Bristol Myers Squibb cut its procurement request timeline from 6–9 months to under 30 days by deploying AI without waiting for perfect data, proving that "good enough" data can unlock faster business results.

What happened: Bristol Myers Squibb launched an AI-driven procurement platform with Palo Alto-based Globality in November 2024 (soft launch) and scaled it organization-wide in February 2025. The company's request-for-proposal timeline dropped from six to nine months to under 30 days. Over $1 billion（約1600億円） flowed through the platform in the first year, far exceeding initial projections, and the company enabled roughly ten times more requests for proposals while bringing previously outsourced work back in-house with about 50% fewer resources.

Why it matters: A December 2025 report from RGP found that only 10% of chief financial officers fully trust their data quality, and over one-third cite data trust as the main obstacle to AI return on investment. Bristol Myers' success—achieved despite moving forward without flawless data—suggests that enterprises can begin extracting value from AI and operational gains sooner than if they wait for data perfection. The contrast highlights a wider debate in business: whether companies should prioritize completing data preparation before deploying AI, or proceed with "good enough" data and improve it as they go.

What to watch: Rhonda Griscti, Bristol Myers' executive director of digital strategy, advocated an 80–20 approach, noting that "one should reach about 80% readiness and then proceed." She emphasized that companies need a data lake (a centralized repository for storing and processing raw data) and the ability to access and clean data as work continues, rather than perfection before launch. Her advice to other executives is to avoid getting bogged down by data concerns, since waiting for perfect data means never getting started.