Dun & Bradstreet Survey of 10,000 Businesses Finds AI Returns Stable for Most, but Data-Ready Companies Pull Ahead on ROI

Dun & Bradstreet's Q4 survey of 10,000 businesses found 28% report broad or strong AI ROI, unchanged from Q3, while 65% plan to increase AI spending and only 8% say their data is fully ready for AI at scale. Broad or strong ROI was reported by 81% of businesses with fully AI-ready data versus 12% with partially ready data; the findings point to data readiness as a barrier to scaling returns.

Analysis

The investable implication is a shift in AI spending mix, not evidence that enterprise AI economics have suddenly improved. If data quality and integration are the constraint, budgets may migrate from model experimentation toward data engineering, master-data management, entity resolution, governance and implementation services. That could favor vendors with embedded enterprise workflows and verified data assets, while pressuring generic AI applications whose differentiation rests mainly on model access. Better context may also improve successful task completion while reducing wasted inference and rework—positive for AI adoption, but not automatically for token or compute intensity.

Treat the survey’s ROI gap as directional, not causal: readiness and returns are self-reported, and better-managed or better-funded firms may score higher on both. The release is also promotional material from a data vendor; its product claims need independent validation. In the near term, this is a weak standalone catalyst. Over 1–3 months, watch earnings commentary for data-preparation and governance demand converting into bookings, and for AI pilots becoming production workloads. Over 6–18 months, a durable shift would appear in recurring data-platform revenue and customer-reported payback, not AI-spending intentions alone.

Contrarian angle: the market may overpay for “data readiness” as a theme before buyers demonstrate willingness to fund it separately. If integration remains labor-intensive, services firms may capture more value than software vendors. Conversely, falling model costs and better tooling could make readiness less of a bottleneck than the survey implies.

Key Decisions for Investors

• No trade on this survey alone. Treat it as a watch item for data-management, integration and governance software and implementation services; the article provides no company-level revenue evidence or independently verified ROI.
• Over the next 1–3 months, favor companies only where earnings or bookings show paid demand for data quality, identity resolution, governance or production integration—not merely AI product announcements. Avoid assigning a ticker based on this release; the supplied company-identity data contains none.
• Potential relative-value expression, conditional on valuation and earnings confirmation: long a diversified data-infrastructure/software basket against a basket of undifferentiated AI application vendors. Keep exposure small until relative bookings or guidance validate the budget shift; do not infer a specific pair from the survey.
• Falsify the thesis if upcoming results show data-platform bookings or implementation demand slowing while AI deployments scale, or if buyers report that model/tool improvements materially reduce data-preparation costs. Also reassess if AI spending plans are cut or production deployments fail to convert into measurable customer payback.