Products Are Functions
Ryan Singer argues that every product is, at its core, a function: a mapping from a user's current state to a desired one.
The frame sounds abstract until you try to apply it — then it starts to dissolve the usual confusion between features and outcomes.
A feature is a mechanism; the product is the function.
Singer's writing has the same spare precision as his Shape Up work: no jargon, no ceremony, only the cleanest statement of the idea that can be made.
Useful for anyone who has ever sat in a roadmap meeting and sensed that the participants were all talking about different things.
Central argument
Singer argues that a product is best understood as a function in the mathematical sense: it takes a user from a current state to a desired state, and every design decision should be evaluated against how well it performs that transformation. The distinction he draws is between the function (the product) and its mechanisms (the features) — a clarification that reframes what teams are actually building and why. This abstraction is not merely conceptual housekeeping; it is a diagnostic tool for identifying when a roadmap has drifted from outcomes into feature accumulation.
Critique
The functional frame is analytically powerful but can obscure the degree to which desired states are not discovered but constructed — products shape what users come to want, and the input/output model risks underweighting that generative, market-shaping dimension of product work. A CPO operating in a nascent category may find the framework too static: it assumes a legible transformation to optimize toward, which is precisely what is often absent in early-stage or platform contexts where the function itself is being defined through iteration. The elegance of the metaphor may also invite false precision, encouraging teams to specify user states with more confidence than the evidence warrants.
Why it matters for product
For a product leader running roadmap or prioritization processes, Singer's frame offers a concrete intervention: any proposed feature should be challenged to articulate which part of the state transformation it advances, and if it cannot, it is a candidate for deferral or removal. This directly addresses the organizational failure mode where engineering, design, and commercial teams attend the same roadmap meeting but are implicitly optimizing for different things — mechanisms, aesthetics, and outputs respectively — because no one has agreed on the function the product is meant to perform. It also sharpens the connection between discovery work and delivery: discovery becomes the act of specifying the function more accurately, not merely generating ideas.