If you're staffing your company with agents rather than headcount, you're not automating a business — you're probably building one that incumbents structurally cannot copy. That distinction is worth thinking through carefully before you price anything.
The Innovator's Solution, Clayton Christensen's follow-up to his disruption diagnosis, is a builder's manual. The core move: find people doing nothing — not using a worse competitor, just going without — and serve them with something simple enough that they'll start. Pair that with jobs-to-be-done thinking, where you stop asking what customers want and start asking what situation they're stuck in and what outcome they're trying to make happen. Then hold integration tight in any dimension where performance still falls short of good enough.
The non-consumption insight hits differently now. Whole categories of work — contract review for a small landlord, competitive analysis for a solo founder, customer research for a two-person team — were never not-done because people lacked motivation. They were not-done because hiring a human to do them once was absurd. Agent-native products can go after that population directly, and incumbents staffed with humans literally cannot match the unit economics to follow. Where Christensen warned against modularizing too early, the same logic applies to your agent stack: wherever quality still varies unpredictably, keep the orchestration in-house and resist the urge to hand it to a generic platform.
- Compete against non-consumption first — your real market is work that was never getting done, not work being done badly elsewhere
- apply jobs-to-be-done framing to the agent interaction itself, not just the end deliverable
- stay integrated at every layer where your agents still produce unreliable output, because that's where your defensibility actually lives.
