# Bmc Canvas Clarity And Iteration > Guides BMC work with a variation logic aimed at clarity over depth: builds 2-3 alternative canvas variants, identifies when a team is stuck on a single canvas, and assesses canvas readiness using the expert's three criteria. - Skill: `pilot2service/bmc-canvas-clarity-and-iteration` (Agent Skill) - Install (CLI): `npx skillmds@latest add pilot2service/bmc-canvas-clarity-and-iteration` - Raw SKILL.md: https://api.skillmd.com/api/skills/pilot2service/bmc-canvas-clarity-and-iteration/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: Pilot2Service (https://skillmd.com/u/pilot2service) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/pilot2service/bmc-canvas-clarity-and-iteration --- # BMC Canvas Clarity and Iteration ## Purpose Prevents the most common pitfall in BMC work: polishing a single canvas endlessly instead of building several alternative models and comparing them. This skill puts the expert's own core principle into practice: value comes from comparing variants, not from the perfection of a single canvas, and clarity always beats depth. ## Anchored in - The owner's April 2026 expert interview, three `status: accepted` / `confidence: high` files in the owner's research notes: - `07_reasoning_model/reasoning_patterns/bmc_iteration_logic.md` — the variation logic, analysis/synthesis phases, the "unique combination" goal, four sufficiency criteria. - `12_quality_model/bmc_canvas_readiness.md` — the "clarity > depth" philosophy, signs of a finished canvas, a clarity scale. - `06_expertise_and_cognition/cognitive_signatures/bmc_cognitive_signature.md` — quick reads on canvas maturity, how a team's background shows up in the canvas, recognizing a stuck team. ## Method 1. **Build the first variant quickly and lightly.** Fill all nine blocks in one light pass — don't polish. Goal: one business model, not a perfect business model. 2. **Build a second (and, if needed, a third) variant with a different combination.** Use the remaining options, or a second pattern chosen from `../bmc-innovation-pattern-matching/SKILL.md`. A minimum of 2-3 variants before the analysis phase — one canvas is never enough to identify patterns. 3. **Identify signs of being stuck mid-work:** the team polishes one canvas deeply without producing variants; the same canvas stays in play for a long time without iteration; too many options per block (several segments/channels/revenue streams "active" at the same time). If you see these signs: enforce a sprint structure, a time limit per phase, move to a new variant instead of continuing to polish. 4. **Identify the team's background bias in the canvas.** People with a marketing/customer background tend to fill in customer segments and channels readily; people with an operations/production background fill in the left side strongly (key activities, key resources); finance-led people naturally think in cost/margin logic. Balance the work by steering the discussion toward whichever block is underdeveloped because of the team's background. 5. **Once 2-3 variants are ready, move to the analysis/synthesis phase:** - Margin perspective: which model produces the best margin? - Competitor perspective: how would competitors react to each? - Novelty perspective: what is genuinely new in each? - Scenario analysis: how does the model hold up to a change in market conditions? - Synthesis: build a combination of the variants' best elements toward a "unique combination" — not a perfect canvas, but one clear, focused customer segment + channel + revenue stream + value proposition. 6. **Assess readiness using the expert's three criteria** before moving forward: - **Tells one thing:** one business model, one set of choices, one focus — not several alternative or overlapping choices per block. - **Clear choices:** no "everything open" state in any block. - **Simple:** a good canvas can be one sentence per block — length isn't a measure of quality. Use this scale: High clarity (one segment, one channel, chosen revenue streams, a focused value proposition) / Low clarity (multiple options per block, no clear choices) / Stuck (a single canvas worked too deep without iteration). 7. **The work is sufficient when:** all blocks have been covered in a light first pass, at least two variants have been produced, choices are starting to crystallize, and the model can be communicated simply. At this point: either go deeper or move to more precise tools (see `../bmc-tool-switching-decisions/SKILL.md`). ## Gotchas - Step 6's readiness criteria are the expert's own simplified rubric, deliberately different from `bmc-canvas-diagnostic-reading`'s broader four-dimension rubric (see "Continue from here"). Passing "High clarity" here does NOT mean the canvas is internally consistent or evidence-backed — the two checks answer different questions and neither substitutes for the other. - Step 4 (background bias) only works if you actually know the team's professional background before judging which blocks look over/underdeveloped — without that context, a legitimately customer-led business can be misread as "unbalanced" when the imbalance is intentional. - The "2-3 variants" floor in steps 2 and 7 is a hard minimum, not a suggestion. Synthesizing from one canvas plus "the user's own alternatives" without ever producing a genuine second and third canvas breaks the core premise this skill exists to enforce: value from comparison, not from polishing one model. - Step 5's four analysis perspectives (margin, competitor, novelty, scenario) are meant to be run against every variant, not just the one the team already prefers — applying only the perspective that flatters a favored variant defeats the point of building alternatives in the first place. - The "Stuck" readiness state (one canvas worked too deep) and "Low clarity" (multiple options per block) need opposite fixes — force variant-building vs. force choices — so misclassifying which state a team is in sends them down the wrong corrective path; see `../bmc-antipattern-and-misunderstanding-correction/SKILL.md` for the fuller set of corrective moves. ## What this skill does NOT do - Doesn't produce patterns or innovation directions itself — it uses the patterns produced by `../bmc-innovation-pattern-matching/SKILL.md`, or the user's own alternatives, as variation material. - Doesn't systematically assess a canvas's internal contradictions or diagnostic signals — that's a narrower, more precise task, see `../bmc-canvas-diagnostic-reading/SKILL.md`. This skill focuses on the PROCESS (how to iterate), not on analytically reading the canvas. Note: this skill's readiness criteria are the expert's own, deliberately simpler and more opinionated view than the research layer's broader four-dimension rubric — see the explicit contrast in "Continue from here". - Doesn't make the final business model choice for you — the synthesis produces a recommendation, the human chooses. ## Continue from here - Previous skill in the same pack: `../bmc-innovation-pattern-matching/SKILL.md` — selecting patterns before building variants. - Next skill in the same pack: `../bmc-canvas-diagnostic-reading/SKILL.md` — once a variant is chosen, read it systematically for internal contradictions and evidence gaps. **Note:** this skill's "clarity > depth" criteria are the expert's own, lighter-weight view; `bmc-canvas-diagnostic-reading` uses the research layer's broader, more precise rubric. Use both — they don't replace each other, they answer different questions ("is this clear?" vs. "is this internally consistent and evidence-based?"). - Related skill in the same pack: `../bmc-antipattern-and-misunderstanding-correction/SKILL.md` — fixing being stuck and going too deep too early is one key antipattern addressed by this skill; see there for a fuller set of corrective moves. - This pack's shared guardrails: `../../CLAUDE.md` ## References - `../../references/bmc-source-material-notes.md` — source material background - `../../CLAUDE.md` — this pack's shared guardrails