What is user story mapping?
User story mapping arranges work as a two-dimensional map: the horizontal axis is the sequence of steps a user takes to accomplish something, and the vertical axis is depth of implementation within each step. Slicing horizontally across the map produces a release that works end to end, which is the entire point — a flat backlog cannot show you whether a release is usable.
The two axes
The horizontal axis is the user's journey in the order it happens. For a marketplace: find an item, evaluate it, buy it, track delivery, get support. This row is the backbone, and it is narrative — you should be able to read it aloud as a sentence.
The vertical axis is depth. Beneath find an item sit the many possible implementations: a search box, filters, saved searches, personalised recommendations, visual similarity search. Higher means more essential, lower means more elaborate.
Why the second axis is the whole idea
A flat backlog ranked by score will happily produce a top ten that contains four excellent search features and nothing about checkout. Every item is individually justified, and the release is unusable, because a user cannot buy anything.
The map makes this impossible to miss. You cut a release by drawing a horizontal line across the map, taking the top row of everything. The result is thin in every step but complete across all of them — the user can get from one end to the other. The second release drops the line lower.
That is the technique. Everything else is stationery.
Where it pays off most
Anywhere a partial journey is worthless. Payments, onboarding, checkout, anything regulated. Half a KYC flow is not half a feature; it is zero features and some wasted engineering.
It is also the fastest way to surface disagreement about scope. Teams that agree on a feature list frequently disagree about the journey it belongs to, and that disagreement stays hidden in a list and becomes obvious on a map.
Gmail's redesign was fundamentally a rearrangement of the mail journey rather than a set of new capabilities — conversation threading changed the sequence of steps a user moved through, which is the kind of change a story map represents well and a ranked backlog represents badly.
The common failures
A backbone that is a feature list. If the top row reads search, filters, notifications, it is a component inventory, not a journey. It should read as things a person does, in order.
Mapping the system instead of the user. The axis is what the user does, not what your services do. As soon as the backbone contains a queue or a service name, the map has stopped answering its question.
Treating the map as permanent. It is a planning conversation that leaves a residue, not a document to maintain. Use it to cut releases, then let the tracker carry the work.
Linear's opinionated refusal to be endlessly configurable is the same instinct at product level: pick the journey, make it excellent, and decline to support every possible variation of it.
Seen in practice
(3)Case studies where this shows up as a real decision, not a definition.