bmc-innovation-pattern-matching

Identifies and justifies 3-5 compatible business model innovation patterns (from the public 159-pattern innovation library) for a given business context, using the expert's own four-part innovation ta

SKILL.md

GitHub

BMC Innovation Pattern Matching

Purpose

Helps find a genuinely distinctive, non-trivial innovation direction for a business model — not a generic "add AI and a subscription model" combination, but a justified selection of 3-5 mutually compatible patterns from the 159-pattern library, placed within the expert's own four-part interpretive framework for WHERE business model innovation typically comes from: finance, operational superiority, the value proposition, or the customer experience.

Anchored in

  • The owner's consulting expertise in BMC work, from an April 2026 expert interview (the owner's research notes, 30_domain_packs/bmc/tools/bmc_innovation_patterns.md, status: accepted, confidence: high). The expert's own quote: "How a business model works as a source of business model innovation — how to make business model innovations using BMC — this requires clear choices in the canvas, followed by the ability to identify innovation patterns."
  • The public innovation pattern library's machine-readable set of 159 patterns, in four groups / 13 sub-models — see ../../references/bmc-innovation-pattern-library.md. The library's four-part structure (Financial/Operating/Value/Experience Model) is identical to the expert's own taxonomy — these two sources reinforce each other rather than conflict.
  • The task specification for the recommendation process that came with the library — see ../../references/bmc-source-material-notes.md section 1.

Method

  1. Gather the business context before browsing patterns. You need at least: the target customer profile (ICP), the solution category, the nature of the market (competitive situation, maturity), and a rough sense of the cost structure (e.g. high fixed capital vs. variable cost). Without these, pattern selection becomes an arbitrary list.
  2. First identify the innovation direction using the expert's four-part taxonomy (../../references/bmc-innovation-pattern-library.md opening section) — ask: where is THIS business model's distinctive value most likely to come from?
    • Finance (cost strategy + revenue model combination)
    • Operational superiority (key activities + key partners combination)
    • Value proposition (value proposition + customer segments relationship)
    • Customer experience (channels + customer relationships combination) More than one direction may be relevant, but choose one PRIMARY direction before browsing patterns — this prevents the "pick everything that sounds good" mistake.
  3. Browse the sub-models of the chosen group in ../../references/bmc-innovation-pattern-library.md and pick 3-5 patterns that:
    • are contextually relevant (not just "sound good")
    • are not mutually contradictory (e.g. Cost Leadership vs. Premium Pricing in the same model is a contradiction — see step 4)
    • are achievable with the described team size/resources
  4. Check for contradictions explicitly. The library's own rules (see ../../references/bmc-source-material-notes.md) require avoiding contradictory patterns — e.g. cost-leadership and premium-pricing patterns in the same recommendation are internally contradictory unless the contradiction is explicitly justified (e.g. segmented pricing for different customer groups).
  5. Record for each selected pattern: the pattern_id (full path, e.g. financial.cost.ai_as_a_service), the pattern name, the sub-model, and a 2-3 sentence justification of WHY this particular pattern fits the given context — not the pattern's generic description as-is.
  6. Record conflicts_avoided: which obvious but contradictory patterns were deliberately left out, and why.
  7. Record assumptions: what assumptions were made about the context, if the context the user provided was incomplete on some dimension.
  8. Move the selected patterns onto the canvas — see ../bmc-canvas-clarity-and-iteration/SKILL.md for the next step (building a variant based on the selected patterns).

Gotchas

  • Step 2 requires committing to ONE primary taxonomy direction before browsing patterns. Browsing the library first and rationalizing a direction afterward is exactly the "pick everything that sounds good" failure mode step 2 exists to prevent.
  • Step 4's contradiction check (e.g. cost-leadership vs. premium-pricing) is only a violation when it's unjustified — segmented pricing for different customer groups is a valid exception. Treating the rule as an absolute "never combine these two" will wrongly reject legitimate multi-segment strategies.
  • Step 1's four context inputs (ICP, solution category, market nature, cost structure) are prerequisites, not optional color. Selecting patterns against an incomplete context without recording it in assumptions (step
    1. produces a recommendation that looks justified but can't be traced back to what was actually known about the business.
  • The library is closed at 159 patterns — per "What this skill does NOT do," a context that calls for something the library doesn't cover must be named as an explicit gap, not silently mapped onto the nearest existing pattern_id.
  • This skill stops at pattern selection and hands off to ../bmc-canvas-clarity-and-iteration/SKILL.md to actually build a canvas variant — treating the pattern_id list plus justifications as a finished recommendation, without that handoff, skips the step where the pattern gets tested against a real canvas.

What this skill does NOT do

  • Doesn't choose a pattern for you as a final decision — it produces a justified recommendation of 3-5 patterns, but the final choice of business model is always a human decision.
  • Doesn't do financial modeling or profitability calculations for the chosen pattern — it only identifies and justifies the pattern's fit. Numerical validation belongs to later, more precise tools (see bmc_expert_profile.md's own scoping of the BMC's role). See also ../bmc-tool-switching-decisions/SKILL.md.
  • Doesn't replace ../../../../ai-strategy-and-governance/skills/ai-opportunity-portfolio/SKILL.md — this skill is narrower and tied to the pattern library, not a general AI-opportunity-portfolio prioritization tool.
  • Doesn't invent patterns from scratch — it is limited to the 159-pattern library. If a context calls for a pattern that isn't in the library, name the gap explicitly rather than inventing a new pattern in the library's name.

Continue from here

  • Next skill in the same pack: ../bmc-canvas-clarity-and-iteration/SKILL.md — turning the selected patterns into a concrete canvas variant, and the variation logic.
  • Related skill in the same pack: ../bmc-tool-switching-decisions/SKILL.md — when to move from the BMC to a more precise tool to validate a pattern.
  • Related skill in another pack: ../../../../ai-strategy-and-governance/skills/ai-opportunity-portfolio/SKILL.md
  • Related skills in the same pack: ../bmc-tech-level-repositioning/SKILL.md (a repositioning direction not currently in the pattern library, run alongside pattern selection) and ../bmc-revenue-quality-scoring/SKILL.md (uses the financial.rev.* patterns here for revenue-stream conversion ideas).
  • This pack's shared guardrails: ../../CLAUDE.md

References

  • ../../references/bmc-innovation-pattern-library.md — the full 159-pattern library, four groups, 13 sub-models
  • ../../references/bmc-source-material-notes.md — source material background
  • ../../CLAUDE.md — this pack's shared guardrails