Structural ยท Commoditized
Services as software is not SaaS
AI-enabled services rollups can become valuable operating businesses, but that does not make them high-margin SaaS or venture-scale software by default.
The category error
A services company that uses AI to lift margins can be a rational buyout or rollup target. The mistake is underwriting it like pure software when revenue still depends on delivery labor, customer-specific work, and operational execution.
Why AI does not erase the services multiple
The model may compress work, but customers still buy outcomes, trust, integration, and accountability. If delivery complexity remains inside the company, the business may deserve PE-style expectations rather than venture-style multiples.
The better test
The right question is whether software-like retention, gross margin, deployment repeatability, and pricing power are actually present. If not, the company is a services operator with AI leverage, not a software platform.
Signals to watch
- FDE-heavy delivery: Revenue expands through forward-deployed engineers or bespoke implementation rather than self-serve software adoption.
- Gross ARR confusion: Headline ARR includes pass-through, services, or delivery labor that should not be valued like net software revenue.
- Margin promise before proof: The pitch assumes AI margin expansion before verified cohort-level margin data exists.
Representative companies
These are examples of the category pressure or survivor pattern, not a claim that the companies have failed.
- UiPath (uipath.com)
- Accenture (accenture.com)
- Harvey (harvey.ai)