New Survey from Collibra by The Harris Poll Finds 72% of Tech Decision-Makers Feel AI Initiatives Today Are Falling Short

Collibra's Harris Poll of 306 U.S. enterprise data, privacy and AI decision-makers found that 76% encountered critical barriers moving autonomous AI agents from pilot to production, while 87% regularly re-verify agent context. Poor or unaligned data foundations were identified as the root cause of failed AI initiatives by 72% of respondents, rising to 96% at companies with at least $100 million in revenue. Organizations are responding by tightening AI accountability and governance: 90% are preparing for evolving AI regulations, while 51% are investing in data lineage and documentation.

Analysis

This is a demand-validation datapoint for the data-governance/control-plane stack, not yet an earnings catalyst for public research vendors. The bottleneck shifts enterprise AI spend away from incremental foundation-model experimentation and toward metadata, lineage, identity, evaluation, observability, and workflow controls; that favors platform vendors with embedded governance assets over point-model providers whose ROI depends on autonomous deployment volumes. The more important second-order effect is a longer sales cycle: buyers will consolidate AI governance under data-office budgets, increasing procurement scrutiny but raising switching costs once standards are embedded.

FORR and IT should not be treated as direct beneficiaries merely because governance becomes more prominent. Their potential upside is indirect—higher demand for CIO advisory, vendor-selection research, and AI-risk consulting—but subscription/research revenue only responds if enterprise budget uncertainty translates into recurring advisory engagement rather than internal governance hiring or systems-integrator spend. Near-term, the cited survey is company-sponsored and small, so it is insufficient to alter estimates or justify a directional position.

Over the next 1-3 months, watch earnings commentary from ServiceNow (NOW), Microsoft (MSFT), Salesforce (CRM), IBM (IBM), Informatica (INFA), and cyber/data-control vendors for attach-rate evidence: paid governance modules, AI workflow deployments, professional-services utilization, and reductions in pilot-to-production timelines. Over 6-18 months, a tougher regulatory and liability environment could create a durable governance spend category, but it may also cap agent adoption and delay the consumption growth embedded in optimistic AI-software valuations. The thesis is falsified if enterprises report production scaling without incremental governance spend, or if vendors show AI bookings but no material services/implementation burden.

Contrarian view: investors may assume governance friction is uniformly bullish for software vendors. In practice, fragmented controls can become an integration-tax problem, favoring hyperscalers and horizontal workflow platforms that bundle identity, data, monitoring, and model controls; standalone governance vendors risk being feature-compressed unless they demonstrate cross-cloud interoperability and measurable labor savings.

Key Decisions for Investors

• No immediate directional trade in FORR or IT: require next-quarter evidence of AI advisory demand in bookings, renewal rates, or management commentary before underwriting an earnings impact.
• Create a 1-3 month watchlist for long NOW/MSFT versus short a basket of high-multiple, agent-exposure software names with limited governance monetization; enter only after two or more vendors disclose paid governance attach rates or elevated AI implementation services. Target 2:1 reward/risk; exit if production-agent adoption accelerates without control-layer spend.
• Monitor INFA for data-lineage and data-quality booking acceleration as a cleaner public proxy for the underlying bottleneck. A long is actionable only if management quantifies AI-driven ARR/bookings and maintains margin guidance; otherwise implementation intensity may offset revenue upside.
• For broad AI software exposure, reduce reliance on pilot-to-production assumptions over the next 6-12 months. Hedge through selective underweights in companies where valuation depends on rapid autonomous-agent consumption but disclosures lack production deployment, governance, or customer-ROI metrics.