Stories as Survivable Experiments

Definition

A user story is a bet on an assumption about business value, and its size should be set by how much a stakeholder is willing to invest to learn whether that assumption holds — not by whether the work fits neatly into an iteration. Drawing on Tim Harford's Adapt, the book reframes small stories as "survivable experiments": small not because a timebox demands it, but because "the world shouldn't end just because a story turns out to be wrong."

In the Book

Chapter 2 argues that describing stories as "small chunks of work that fit in an iteration" causes teams to optimize for size over value, producing disconnected technical stories nobody can deploy or get feedback on. The book's worked example is a mobile-engagement initiative: rather than building a full mobile app (a huge, unverified bet), the team isolates the underlying assumption — that a mobile-optimized homepage keeps users engaged longer than a desktop-optimized one — and tests it cheaply, first by routing a small group of mobile users to a mobile homepage, then narrowing further to a single city, then to a hand-crafted static page for that city if even that is too costly. Each narrowing step is a smaller, survivable experiment that still produces a real, deployable answer; a follow-up story can then extend the winning slice (more cities) or reduce its cost (automating the content) — work that would be needed for the final product regardless of the experiment's outcome.

Why It Matters

This reframes story splitting: instead of asking "how do I cut this to fit two weeks," ask "what's the smallest bet that would tell us if the underlying assumption is right, and what can we survive losing if it's wrong." It converts iterative delivery's value proposition from a scheduling technique into a risk-management one, useful anywhere plans rest on unverified assumptions about how people or systems will actually respond — product bets, policy pilots, or organizational changes alike.