Sequoia Capital's early-stage investing thesis suggests the next trillion-dollar company won't sell software or hardware as a product at all. Instead, it will sell outcomes, using AI-powered software alongside human expertise to deliver results. This represents a fundamental shift in how venture capitalists view the future of AI-driven business models.

The idea comes from Julien Bek, an early-stage investor in Sequoia's London office, who published a viral essay titled "Services: The New Software" that garnered nearly 3 million views on X and over 450,000 impressions on LinkedIn. Bek's thesis challenges the conventional wisdom that software businesses represent the highest-margin, most scalable path to massive valuations. Instead, he argues that AI-native service companies, built from the ground up to leverage artificial intelligence, will capture enormous value by delivering professional services more efficiently than traditional firms.

What Makes AI-Native Services Different From Traditional Software?

The core insight behind Bek's argument rests on a simple economic observation: for every dollar enterprises spend on software, they spend six dollars on services. This massive market gap exists because many critical business functions, from legal work to accounting to insurance brokerage, have traditionally been delivered by human professionals rather than automated software tools.

Bek distinguishes between two types of work that matter for this analysis. Intelligence refers to tasks with clear right and wrong answers, like coding, mathematics, or accounting calculations. Judgment, by contrast, involves taste, professional intuition, and subtle qualitative distinctions that require talent and experience. AI models are becoming proficient at intelligence tasks, but judgment remains largely in the human domain.

The sweet spot for AI-native service companies, according to Bek, lies in services that are already outsourced by enterprises and require mostly intelligence with just a dash of human judgment. These include:

• Insurance Services: Insurance brokerages, claims adjustment, and underwriting where AI can handle routine analysis while humans manage complex cases
• Professional Services: Tax advisory, accounting, audit services, and simple legal work where AI can automate routine processes
• IT and Compliance: Managed IT services and certain compliance functions where AI can monitor systems and flag issues for human review
• Payroll and Administration: Payroll processing and administrative tasks that are highly standardized and rule-based

Bek calls startups targeting these categories "autopilots," though he clarifies this doesn't mean fully autonomous AI agents replacing human experts entirely. Rather, it means the processes delivering these services can be largely automated, similar to how airplane autopilots function. A human expert remains present, monitoring systems and handling the hardest tasks, ready to intervene if something goes wrong.

How Can AI-Native Services Compete Against Established Firms?

The competitive advantage for AI-native service companies comes from multiple angles. First, they can undercut traditional firms on price. If a customer currently pays $100 for a service, an AI-native startup might offer the same service for $80 while maintaining healthy profit margins through AI-driven efficiency.

Beyond pricing, these companies can fundamentally change how they charge for services. Traditional professional services firms bill by the hour, a model that has persisted for decades despite widespread criticism. AI-native firms can instead bill by outcome, which changes the economics entirely. This shift is particularly powerful for smaller companies competing against larger incumbents, as it allows them to disrupt on pricing while offering better alignment with customer success.

"When you're a smaller company, the best thing you can do to compete with the larger ones is actually disrupt them on pricing," said Julien Bek, early-stage investor at Sequoia Capital.

Julien Bek, Early-Stage Investor at Sequoia Capital

However, Bek acknowledges that changing customer expectations around billing takes time. The billable hour in legal services, for example, has been criticized for decades, yet it remains the dominant model at most corporate law firms. That said, AI appears to be accelerating the shift away from hourly billing, particularly in legal services where AI tools can dramatically reduce the time required for routine work.

What About Profit Margins and Scalability?

One concern investors raise about service-based businesses is that they don't scale like software. Software has near-zero marginal costs once built, while services require human labor. However, Bek argues the efficiency gains from AI change this equation significantly. In insurance brokering, for instance, AI-native startups can enable each human expert to sell 10 times more business than traditional insurance brokers, dramatically improving the leverage per employee.

This efficiency gain suggests that while AI-native service companies won't match pure software margins, they can still achieve attractive unit economics. The key is that AI handles the high-volume, routine work, freeing human experts to focus on higher-value judgment calls and relationship management.

Two costs present ongoing challenges for this model. First, AI inference costs, the expense of running AI models to process customer requests, can consume a substantial portion of revenue in some cases. Second, go-to-market costs for selling services tend to be higher than for software, since services require more direct customer relationships and ongoing support.

Why Won't Companies Just Build These Services Themselves?

A natural question is whether enterprises will simply use AI tools to build these services in-house rather than outsourcing to AI-native startups. Bek acknowledges this may happen for some functions, but argues many services will remain outsourced for regulatory or practical reasons. Financial auditing, for example, requires independent third-party firms by law. Management consulting persists partly because it provides external validation for decisions management already wanted to make, offering cover if decisions turn out poorly.

Similar dynamics apply to other professional services. Companies may lack the specialized talent to build certain capabilities in-house, or they may prefer the risk transfer that comes with outsourcing to a specialized firm. These "softer" reasons for outsourcing are often as important as pure economics.

Real-world examples already demonstrate this model working. Companies like Robin AI and Legora in legal services, Dwelly in real estate, Dystyl AI in consulting, Rogo in financial services, and WithCoverage in insurance brokerage are all pursuing the AI-native service model. Their success suggests Bek's thesis has merit, though the model is still in early stages.

The broader funding environment supports this shift. Sequoia's backing of this thesis, combined with substantial capital flowing into AI infrastructure and embodied AI systems, signals that venture capitalists see genuine opportunity in companies that combine AI intelligence with human judgment to deliver better outcomes at lower cost. As AI capabilities improve and inference costs decline, the economics of AI-native services will only improve, potentially creating the next generation of massive, high-margin businesses.