SaaS doomsday approaching? In the age of AI, software contracts are quietly getting shorter.

The rapid iteration of artificial intelligence is fundamentally changing the logic behind enterprise software procurement. Customers are no longer willing to sign multi-year contracts with vendors, but are instead seeking shorter, more flexible agreements—a trend that is shaking the SaaS business model that Silicon Valley relies upon.

On March 25, according to Bloomberg, this transformation has already begun to take place at the contract level. From healthcare to legal tech, enterprise buyers are generally demanding that contract periods be compressed from three or five years to one year or even month-to-month renewals, in order to retain the flexibility to switch vendors at any time. Julie Yoo, a healthcare investor at Andreessen Horowitz, said that some hospitals are running pilot programs with multiple AI startups simultaneously, assessing products through a "competition" approach: "Nobody wants to tie themselves to one dance partner for long anymore—what if they pick the wrong one?"

This shift has already had a direct impact on the capital markets. ETFs tracking the software sector continue to be under valuation pressure, and the so-called "SaaSpocalypse" has caused the market value of many publicly listed software companies to shrink significantly. This week, Amazon announced that its cloud computing division is developing AI agents capable of automating sales and other business functions. On the day the announcement was made, software-tracking ETFs dropped over 4% in a single day, with some stocks falling nearly 9%.

Pressure on SaaS Model: Per-user Pricing Logic Challenged

Over the past decade, selling software services to enterprises has been Silicon Valley’s core path to predictable revenue and high profit margins. The subscription model based on user counts underpins the valuation system of many industry giants.

However, the proliferation of AI tools is eroding this logic’s foundation. The industry widely worries that when AI can replace or dramatically reduce manual operations, pricing models based on "headcount" will face structural shrinkage. Meanwhile, changes at the contract level are arriving ahead of commercial model restructuring, with customers refusing to make long-term commitments to any party while it is still unclear which vendors will survive.

Lisa Singer, vice president and principal analyst at Forrester, said bluntly: "I tell enterprises, you shouldn’t sign a three-year contract that includes AI product pricing. Three years from now, who knows what the cost of AI will be—it’s impossible to judge at this point."

Contract Cycle Shrinks: From Multi-year to Month-to-month Renewals

This change is especially evident in verticals like healthcare and legal. Julie Yoo points out that startups selling software to hospitals used to aim for multi-year contracts, but now many have shortened agreements to a year, and some are even moving to monthly renewals to provide maximum flexibility to clients.

The legal tech field is also under pressure. Filevine, a 12-year-old legal tech company, previously built a competitive moat by selling AI-enhanced system-level software to law firms. Now it faces direct competition from unicorn startups like Harvey and Legora. CEO Ryan Anderson said the company is preparing to launch an AI agent layer that allows lawyers to directly call existing software systems, aiming to speed up new client adoption and avoid losing bids due to lengthy onboarding cycles.

Consumerization Trend: Enterprise Procurement Logic Shifts Toward Personal Consumerization

This transformation is also challenging a long-held VC assumption—that selling to enterprises is far more stable than selling to individual consumers. In the early 2020s, startups directly targeting consumers lost investor favor due to high customer acquisition costs and thin profit margins. However, enterprise-side transactions are increasingly showing consumer product characteristics: rising churn rates, increased income volatility, and weakening product moats.

Derek Xiao, head of Menlo Ventures, points out that many enterprise AI contracts actually originate from individual consumers. Employees develop applications using AI tools like Lovable or Anthropic’s Claude over the weekend, then bring them into the workplace on Monday, ultimately driving employers to incorporate these tools into enterprise systems. Menlo Ventures is an investor in Lovable and Anthropic. Meanwhile, sales cycles for leading AI companies have been compressed from several months to just a few weeks.

Capital Repositioning: Betting on AI Startups

Despite the pressure on existing software companies, capital is rapidly gathering around AI-native firms. Veteran venture capital firm Kleiner Perkins is raising a new $3.5 billion fund, including a $1 billion allocation for its 22nd early-stage fund focused on high-potential AI startups. Kleiner Perkins is known for early investments in Google and Amazon.

This strategy reflects the market’s overall judgment: The process of AI reshaping the software industry is irreversible. The question is not whether the change will happen, but which emerging companies will establish new business orders on the ruins of the old model.

Risk Warning and DisclaimerThe market carries risks. Investments require caution. This article does not constitute individual investment advice, nor does it take into account special investment goals, financial situations, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article suit their specific circumstances. Investing based on this content is your own responsibility.