Enterprises are still grappling with measuring concrete returns from AI investments despite earlier hype
Executive summary: NEA partner Tiffany Luck said enterprises are still figuring out how to measure ROI from AI, citing early over‑hype like tokenmaxxing and budget overruns at companies such as Uber. Mis‑aligned expectations could lead to wasted spending, pressure on AI vendors, and potential regulatory scrutiny of AI investment disclosures.
Who is involved: NEA’s Tiffany Luck; enterprise CIOs and AI vendors; companies including Uber that exhausted AI budgets.
Likely next: Enterprises are likely to pause large AI spend until robust ROI frameworks emerge, while regulators may increase oversight of AI budgeting practices.
Firms launched aggressive AI pilots in 2024, encouraging widespread usage, but early results show limited measurable profit gains. Recent statements from venture investors indicate budgets are being exhausted faster than anticipated, prompting tighter scrutiny. Analysts expect capital allocation to shift toward ROI‑focused projects before further spending resumes.
What's next — scenarios
The ROI Correction (Base Case) (50%)
Enterprise software budgets shift from general-purpose LLM experimentation to specialized, high-utility vertical AI tools.
- Reduction in seat-based enterprise AI licensing renewals
- Increased demand for 'agentic' workflow automation tools
The Efficiency Plateau (Downside) (30%)
Corporate capital expenditure on AI hits a multi-quarter freeze as investors demand proof of margin expansion.
- Downsizing of internal AI task forces
- Reported depletion of venture-backed AI pilot budgets
The Productivity Breakthrough (Upside) (20%)
Early adopters unlock significant labor-cost savings, triggering a second wave of massive capital reallocation.
- Public earnings calls citing AI-driven operational margin improvements
- Rapid scaling of successful pilot programs to company-wide deployment
What to watch
- Q3 and Q4 enterprise software spending reports for signs of 'pilot fatigue'
- Earnings calls from major cloud providers regarding AI-driven revenue growth vs. infrastructure cost
- Venture capital funding shifts from foundational models to application-layer ROI tools by end of year
Timeline
- — NEA’s Tiffany Luck says enterprises are still figuring out their AI ROI (TechCrunch)
Analysis — what this means
Likely next events
- Delayed AI project approvals
- Reallocation of AI budgets toward cost‑saving initiatives
- Increased regulator queries on AI spend
- Potential announcements of AI ROI benchmarks
Sectors affected
- Technology
- Finance
- Venture Capital
Regulatory implications
- Greater scrutiny of AI expense reporting
- Possible tax considerations for AI R&D
Historical parallels
- Dot‑com bubble excesses in 2000
- Early 2000s telecom over‑investment