Coffee aficionados more often left the coffee house disappointed when their favourite Starbucks beverage was unavailable due to stockouts. And as every retailer knows, a disappointed customer is a competitive disadvantage. To aid this recurring issue of stockouts across North America, Starbucks turned to what many companies see as the answer to operational inefficiencies today: artificial intelligence.
So, in September 2025, Starbucks rolled out an AI-powered inventory counting system across more than 11,000 North American stores, positioning the technology as a key step toward improving product availability, reducing stockouts, and freeing store employees from time-consuming manual inventory checks.
However, only 9 months in, the company scrapped the initiative, but why?
The answer lies in the rise and fall of Starbucks’ Automated Counting programme – one of the brand’s most visible and ambitious AI supply chain projects.
The Vision:
Starbucks developed the system in partnership with technology startup NomadGo and combined computer vision, 3D spatial intelligence, and augmented reality to automate inventory counting inside its stores. Moreover, early pilots of the automated counting programme delivered counts eight times faster and with 99% accuracy, according to both companies.
Employees would use handheld devices and scan shelves and storage areas while the system identified and counted products such as milk cartons, syrups, toppings, cups, and other inventory items.
As per Starbucks, the technology could complete inventory counts up to eight times faster than traditional manual methods. Not only saving on labour hours but also generating more accurate and frequent inventory data that could improve replenishment decisions across its store network.
For a company the size of Starbucks, better inventory visibility was expected to help reduce stockouts, improve customer experience, and provide supply chain teams with a clearer picture of product movement across thousands of locations.
As retailers increasingly look to AI to improve forecasting and supply chain responsiveness, Starbucks’ initiative stands as an example of how advanced technologies could modernise operational decision-making. But things didn’t quite go as planned.
When Scale Meets Reality
Once the technology was deployed across thousands of stores, employees encountered recurring issues with inventory accuracy, which defeated the purpose of the AI-powered inventory counting system.
The system sometimes misidentified products, confused similar-looking items, or failed to recognise certain inventory altogether.
Instead of eliminating manual work, some store teams reportedly found themselves spending additional time verifying counts and reorganising inventory to help the system function properly.
Therefore, in May 2026, Starbucks informed employees that it would discontinue the automated counting programme across North America. While Starbucks continues to invest in technology and automation, the move demonstrated that even well-funded AI initiatives can struggle to deliver consistent results when scaled across complex networks.
This rapid rise and fall of Starbucks’ automated counting system reflects a broader trend that’s unfolding across industries. While everyone is under increasing pressure to embrace AI, successful implementation depends on more than sophisticated algorithms. Data quality, process standardisation, employee adoption, and operational consistency require equal consideration.
The Starbucks experience serves as a reminder. Yes, technology does create sustainable value, but organisations must also recognise where human oversight remains indispensable.
