P&G targets $200M–$300M in savings with AI-powered dynamic routing and sourcing optimization

P&G, the world's second-largest publicly owned consumer goods company with operations spanning dozens of countries and product categories, saw its supply chain productivity fall below pre-pandemic benchmarks. Logistics inefficiencies compounded across the network: truck scheduling gaps were generating driver idle time, fill rates were underperforming, and sourcing decisions lacked real-time data support. Critically, optimization was fragmented — teams addressed isolated segments of the logistics network rather than working across the full chain, including the retailer layer. Without a unified ML infrastructure, savings opportunities remained invisible at scale, and the company lacked the tooling to quantify or close them systematically.

P&G responded with its global Supply Chain 3.0 initiative, deploying machine learning and predictive analytics across end-to-end logistics operations. The centerpiece was an internal 'AI Factory' — a centralized ML platform built under CIO Vittorio Cretella's direction and deployed across 80% of P&G's global business. Rather than siloed pilots, the platform was architected for enterprise-wide reuse, allowing data science teams to develop and deploy models rapidly without rebuilding infrastructure per project. Specific applications include ML-driven truck scheduling to eliminate driver idle time, AI-powered dynamic routing for carrier and lane selection, sourcing optimization to identify cost-efficient supply inputs, and fill rate tools that align supply availability with downstream retail demand. The initiative was designed from the outset to extend beyond P&G's internal operations, enabling joint optimization with retail partners across the full supply network.

P&G anticipates $200M–$300M in cumulative cost savings from AI-driven routing, sourcing, and fill rate improvements — figures cited by CEO Jon Moeller directly in an investor earnings call. The AI Factory has delivered a 10x improvement in data scientist speed and efficiency, compressing model development cycles that previously took months. With 80% global business coverage, impact is systemic rather than localized. Key outcomes include:

• Routing efficiency: Reduced driver idle time through ML-optimized truck scheduling
• Fill rate improvement: Tighter supply-demand alignment with retail partners
• Collaborative supply chain: Retailers now participate in network-wide optimization alongside P&G rather than managing demand signals in isolation

• Centralized ML infrastructure before individual use cases is the critical enabler — P&G's AI Factory delivered 10x data science productivity by giving teams a shared foundation rather than rebuilding per project.
• Extending AI optimization to retail partners, not just internal logistics, unlocks network-wide savings that siloed approaches cannot reach.
• Dynamic routing and sourcing work best as integrated capabilities — the $200M–$300M savings target reflects the compounding effect of optimizing across multiple logistics dimensions simultaneously.
• CPG enterprises should sequence AI investments around the highest-idle-cost nodes first — truck scheduling and fill rate improvements offer near-term, measurable ROI that builds internal confidence for broader platform rollout.