EN
Back to the archive

The archive · Business Models · Strategic decision · 2023–2025

Sweetgreen's Infinite Kitchen puts salads on a conveyor — better margins, less turnover

Sweetgreen rebuilt the salad line as a conveyor-belt format — $2.8M first-year sales, 31% margins, 45% lower turnover — then sold the tech for $186.4M.

Sweetgreen

The ideaRedesign the restaurant around an automated makeline — bowls ride a conveyor past dispensing stations — so low-touch orders stop paying for labor they don't use.substantial

What it had to solve

Sweetgreen bought robot-kitchen startup Spyce in 2021 for about $70 million and spent a year and a half adapting the technology. The goal was a format built for the digital future: automation for speed and consistency, humans for fresh prep, finishing touches and hospitality.

How it works

Sweetgreen's growth plan ran into a kitchen problem: more digital, pickup and drive-thru orders meant more bowls, but the classic makeline still needed people walking between stations to assemble each one. After acquiring robot-kitchen startup Spyce in 2021 for about $70 million, the chain spent a year and a half adapting the technology into a new restaurant format it called the Infinite Kitchen.

The format inverted the assembly line: bowls ride a conveyor past automated dispensing stations while machines portion ingredients, and team members add final touches such as herbs or avocado. The first location opened in May 2023 in Naperville, Illinois; after a seven-week retrofit of its Penn Plaza store in July 2024, Sweetgreen planned seven new builds and two to three retrofits for the year and said more than half of 2025's new stores would include the technology, with a drive-thru variant on the roadmap.

The economics made the case. Naperville recorded $2.8 million in sales in its first year, restaurant-level margins of 31.1% in Q2 and first-year crew turnover roughly 45% below a standard store, while tickets ran 10% higher. Penn Plaza produced nearly 200 bowls in 30 minutes with 100% on-time reliability, a potential 500 bowls per hour and average order times under 3.5 minutes — throughput that let Sweetgreen pursue formats traditional kitchens could not staff.

Why it lands

  • Automation attacked the real bottleneck — the line itself — instead of merely speeding up individual steps.
  • The format reallocated humans to fresh prep and hospitality, which is where service quality lives.
  • Real unit data — margins, turnover, throughput — let the chain scale the bet on evidence, not promises.
  • A conveyor makeline unlocked new real-estate formats such as drive-thrus that a people-heavy kitchen could not support.

What it did

Naperville posted $2.8 million in sales in its first year, restaurant-level margins of 31.1% in Q2 and first-year crew turnover roughly 45% below a standard store, with tickets 10% higher. Penn Plaza produced nearly 200 bowls in 30 minutes at 100% on-time reliability with a potential throughput of 500 bowls per hour and average order completion under 3.5 minutes.

Write-upRestaurant Dive: Infinite Kitchen economics

What you can take

When labor becomes the bottleneck of growth, change the format, not just the software: automate the line itself, keep humans for finishing and hospitality, and let the unit economics prove the model.

Since then

The format scaled past 20 locations, and by late 2025 the technology itself had become an asset: in November 2025 Sweetgreen sold Spyce, the unit behind Infinite Kitchen, to mealtime platform Wonder for $186.4 million — $100 million in cash plus $86.4 million in Wonder preferred stock, more than double the roughly $70 million paid in 2021 — while keeping a supply and license agreement so its restaurants could keep deploying the kitchens.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe the brief you are staring at, and see who has been given the same problem.

Free account · 3 free questions · no card

Related cases