What If Starbucks Is Solving the Wrong Problem?

Starbucks announced this week that it will close roughly 250 North American stores, about 1% of its footprint, after concluding that those locations could not consistently deliver the experience it wants or reach acceptable financial performance.¹ The closures carry roughly $300 million in charges, two-thirds of it from exiting leases early and paying separation benefits.²

At the same time, the company is accelerating something much larger. It has redesigned more than 1,000 coffeehouses across the U.S. and Canada since late 2025 and intends to complete at least 1,500 of these "uplifts" by the end of fiscal 2026 — softer seating, artwork, greenery, local design — with the explicit goal of making people stay longer and restoring Starbucks as the "Third Place."³

Note

The eventual program is reported at roughly $1 billion across as many as 9,000 stores.⁴

Brian Niccol's premise is that Starbucks lost what made it special: the stores became transactional, mobile orders crowded the counters, and customers stopped seeing the coffeehouse as a place to spend time. There is considerable truth in that.

But Starbucks may be fixing the experience while the larger problem is the economics of the customer — rebuilding the coffeehouse for a consumer who can no longer afford to use it the way they once did.

The Consumer Never Got the Old Dollar Back

Slower inflation means prices stopped rising as quickly; the higher price level stays. The Consumer Price Index stood at 257.971 in January 2020 and 334.980 in August 2026, still climbing at 3.4% a year.⁵ Prices are almost 30% higher than immediately before the pandemic, which means a 2020 dollar buys about 23% less.¹⁵

Coffee at Starbucks is not rent, insurance or groceries. It is discretionary, and discretionary spending is what remains after everything else has been paid.

Consider a customer who once visited four times a week at $4.50, about $72 a month. If the same order approaches $7, keeping the habit costs $112. The customer doesn't have to stop liking Starbucks or complain about the experience. They only have to go less often. Eight visits at $7 comes to $56.

The customer can remain loyal while becoming economically less valuable.

Starbucks May Be Diagnosing the Wrong Variable

Starbucks has evidence the turnaround is working: four consecutive quarters of comparable-sales growth, with U.S. comps up 7.9% in fiscal Q3 2026, led by a 4.2% rise in transactions.⁶ That should not be dismissed, but it doesn't show that the couches did it.

Comparable transactions are counted per store, not per customer, and a chain closing hundreds of locations pushes displaced traffic into the stores that survive. The same quarter also included new service standards, a redesigned Rewards program, menu changes and a growing delivery business.⁷ The mistake would be watching a business improve after ten changes and crediting the one that costs a billion dollars.

The larger question is whether Starbucks is investing in a behavior that has structurally changed.¹⁶ The Starbucks of the 1990s and 2000s lived inside a daily route — home, Starbucks, office — and turned that stop into an experience. COVID broke the route: work dispersed, mobile ordering and delivery became normal, and the purchase became transactional just as the purchasing power behind it shifted.

Starbucks may be trying to restore both an economy and a routine that no longer fully exist.

Then Look at Luckin

Luckin is making almost the opposite bet. At the end of 2025, 99.1% of its stores were pickup stores of 20 to 60 square meters — roughly 215 to 645 square feet — with limited seating, a format Luckin says lets it expand rapidly at lower rental and decoration cost.⁸ Ordering, payment, and promotion all run through the app.

Luckin isn't building coffee shops. It is building coffee distribution nodes, where the store exists to manufacture the beverage and hand it over.

• Starbucks monetizes coffee plus labor, real estate, atmosphere, service, and brand.
• Luckin monetizes coffee plus convenience, density, software and price.

Those architectures behave very differently once consumers become price-sensitive — and the contrast has already been tested. In China, Luckin's home market, Starbucks spent years losing ground to chains selling drinks at roughly half its price,⁹ and in fiscal Q3 2026 it converted its China retail operations to a licensed joint venture.⁷ The experience-first model did not win where it met the distribution-first model head-on.

A large Starbucks has to justify its rent, remodel and labor, which works when consumers gladly pay a premium for the whole package. When budgets tighten, the gap between a $7 experience and a $3 caffeine transaction becomes the thing the customer sees.

What the Alternative Looks Like at Scale

Luckin ended Q2 2026 with 36,310 stores after adding 2,714 in a single quarter, with 112.7 million average monthly transacting customers and revenue up 28.5% to about $2.34 billion.¹⁰

The caveats are real: same-store sales at self-operated locations fell 5.3%, which Luckin attributes mostly to heavy delivery subsidies a year earlier,¹⁰ and store-level margin had already compressed from 17.0% to 13.6% in Q1.¹¹

Density is not free. But a network business can tolerate weaker individual stores, because Store B cannibalizing Store A is acceptable if the network captures more total consumption.

Important

A Starbucks location has historically needed to matter as a destination. A Luckin location only needs to make the network denser.

The Consumer Is Already Bifurcating

Households earning $250,000 or more now account for nearly half of all consumer spending, and the top 10% of earners spend as much on discretionary items as the bottom 70% combined.¹² Lower-income cohorts stopped cutting nominal spending in the first half of 2026,¹³ but nominal spending barely outrunning 3.4% inflation is not a customer with room for a bigger coffee habit.¹⁴

Starbucks sits directly on that divide. For an affluent household, a $7 drink is noise; for another, four a week is a $100-plus monthly line item.

Starbucks can retain customers, grow revenue, and post transaction-led comps while losing frequency in the broad middle as its mix drifts upscale, and rising prices can conceal that for a long time.

Important

The number I would want to see is not comparable-store sales. It is visits per customer per month, segmented by income.

Maybe There Shouldn't Be One Starbucks

None of this means abandoning the third place. There are downtowns, campuses and affluent neighborhoods where a beautiful coffeehouse serves exactly that function. The mistake is imposing its economics on the whole network. The logical endpoint is three formats:

• A Third Place store built around experience and premium beverages
• A Traditional café and drive-through hybrid for suburban and commuter traffic
• A Starbucks Express of 300 to 600 square feet built for mobile pickup and delivery, with minimal seating, lower rent, lower build-out costs, and potentially lower prices.

Luckin already runs this split with the ratio inverted. Its 0.9% of non-pickup stores exist for branding,⁸ and this year it opened a 420-square-meter flagship in Shenzhen.⁹

Luckin builds the showpiece to support the network. Starbucks is building the network to support the showpiece.

This Is Really a Bet on Purchasing Power

Starbucks is betting that consumers still have enough discretionary income to pay for an experience wrapped around coffee. Luckin is betting that coffee can become a high-frequency convenience product by stripping that experience, and its cost, away.

One model says make the $7 purchase feel worth $7. The other says don't require the customer to spend $7 in the first place.

Wages rose too, but averages obscure exactly the customers who matter most to a mass-market discretionary business. A customer does not need to become poor to change a Starbucks habit. They only need to become conscious of the transaction, and once a purchase is something the consumer thinks about, frequency changes.

• Starbucks is rebuilding a place worth paying more to visit.
• Luckin is building a system cheap enough that visiting barely requires a decision.

If purchasing power stays under pressure, Starbucks could execute its return to its roots exceptionally well...

And still discover that the roots grew in a different economy.