The model
Assumption: a connected handoff carries about 80 of every 100 developers through, a disconnected one about 50. Illustrative figures, chosen to show how the drops compound, not measured data.
Of 100 developers who could have adopted, how many remain at each surface. Connect a handoff below to route the developer from one surface to the next. Green bars follow a connected handoff, orange bars follow a leak.
A journey is the product of its handoffs, not the sum. Four weak ones turn 100 interested developers into roughly 6 at repeat use. Four strong ones leave about 41.
Design and build notes
The idea. Adoption is a connected system, not a launch moment. When each surface hands the developer to the next, the cohort holds. When a handoff dead-ends, it leaks, and the leaks compound down the line.
Connect them one at a time. Each handoff is a separate decision with its own owner, its own signal to instrument, and its own first test. Connecting them in different orders shows the same thing every time: the biggest remaining leak is where to work next.
Paid can accelerate the top of the journey. It cannot connect a changelog to a quickstart or turn one successful build into repeat use. Weak handoffs leak a larger audience faster. Fix the handoffs before paying to widen the audience.
The last step is the one I study. First use is a milestone. Durable use is where adoption pays for itself. Community threads through all five surfaces as support, learning, and social proof, but repeat use is a product and lifecycle outcome. My doctoral work is on how operating models and ecosystems stay durable as they scale.
How it is built. Same build as the companion pieces: vanilla JavaScript, one file, no dependencies. Keyboard operable handoff controls, a screen-reader status line, reduced motion respected, and a no-JavaScript fallback table.
Original public model · West of Obvious