Prologis CEO says warehouse demands are shifting fast. Here's how he's keeping up

A Prologis-Harris Poll survey found 91% of business leaders believe retailers and logistics providers urgently need more decentralized fulfillment networks; the share planning to retain their existing global footprint fell 14 percentage points year over year to 36%. Companies are prioritizing faster delivery, regionalized networks and reliable power, while three-quarters of respondents plan major investment in AI demand forecasting and inventory replenishment. Prologis CEO Dan Letter said the company generates more than 1.4 gigawatts from solar and storage and has sold over 200 million square feet of property that did not fit customers’ long-term needs.

Analysis

The investable signal is not simply “more warehouses”: it is a potential widening gap between well-located, power-capable infill assets and generic industrial space. If customers pay for delivery speed and resilience, scarcity value could accrue to landlords with the right sites and credible power access; PLD’s scale may help it match customer networks to those locations. But the survey is commissioned by PLD with an obvious strategic interest, and its stated energy capacity does not establish returns, utilization, or incremental earnings.

The counterweight is AI efficiency. Better forecasting and replenishment may reduce buffer inventory and total space per unit of sales, even as decentralized networks require more locations. Net demand depends on whether node proliferation outweighs inventory compression—an outcome the survey does not quantify. Power availability is also not equivalent to deliverable capacity: interconnection delays, grid limits, and storage economics could make energy investment capital-intensive rather than accretive. Less specialized industrial landlords may face higher leasing and retrofit costs, while other logistics REITs with dense urban footprints could compete for the same tenants.

Near term, treat the survey as sentiment, not an earnings catalyst. Over 1–3 months, monitor PLD leasing spreads, occupancy, development starts and customer commitments to powered sites. Over 6–18 months, the thesis requires demonstrable rent premiums and returns on energy capex. Falsification: management signals weaker leasing/pricing, powered-site commitments fail to convert, or energy investment rises without improving returns. A plausible contrarian risk is that the market capitalizes “AI + power” before either produces measurable cash flow.

Key Decisions for Investors

• No trade on the survey alone. Keep PLD on a relative-value watchlist; consider a modest long PLD versus less urban-focused industrial landlords only if upcoming leasing data show persistent rent/occupancy outperformance and the valuation premium remains reasonable.
• Over the next 1–3 months, track PLD’s leasing spreads, occupancy, development pipeline and disclosed customer demand for powered locations. Verify the economics behind the company’s solar/storage capacity: capex, utilization, contracted revenue and returns—not just gigawatts installed.
• Treat AI as two-sided: faster delivery may increase demand for distributed sites, while forecasting can reduce inventory and warehouse space per dollar of sales. Reassess the thesis if customer space commitments weaken despite continued AI investment.
• Avoid extrapolating survey intentions into earnings. Revisit exposure if grid interconnection constraints delay projects, energy capex expands without improving returns, or management reports weaker leasing/pricing.