Starbucks Develops In-House AI to Replace IBM Maintenance Tool

What happened

Starbucks disclosed plans to replace third-party enterprise software from IBM and Microsoft with internally developed AI-powered alternatives, according to an internal presentation. The company is building a replacement for IBM's maintenance management tool using AI-assisted coding, as well as an alternative to Microsoft's inventory tracking system. Chief Technology Officer Anand Varadarajan told employees that Starbucks spends roughly $400 million per year on software and sees clear opportunities to reduce that spending. Some of the internally built software could roll out by the end of next year pending successful testing. The company's enterprise technology team is on track to reduce its budget by approximately $30 million in the fiscal year ending late September, including about $10 million from software spending. Starbucks has also been working for several years on creating a point-of-sale system to replace Oracle Simphony. The company is encouraging technology workers to use artificial intelligence tools and has factored AI usage into employee bonus evaluations.

Why it matters

For IBM, this development represents a tangible example of how AI-enabled in-house development is disrupting traditional enterprise software vendors. Starbucks's shift away from IBM's maintenance management platform-driven by the lower cost and customization advantages of AI-assisted coding-signals a broader risk to IBM's enterprise software business model. The coffee chain's ability to build replacement tools internally, despite not being a technology company, demonstrates how accessible AI development platforms have made it for large enterprises to reduce reliance on legacy vendors. This trend could pressure IBM's software revenue and pricing power, particularly for maintenance and operational tools that large customers can now replicate more easily. Research firm Gartner estimates that up to $234 billion of enterprise application software spending will be exposed to what it calls agentic arbitrage by the end of 2030, representing roughly 20% of all enterprise software-as-a-service spending. IBM shares fell 5.2% following the Bloomberg report, reflecting investor concern about the company's vulnerability to AI-driven disruption.

Bigger picture

Starbucks's initiative reflects a structural shift in enterprise software economics driven by advances in artificial intelligence. For two decades, businesses accepted vendor feature sets, paid per seat, and hired specialists to manage platforms. AI coding tools from providers like Anthropic, Lovable, and Replit are changing that calculation. Five small companies switched from Salesforce and HubSpot to in-house applications in the last six months, reducing software costs by 40% to 80%. Low-code platform Retool found that 35% of enterprises have already replaced at least one SaaS tool with a custom-built alternative, with 78% intending to develop more this year. This trend poses a threat to established enterprise software vendors including IBM, Microsoft, Oracle, and Salesforce. Before advanced AI models, businesses were tethered to technology vendors due to fear of disruption and complexity of building in-house tools. AI is shifting that calculus by making it easier to develop applications from scratch. However, risks remain: Starbucks recently pulled an AI-powered inventory tracking system after it frequently miscounted and mislabeled items. While in-house software can be cheaper upfront, building can lead companies to pay higher maintenance and labor costs over time.

What to watch

Monitor whether Starbucks successfully deploys its IBM and Microsoft replacement tools by the end of next year and whether the company achieves its $10 million software spending reduction target. Watch for similar announcements from other large enterprises shifting away from IBM's software platforms. Track IBM's enterprise software revenue trends and pricing power in upcoming quarterly results, particularly for operational and maintenance tools vulnerable to AI-driven in-house development. Observe whether IBM accelerates its own AI capabilities to defend against customer defection. Watch for commentary from IBM management on competitive threats from AI-assisted coding and customer build-versus-buy decisions. Monitor whether the broader $234 billion agentic arbitrage threat materializes across the enterprise software sector through 2030.