If your agents can deploy a feature before your team has decided whether it's worth building, you have a sequencing problem that no amount of speed will fix. Cagan's framework is the antidote you need before your backlog becomes a liability.
Inspired builds its argument around a deceptively simple claim: most product teams are delivery teams in disguise, burning capacity shipping things stakeholders requested rather than solving problems customers actually have. Marty Cagan insists that valuable, usable, feasible, and viable are four constraints that must be tested together, continuously, before a single line of production code gets written. The work of discovery is not a phase — it is a permanent parallel track, and skipping it is how you build fast toward the wrong destination.
Agents collapse feasibility almost to zero, which is exactly where the danger hides. When building a thing costs almost nothing, the discipline of asking whether it should be built at all has to get louder, not quieter. The viable constraint — will the business actually sustain this — now needs a human owner who is explicitly accountable, because no agent will volunteer that a feature undermines your pricing model. Cagan's empowered-team structure becomes the governance layer you design on purpose, or the one you discover you needed after the fact.
- Treat discovery as infrastructure, not ceremony — agents accelerate delivery but cannot substitute for it
- assign explicit human ownership of the viability constraint before the first agent task is written, not after
- measure your team by outcomes shipped, and audit whether your agent workflows are even pointed at an outcome someone tested first.
