Starbucks Taps AI to Reduce Reliance on Microsoft, IBM Software
Starbucks is developing AI tools to reduce its reliance on Microsoft and IBM software. The move sent software stocks lower as companies like Microsoft may lose a major customer.
SBUX is building in-house AI tools designed to replace commercial software it currently licenses from MSFT and IBM, Bloomberg reported. The coffee chain is specifically targeting a Microsoft Dynamics 365 system used to track store inventory and an IBM TRIRIGA tool used to manage facility maintenance, aiming to build internal replacements with the help of AI-assisted coding. News of the plan hit both vendors' shares on July 9, 2026: IBM fell as much as roughly 4 to 5.6% in premarket trading, while Microsoft slipped about 1.5%, as investors weighed the risk of losing a large enterprise customer.
The initiative sits inside a broader $2 billion cost-cutting push at Starbucks, which spends roughly $400 million a year on software. Chief Technology Officer Anand Varadarajan has told employees there are "clear opportunities to reduce the spend in software," and the enterprise technology team is targeting about $30 million in budget cuts for the fiscal year ending in late September, including roughly $10 million from software licensing and another $13 million from lower contractor spending. This is not Starbucks' first attempt to insource enterprise software: the company has separately spent several years developing a replacement for Oracle's (ORCL) Simphony point-of-sale system, suggesting a deliberate, multi-front strategy of using AI development tools to reduce dependence on outside vendors rather than a one-off decision.
For Microsoft and IBM, the episode is an early, concrete example of a risk that has loomed over the enterprise software sector: AI-assisted coding could make it cheaper and faster for large customers to build their own replacements for licensed systems. The rollout is not immediate. Starbucks expects to test the new inventory and maintenance tools before any rollout, with some systems potentially live by the end of 2026, and the company has already shown this approach carries execution risk, having previously pulled an AI-based inventory-tracking system after operational issues and reverted to manual counting. What to watch next: whether the pilot software clears testing without disruption to stores, whether other large retailers or enterprises follow with their own AI-built replacements for licensed software, and how Microsoft and IBM address customer-insourcing risk on upcoming earnings calls.
Related Stocks
Powered by SentiSense - Intelligent Market Analysis