Why Turning Services into “AI-Powered Software” Isn't as Easy as VCs Think
Venture capitalists are increasingly treating service businesses—legal, managed IT, consulting—as the next frontier for software-style margins. The blueprint is to acquire established companies, insert AI tools to automate routine tasks, cut costs, and then scale through additional acquisitions. General Catalyst (GC) is a major player here: with a $1.5 billion fund carve-out, it is backing companies that both build new “AI-native” software and use them to absorb existing service firms. This dual approach is supposed to unlock higher profit margins than traditional service businesses typically manage.
Examples in Play
There are real companies already following this model:
- Titan MSP is in GC’s portfolio. After investing, GC helped Titan introduce AI tools across its managed service providers (MSP) segment, and then acquire RFA, a legacy IT services firm. Titan claims 38% of typical MSP tasks can be automated—a big step for margin improvement.
- Eudia, another example, is focused on servicing in-house legal departments (not traditional law firms) via fixed-fee models supported by AI. It has also made acquisitions to expand its reach.
Others like Mayfield are backing similar “AI teammates” strategies, acquiring mature firms, automating key tasks, and hoping margins creep toward the 60-70% gross level.
The Hidden Costs and Practical Hurdles
Transforming service businesses with AI isn’t just about buying the company and flipping a switch. Several issues are complicating the picture:
- “Workslop”: A recent survey of over 1,100 full-time employees showed that a lot of AI-produced content or outputs require follow-up, corrections, or sometimes total redo. That extra friction eats into the gains. One report put the cost of this overhead at about $186 per employee per month. For large organizations, that can amount to millions in lost efficiency.
- Skill gap: Automating tasks reliably demands technical finesse—knowing which AI models to use, when to use them, how to integrate them cleanly into workflows. It's not just about having AI tools; it's about deploying them well. TechCrunch
- Error correction burdens: As AI tools produce flawed or incomplete work, existing teams often have to pivot into correcting or validating outputs. If companies reduce staff too much expecting perfect automation, those gaps can become major bottlenecks. If they keep staffing levels high to manage the “AI mess,” margins may not improve much.
What This Means for VC Models & Expectations
The promise of transforming services into high-margin, AI-driven operations is seductive. But the margins may be thinner and the path rockier than many VCs admit. Some implications:
- Roll-ups (buying multiple service providers) will require more patience; scaling fast may expose flaws in integration, quality control, and talent pipelines.
- Investors may need to expect longer timelines to realize return on investment, especially if initial automation efforts trigger customer backlash or quality dips.
- Companies that adopt AI must balance automation with human oversight, rather than pursuing pure cost cutting. The most successful operators may be those who strategically maintain critical human roles for judgement, oversight, corrections.
General Catalyst itself acknowledges this complexity. According to GC, the know-how of applying AI properly is precisely why they pair AI-specialist talent with domain experts when building companies. The idea is: proprietary expertise + good engineers + solid AI tools = better chance of navigating the transformation successfully.
Final Thoughts
There’s big money, big promise, and high expectations pushing AI-powered service transformations forward. But the real world is more messy than many forecasts suggest. “If it were easy,” one VC observer noted, “we’d already have hundreds of service firms running like Silicon Valley software startups.” What’s clear is that success is not just about automation; it’s about quality, structure, people, and patience.
