A successful AI pilot doesn’t mean your AI strategy is working

McKinsey’s global 2026 survey found that 37% of respondents said AI had improved their company’s earnings, unchanged from the previous year. Eight in ten said AI had helped their companies become more [text is truncated], highlighting a gap between reported operational benefits and earnings impact.

Analysis

The investable implication is a likely shift in AI spending mix, not a collapse in adoption. If deployment remains difficult, budgets may migrate from model and compute providers toward systems integration, data cleanup, workflow redesign, and governance. That can benefit implementation-heavy service providers and established software vendors with distribution and embedded customer workflows, while challenging AI application vendors whose pricing assumes rapid, provable productivity gains. The offset: integration-heavy projects can be slower and less repeatable, so services revenue may rise without software-like margin expansion.

Near term (days to weeks), the survey is a modest check on “AI spend converts quickly to earnings” positioning, especially in high-multiple names priced on monetization rather than usage. Over 1–3 months, watch earnings commentary for AI-related bookings converting into renewals, paid seats, or measurable productivity—and whether cloud providers’ capex is matched by monetizable workloads. Over 6–18 months, the durable winners should be those controlling workflow distribution and customer data, not necessarily those supplying the most model capacity.

Contrarian point: a flat share reporting earnings improvement is not evidence that AI has no value; self-reported, broad-company results may lag deployment, and benefits can accrue first as avoided hiring or capacity growth rather than reported earnings. The survey is not enough to identify a security-level trade or infer company-specific economics.

Key Decisions for Investors

• Do not add broad AI exposure solely on adoption or usage indicators; require evidence of paid conversion, renewal uplift, or margin/productivity gains in company disclosures.
• Treat high-valuation AI infrastructure and application exposures as more vulnerable to earnings-conversion disappointment over the next 1–3 months; avoid initiating a sector-wide short from this survey alone.
• Watch for relative opportunities in implementation services and incumbent workflow software only where reported AI-related bookings convert to recurring revenue; verify backlog quality, delivery capacity, and margins before positioning.
• Falsify the cautious monetization thesis if upcoming results show sustained AI-driven paid-seat growth, renewal uplift, or productivity-led margin expansion; strengthen it if capex rises while monetizable AI revenue or guidance fails to follow.