Starbucks stops using its AI inventory counting app after errors in product counts and item identification.
Starbucks has stopped using an AI-powered inventory counting tool across its North American stores after the system produced counting and labeling errors.
The app, developed by NomadGo, used tablet cameras and LiDAR to scan shelves and count store inventory. Starbucks began deploying the technology in September 2025 as part of a wider push to improve product availability and reduce manual store tasks.
The tool was designed to help employees count inventory faster and identify low-stock items. NomadGo previously said its Inventory AI could perform counts up to eight times faster than manual processes and with high accuracy.
However, Reuters reported that workers and managers had flagged issues with the system, including missed items and confused product names. Starbucks later said it would end use of the app to standardize inventory counting across stores and focus on consistency at scale.
The decision comes as Starbucks continues to invest in technology as part of its turnaround strategy under CEO Brian Niccol. The company has said digital tools remain important to improving efficiency across stores and support operations.
The rollback highlights a broader challenge for retailers adopting AI in store operations. While automation can reduce manual work, tools used in live retail environments must be accurate enough to support frontline execution without adding confusion or rework.
