Companies Rebuild IT Systems Before Adding AI Tools

Seagate replaced its decade-old IT service management platform in three months after a contract deadline. The company moved 30,000 employees onto a new system by rebuilding service catalogs, setting consistent service-level agreements across regions, and rewriting ticket categories.

The rebuilt structure allowed an AI agent to deflect about one-third of incoming tickets. First-contact resolution rose 27 percent above the industry average.

Group research from September 2025 found that 60 percent of companies generate no material value from AI. Freshworks' upcoming Cost of Complexity study estimates that one quarter of AI budgets is spent on integration and data cleanup. MIT reported that 95 percent of generative AI pilots fail to reach production. The pattern appears across industries where legacy systems remain fragmented.

Balance consolidated its IT stack onto one platform with a single source of truth. The company has 9,000 employees compared with Nike's 80,000. Katz Media Group CTO Robert Lyons maps potential AI projects on a value and effort matrix. His team cleaned and labeled data and ran an AI primer webinar for all employees before any tool deployment.

Lyons said the sequenced approach separates companies that are getting AI to work from those still discussing it.

Successful mid-sized firms reduced system fragmentation so data can move without manual handoffs. They applied AI to measurable workflow tasks such as ticket resolution, demand planning, and predictive maintenance. These companies treated AI adoption as an operating discipline measured by concrete business outcomes rather than a standalone technology project.