Starbucks: The Store Across the Street
Why nearby locations can capture different flows—and when density becomes too much
Brand Learning · Location Changed Everything · Case 4 · Collection I · Prepared October 2026
Photo: Larry LaRose via Wikimedia Commons. Displayed with responsive sizing.
Starbucks learned that proximity is not the same as duplication. Different corners, walking paths, queues, and customer missions can sustain stores that appear dangerously close.
The location decision
In early-1990s Vancouver, Starbucks faced a practical problem. A small, high-performing store sat in a building scheduled for renovation. The company opened a larger replacement roughly 30 yards away. When the original location remained open longer than expected, the supposedly temporary overlap became an experiment.
Both stores performed. The explanation was physical behavior. Customers approached the intersection from different directions. Many would not cross, reverse course, or wait through a long line for a low-ticket convenience purchase. A second store could reduce lost demand even while taking some transactions from the first.
The company extended the idea into a broader density strategy. The objective was not to maximize the sales of each individual store in isolation. It was to maximize chain-level market capture, customer convenience, brand presence, and operational density. Marketing Science research later estimated that a neighboring Starbucks imposed average cannibalization of about 1.2 percent within one mile and 0.4 percent within one to three miles, while remote stores experienced higher operating costs as distance from the network increased.
Density, however, is not self-justifying. In 2008 Starbucks committed to close approximately 600 underperforming U.S. company-operated stores. About 70 percent had opened since the beginning of fiscal 2006. The company recorded major asset, lease, and restructuring costs. The same organization that proved close stores could work also proved that expansion speed, market conditions, and store-level returns still matter.
The strategic model
Micro flows
Corners, transit exits, office entrances, turn restrictions, and pedestrian direction can matter more than straight-line distance.
Queue capacity
A nearby store can recover customers who would abandon a long line.
Network economics
Dense stores may share brand awareness, management, delivery, labor pools, and local knowledge.
Chain-level optimization
Some cannibalization may be rational when the network captures more total demand or blocks a competitor.
What the strategy got right
It tested observed behavior instead of relying on a universal spacing rule.
It evaluated the market as a network rather than protecting every store from overlap.
It recognized that convenience purchases have friction at very small distances.
Where the model can fail
Cannibalization is real even when density is beneficial. The relevant question is whether network gains exceed the shifted sales and added cost.
A successful dense urban cluster cannot be copied into lower-volume markets without accounting for customer flow and fixed costs.
The 2008 correction showed that fast openings can outrun site quality, economic conditions, and acceptable returns.
The impact
Starbucks turned a counterintuitive overlap into a repeatable growth idea, then learned the cost of treating growth velocity as proof of demand. The best lesson is neither “cluster stores” nor “protect every trade area.” It is to model customer flow, lost demand, network effects, and cannibalization together.
Questions for leaders and franchise consultants
Do nearby sites intercept the same trip or different trips?
Which direction do customers approach from, and what barriers alter their path?
How much demand is lost to queues, parking, turns, or street crossings?
What network benefits come from density?
At what point do incremental stores stop adding market capture?
Applying the lesson with Wanderland
Wanderland can map store clusters, customer or business lists, walk or drive-time catchments, demographic fit, competitors, and street context. A useful comparison treats cannibalization as one weighted factor—not a veto—and makes the assumed benefit of density visible to the decision team.
Research sources
Independent educational analysis. Company names identify the subjects of the stories; no endorsement or affiliation is implied. Historical results retain their stated source periods.
Comments