Comparison
Minimum viable productvsWalking skeleton
Minimum viable product
you shipped the smallest thing that would tell you whether anyone wanted it, and half the team called it unfinished.
The smallest build that produces validated learning about whether the idea works. The original meaning is about learning, not about a cheap first version, and the word has drifted so far that it is worth restating in any room where it is used. The common failure is shipping something minimum and not viable, which teaches nothing except that a broken version of the idea is unpopular.
Full entry →Walking skeleton
the first thing built was a thin path all the way from the button to the database, doing almost nothing.
A minimal end-to-end implementation that exercises every layer and deployment step before any of them are fleshed out. It front-loads the integration risk — the wiring, the environments, the authentication, the pipeline — which is where surprises usually live, rather than leaving it for the final week. It is the standard antidote to a project that is ninety percent done on every component and has never once run.
Full entry →