The system, developed by Seattle startup NomadGo, used LiDAR sensors and tablet cameras to automatically scan and count products like syrups and milk stored behind counters. Starbucks expanded the rollout shortly after Niccol became CEO in September 2024, positioning the technology as part of a broader effort to improve store operations, reduce errors, and keep high-demand menu items consistently in stock.
Instead, the system struggled with basic product recognition tasks during real-world use. Reuters reported that employees frequently encountered inaccurate counts, with the software confusing similar-looking products or failing to detect items already sitting on shelves. One Starbucks launch video reportedly showed the tool missing a peppermint syrup bottle entirely while scanning surrounding inventory.
The rollout highlighted a recurring challenge facing enterprise AI deployments: systems that perform well in controlled demonstrations often struggle under the variability of large-scale retail operations. Inventory management inside Starbucks stores involves constantly changing stock levels, visually similar products, different store layouts, and fast-moving employee workflows that leave little room for inaccurate recommendations or added friction.
Starbucks publicly described the decision as part of a broader operational standardization effort rather than a failed AI initiative. In a statement to Reuters, the company said the move was tied to “a decision to standardize how inventory is counted across coffeehouses as we continue to focus on consistency and execution at scale.”
The company added that it remains focused on improving supply chain reliability and daily store restocking. “Our goal is simple — if it’s on the menu, customers should be able to order it,” Starbucks said.
Employee reactions shared internally suggested growing frustration with the tool long before the shutdown. One store worker wrote: “Thanks for discontinuing Automatic Counting! The thought behind it was great, but the execution was proving difficult.”
Inventory shortages and operational consistency have remained ongoing issues for Starbucks leadership over multiple administrations, with executives repeatedly citing product availability problems as a drag on customer experience and revenue. Automated Counting had been positioned as one part of a wider technology push that also includes AI-assisted drink sequencing and tools designed to help baristas manage store traffic during peak periods.
The decision also underscores how quickly companies are reassessing AI projects that fail to produce measurable operational improvements. While businesses across retail and food service continue investing heavily in automation and AI-assisted workflows, many deployments are still running into the harder realities of scaling those systems reliably across thousands of physical locations.
NomadGo said in a statement that the company is “continuously learning from customer and user feedback” as it continues refining its technology.
This analysis is based on reporting from Yahoo Finance.
Image courtesy of Unsplash.
This article was generated with AI assistance and reviewed for accuracy and quality.
