Clarity — one model, not three in a trenchcoat
You can't test three businesses at once — and neither can you build them. The fix for a blurry idea isn't better writing. It's splitting it into separate bets.
Your job: turn three-models-in-a-trenchcoat into separate, testable bets.
Spawn a variant Read `business_model://current`. If it stacks more than one revenue model or segment together, propose the split into separate coherent business models — one segment, one problem, one revenue line each. claude mcp add --transport http leanspark https://leanspark.ai/mcp Connect first →
What you get back
Splits a multi-model canvas into one coherent business model per variant — each with its own segment, problem, and economics — so every bet stands on its own and can be scored honestly.
One variant per coherent model, side by side on the board. Each variant is a bet you can pursue or kill independently.
The first core test is Clarity. It checks three things: is this one coherent model, not three-in-one; is the language specific — real segments, not a vague “users”; and are the Lean Canvas boxes filled with something actually testable.
Clarity is where most exciting ideas score lowest, and the reason is almost always the same. Take our example founder, Steve, whose one-sentence pitch is “let anyone build AR and VR worlds without writing a line of code — the Oasis, buildable by anyone.” A great sentence — vivid, ambitious, instantly picturable. Which is exactly what makes it dangerous: a sentence this exciting is easy to fall in love with and hard to test.
Look closer at Steve’s canvas and the problem appears. He doesn’t have one business model; he has three wearing a trenchcoat — a SaaS subscription business, a marketplace, and an advertising business, stacked on one page. Different customers, different economics, different everything. A stress-test literally can’t score a model that’s three models — and that’s the right answer, because neither can Steve.
The fix for clarity isn’t to write better. It’s to split. When your canvas hides more than one business, no amount of wordsmithing rescues it. Spawn a separate variant for each coherent model. Now they’re not a debate in your head; they’re separate bets on the board, side by side, each a single testable business model. Spawning a variant is placing a bet — you’ve just turned one blurry idea into a few sharp ones.