How IBM Uses watsonx to Unlock $4.5B in Productivity Gains

IBM entered its own transformation program with a clear premise: if AI-led productivity works at enterprise scale, prove it on yourself first. With 280,000 employees distributed across 170+ countries and operations spanning every major business function, IBM represented as demanding a test environment as any client it serves.

Across the business, the pre-AI picture was one of fragmentation. HR employees collectively spent hours navigating policy documents across 175 countries. Procurement teams worked across 40+ disconnected systems. IT support desks were buried in routine tickets. Sales teams switched between five separate applications to complete basic prospecting tasks. In supply chain, 10 million annual shipments and 350,000 SKUs created a data environment where decision-making lagged and costs were difficult to optimize.

IBM’s response was architectural and cultural simultaneously. The company deployed its watsonx portfolio — including watsonx.ai, watsonx Orchestrate, and watsonx Code Assistant — alongside IBM Watson Discovery, IBM Turbonomic, and IBM Apptio to cover distinct functional needs. The execution model emphasized agile two-week sprints, CEO-level sponsorship, and mandatory business-outcome measurement for every initiative. Employees were invited to propose AI use cases, generating 15,000 submissions during the 2025 IBM watsonx Challenge.

The numbers that emerged were difficult to overstate. IT support tickets fell 56% from 2022 to 2024. The AskHR AI agent now resolves 94% of common HR inquiries, cutting support volume by 75%. Supply chain logistics costs dropped approximately 30% over three years. The finance team achieved an estimated 90% reduction in cycle time for journal processing. In sales, the AskSales assistant answered over 250,000 seller questions in 2025 alone. Total productivity gains across all functions reached $4.5 billion.

IBM is now building AskIBM, an orchestration layer designed to unite its domain-specific AI agents into a unified interface. The company has effectively converted internal transformation into a product playbook — a replicable blueprint it can offer clients who face the same complexity it just navigated.