Is there evidence for a SaaS that runs a whole organization?

People pay for AI that does one business job, but the sources reviewed did not establish demand for one product that builds and runs any organization.

The numbers

  • more than $150 million

    annual recurring revenue at Sierra, February 2026

    The company's own figure, not audited

  • 40%

    of firms above $1 billion in revenue scale agents in at least one function

    Against 22% of smaller ones. McKinsey, global

  • 32%

    of respondents say their organization passed on at least one software product or feature because coding agents let them build it

    McKinsey, same survey, confirmed through secondary coverage

Two counts of AI use, two populations

  • of businesses on Ramp paid for AI in September 202655.9%
  • of US businesses used AI in May 202619.8%

Ramp counts payments and the Census counts use, in different populations and months, so the two should not be averaged.

What leaders trust agents with

  • of leaders trust agents with data analysis38%
  • trust them with autonomous interactions with employees22%
  • trust them with financial transactions20%

Shares of 308 senior US business leaders, PwC, 2025.

I asked for evidence for an idea I find attractive: one software product that sets up and operates any kind of organization with AI. I asked the tool to be honest if the evidence was not there.

I found buyers for narrower products: customer service agents, AI teams for small businesses and tools for launching a business. That does not establish demand for the whole idea.

What I found

  • Demand for AI that does a defined business task is strong. Sierra, which sells customer service agents, reported more than $150 million in annual recurring revenue in February 2026. The figure is the company’s own and not audited.
  • Demand for an “AI team” for small businesses is credible. Sintra was reported to have more than 40,000 paying customers and $12 million in annual recurring revenue when it announced its seed round in June 2025. The press coverage names no source for the figures, so treat them as the company’s claim. The study argues that a much dearer product has to prove it does a much more valuable job.
  • Demand for AI that helps people launch a business is real. Lovable says nearly eight in ten of its users are building a business or side project they hope to earn from, and more than a third of those already earn. The post gives no sample size or survey date. Lovable also says it wants to become a place to build and run a business, so it is a competitor too.
  • Payment data shows AI is bought widely, but the picture depends on who is counted. The Ramp AI Index, built from payments by more than 70,000 businesses using Ramp, had 55.9% paying for AI in September 2026. The US Census Bureau’s survey put use at 19.8% of US businesses in May 2026, and between 17% and 20% from December 2025 to May 2026. They measure different populations and should not be averaged.
  • Large companies report more scaling of agents. McKinsey’s 2026 global survey (1,719 respondents) found 40% of firms above $1 billion in revenue scaling agents in at least one function, against 22% of smaller ones. About 6% counted as high performers.
  • Stated trust varies by task. In PwC’s 2025 survey of 308 senior US business leaders, 38% trusted AI agents with data analysis, 22% with autonomous interactions with employees and 20% with financial transactions.
  • Custom building replaces some purchases. In the same McKinsey survey, 32% of respondents said their organizations had decided against buying at least one software product or feature because coding agents let them build it themselves.

The sources reviewed did not establish demand for a product that builds and runs any organization. The better first test is one existing business with a recurring bottleneck, one buyer and one paid outcome, with a human approving what matters. No customer interviews or willingness-to-pay tests were part of this study.

How it was made

I asked Perplexity Deep Research on 24 October 2026 and it showed 384 sources. The AI gathered the material and drafted the study from secondary research, mostly about US businesses, though the McKinsey figures are global. I asked the question. Several figures come from the companies themselves. Figures were current at that date.

Sources

The question
What evidence is there that buyers want one software product that builds and runs an organization on its own, and where does the evidence stop?
How it was made
Perplexity Deep Research, 384 sources.