Industries with historically low software adoption (like trial law or dentistry) are now viable markets. Instead of selling a tool, AI startups are selling an outcome—the automation of a specific labor role. This shifts the value proposition from a software expense to a direct labor cost replacement.

For mature companies struggling with AI inference costs, the solution isn't feature parity. They must develop an AI agent so valuable—one that replaces multiple employees and shows ROI in weeks—that customers will pay a significant premium, thereby financing the high operational costs of AI.

AI enables "software does labor" business models in industries previously deemed too small for specialized software, like dental offices or trial law. By replacing or augmenting specific labor tasks, startups can justify high-value contracts in markets that historically wouldn't pay for traditional SaaS tools.

Confusing credit-based AI pricing models will likely be replaced by a straightforward value proposition: selling AI agents at a fixed price equivalent to the cost of one human worker who can perform the work of ten. This simplifies budgeting and clearly communicates ROI to CFOs.

Companies like Sierra can't justify a 100x ARR valuation by targeting the existing software market (e.g., $8B Service Cloud). The bet is that they will capture a significant portion of the much larger human labor market ($200B+ for support agents). This represents a fundamental transition of spend from human capital to software.

A new generation of AI application companies are being run with extreme leanness and efficiency. They are achieving revenue-per-employee figures between $500K and $5M, dwarfing the public software company average of ~$400K and signaling a fundamental shift in scalable operating models.

Investor Jason Lemkin's thesis for niche B2B software is that AI must enable a massive increase in price to be compelling. If an AI-powered product can eliminate the need for 10 back-office employees, a tool that previously sold for $8,000 a year can now command an $80,000 price tag, transforming its unit economics into something venture-backable.

Unlike SaaS which sells to limited software budgets (e.g., 1% of revenue), vertical AI agents automate core business functions. This allows them to tap into much larger operational and labor budgets. Companies can capture 4-10% of a customer's total spend by replacing expensive human-led tasks like customer support.

AI tools aren't just making employees more efficient; they are replacing human labor. This allows software companies to move from cheap per-seat pricing to a new model based on outcomes, like charging per support ticket resolved, capturing a much larger share of the value.