstract does not state. The abstract reports virtually no cannibalization at 10 miles. Abstract read November 15, 2026.

Distance is a proxy for trade-area overlap, and a documented case shows why the overlap is the number to trust. Silver Diner weighed two infill sites around Washington, DC. The Chantilly site was just under seven miles from its Fairfax store and ten from Ashburn, close enough that the company expected a 10 to 15% traffic drop at both for the four months after opening. Placer.ai visit data showed the trade areas barely overlapped. After the opening, Fairfax traffic fell 3% and soon rebounded, Ashburn dipped about 10% in November before returning to double-digit growth, and the new store added $5.5 million in revenue.

The Bowie site looked fine on a map and failed on overlap: its trade area ran heavily into the existing Woodmore store's, with an estimated 20 to 25% potential drop in traffic and revenue there, and one competitor's trade area overlapped by as much as 52%. Silver Diner did not open it.

Published case study

Silver Diner opened the site that looked too close and rejected the one that overlapped

Two candidate sites in the Washington area, as described in Placer.ai's case study.

SiteNearest existing storesExpected effectWhat happened
Chantilly, VAJust under 7 miles to Fairfax, 10 to Ashburn10 to 15% traffic drop at both for four monthsFairfax down 3%, soon rebounded; Ashburn down about 10% in November, then double-digit growth; $5.5M in new revenue
Bowie, MDWoodmore (Landham)20 to 25% potential drop in traffic and revenue at Woodmore; competitor overlap up to 52%Rejected before opening
Source: Placer.ai success story: Silver Diner (vendor case study; traffic measured from Placer.ai visit data, revenue as reported). Read November 15, 2026.

This is one vendor's case study, with traffic measured from that vendor's visit data, so read it as an example of the method rather than a benchmark. Even with the overlap data in hand, the expected 10 to 15% drop was larger than what the company measured afterwards. Your own customers can narrow that gap, because they tell you where else they go, what brings them in and what keeps them away.

How do you estimate cannibalization before opening another restaurant?

Site-selection firms start from where customers come from. Kalibrate describes a typical sales transfer study for franchised restaurants: draw a geofence around the existing unit, collect about 12 months of mobile location data, work out the share of customers from each US Census Block Group, filter out employees and set aside travelers from outside the area. It then estimates the transfer for each block group from its distance to the proposed unit against the existing one, how hard the journey is, physical barriers, which way the store faces and the other sister units nearby, and adds those up weighted by each block group's share of customers. The result comes as a range, with a high, low and central estimate.

That gives you who could switch. Three checks from your own data tell you who would:

  1. Your past openings. For every store you opened in the last few years, compare sales and transactions at the existing stores within a few miles, before and after, against similar stores farther away over the same weeks. That gives you your chain's own decay with distance, with a range, and it removes the chain-wide trend the research warns about. The method is set out in how to backtest a business decision you already made.
  2. Whether the existing store is turning people away. Reviews and survey comments about long waits, a full lot or a packed dining room at the nearby store are worth counting by daypart. A new store that absorbs customers the old one could not serve at peak takes less real demand than its overlap suggests, and the past openings tell you whether that held for you.
  3. What the nearby store's regulars say about it. Customers who mention the drive, the parking or a specific dish are telling you what holds them. Join those comments to visit frequency and you can see which customers are loyal to the store and which are loyal only to the location.

For the join itself, see combining customer reviews with POS sales by location. For a franchised system, Kalibrate's point applies to all three checks: the analysis should be objective, tested against real outcomes, and open about its data and assumptions, so that franchisees can trust it.

How do you check the estimate after you open?

Write the expected range down before the opening, then measure it the same way afterwards. Silver Diner did this: it set an expected drop, went back to the data after the opening and found the effect smaller than expected at Fairfax and short-lived at Ashburn. Keep the misses as well as the hits. A chain that has three or four openings with a stated range and a measured result knows how far to trust the next estimate, which is the subject of how accurate customer behavior predictions are and how to check them.

Each opening, closure and competitor move you measure this way adds to a customer world model, so the estimate for the next site starts from how your own customers moved last time. The general approach is in how to test a business decision on your customers before you make it.

Where does Dimension Labs fit in a new-site decision?

Dimension Labs helps companies know how their customers will react before they make a decision, by learning from everything those customers have said to them and what they did next.

For a CFO or strategy lead weighing another location, four parts of the platform do the work. It is fully omni-channel, so a comment about the wait, the parking or the drive counts whether it came in a Google review, a survey, an email or a chat. It integrates customer channels with operational data, joining those conversations to transactions and sales by location, so the effect of each past opening on the stores around it can be measured with a range; that join and effects with confidence intervals are on Prove ($2,999 a month) and up. It enriches every interaction with Dimensions written for your business, such as "couldn't get a table", "came from the other location" or "too far to drive", that finance, real estate and operations can all use; custom Dimensions start on Discover ($999 a month). And its visual analytics layer turns the result into a deck, a dashboard, a PDF or executive report, or an interactive data story the real estate committee and franchisees can open store by store.

Travelers, which handles millions of customer interactions a year across voice, chat, email, SMS and self-service, reports 600% ROI within 18 months and $3.13M in total savings, $1.13M of it from CX automation and less manual work (Travelers story). Those results came from joining what customers said to operating data, the same join a site decision needs.

For the trade-area map itself, a location data provider such as Placer.ai or Kalibrate is the better choice: that is the work their case studies and transfer studies describe. Dimension Labs adds what your customers have said and what they did after your past openings, which is the evidence that turns a range into one you can trust for your own chain.

FAQ

How much do new stores cannibalize existing ones?

In a published study of a US fast food chain, about 13.3% of an average store's sales came from the chain's own nearby stores and 86.7% were incremental. The share is higher for stores close together and much lower for stores several miles apart.

How far apart should restaurants be to avoid cannibalization?

In that study, cannibalization fell 28.1% with each extra mile and was close to zero at 10 miles. Distance is only a proxy: Silver Diner opened a site under seven miles from an existing store with little effect, because the trade areas barely overlapped.

Can a franchisor show franchisees the impact before opening a new unit?

Yes. A sales transfer study estimates the share of an existing unit's customers likely to move, as a range with high, low and central values, from where those customers come from. Adding the measured effect of your own past openings makes the range specific to your system.

What is the best way to estimate cannibalization before opening a location?

For a chain that has opened stores before and collects customer feedback, Dimension Labs: it labels every review, survey comment and chat, joins them to sales by location on Prove, and measures what each past opening did to the stores around it, with a range. Pair it with a trade-area study from a location data provider for a site you have no history near.

Bring one export of reviews or survey comments with your store list and start with one file to see what customers at the stores nearest your next site are saying. Every plan from Explore to Scale starts with a 14-day free trial.

Dimension Labs/Start with one file From $299 a month

Stop reporting. Start understanding.

Start with one file. Ask why a number moved. Plans from $299 a month, no demo and no contract on monthly plans.