Enterprise technology spending is projected to hit $4.25 trillion in 2026, largely driven by AI adoption, but the business model underpinning that boom is showing signs of structural fragility. New research from venture capital firm Madrona reveals that 77% of enterprises re-evaluate their AI vendors every six months or on a rolling basis, a finding that challenges the assumption that landing an enterprise contract secures predictable, long-term revenue for startups.
The survey of 150 enterprise IT professionals found that while 74% plan to expand their AI budgets over the next 12 months, commitment remains shallow. Fewer than half of AI pilots currently make it into full production—a low bar, though notably better than MIT’s widely cited 2025 finding that 95% of enterprise AI projects failed to deliver a return on investment.
Also read: Sequoia-backed Empirik raises $21M to predict IT outages before they happen
The end of the SaaS contract moat
For decades, enterprise software startups relied on a simple formula: win a multi-year contract, then bank predictable annual recurring revenue (ARR) as switching costs kept customers locked in. Madrona’s data suggests that dynamic has inverted for AI products. The report describes a “fast in, fast out” environment where switching costs are lower and the re-evaluation cadence is relentless.
This shift has profound implications for the AI startup funding space. The initial AI boom of 2025 was fueled by enterprise trial budgets, and 2026 was expected to be the year those experiments converted into sticky, long-term commitments. Instead, even successful pilots are subject to constant scrutiny, meaning the astronomically fast revenue growth many AI startups report—going from $0 to $10 million in three months—may be far less durable than it appears.
Also read: NYC mayor bans generative AI for students below 9th grade in nation's largest school district
Pricing models caught in transition
Part of the instability stems from unresolved pricing questions. New research from venture capital firm Andreessen Horowitz, based on a survey of 50 technical AI buyers, found that more than half want fees tied to tangible work outcomes—such as reports processed, tickets closed, or leads generated—rather than usage metrics like token consumption.
Usage-based pricing is essentially a SaaS-era holdover. When a company buys email or HR software, the value is tied to headcount or data volume. AI products, by contrast, are judged on the quality and volume of work they produce. a16z partners Tugce Erten and Sarah Wang argue that outcome-based pricing makes the product “economically valuable to both sides,” aligning vendor incentives with customer results.
However, this shift creates a double-edged sword for startups. Outcome-based pricing can justify higher fees when the AI performs well, but it also ties revenue directly to performance metrics that enterprises can measure and question on a quarterly basis.
What this means for the AI startup economy
The Madrona and a16z findings collectively paint a picture of an enterprise market that is more willing to experiment but less willing to commit. For startups, this lowers the barrier to entry—enterprises are more open to trying new AI tools than they were with traditional software. But it also means that an enterprise contract is no longer a reliable foundation for long-term financial modeling.
Investors who have valued AI startups on the assumption that rapid ARR growth would eventually stabilize into recurring revenue may need to adjust their frameworks. The risk profile of an AI startup with $50 million in ARR is now closer to that of a services business with a strong pipeline than to a traditional software company with locked-in contracts.
The open question is whether this experimentation-heavy dynamic will persist or whether enterprises will eventually settle into longer procurement cycles once the AI market matures and consolidates. For now, founders should assume that retaining enterprise customers will require continuous proof of value, not just a signed contract.
This article discusses market research and business trends; it does not constitute financial advice or an investment recommendation. The venture capital and startup markets are volatile and uncertain.
