business-strategy

name	business-strategy
version	0.1.0
description	Route a founder problem to the right business framework, then load that framework on demand. Distills 14 business books (customer discovery, positioning, pricing, sales, growth, persuasion) into scored, decision-tree skills. Use when you say diagnose my business problem, what framework applies, how do I validate demand, price this, position this, generate leads, or close deals. Do NOT use for software design (use the engineering rules) or for a single named decision (use decision-critic).
license	MIT
user-invocable	true
metadata	{"pack":"business-strategy","jtbd_stages":["Discovery","Validation","Positioning","GoToMarket","Persuasion"],"source_pattern":"getagentseal/founder-playbook (MIT)"}

Business Strategy

The front door to a 14-book business-strategy pack. A founder arrives with a felt problem ("sales are flat", "nobody replies", "I do not know what to build") and no idea which framework applies. This skill names the real problem, routes to the one reference that addresses it, and refuses to guess when the symptom is ambiguous.

Each reference is an original distillation (decision tree, scored checklist, honest limitations, cross-book conflicts, worked example) of one book. They are loaded on demand: read only the one this router points you to.

When to use

Start here whenever the problem is a symptom ("growth stalled") rather than a named method ("run SPIN Selling"). If you already know the framework, skip the router and read its reference directly.

Triggers

Trigger phrase	Action
diagnose my business problem	Walk the decision tree below
what framework applies	Walk the decision tree below
validate demand or find customers	Start with mom-test; use four-steps only after interview evidence names a repeatable buyer
price or position this	Route to monetizing-innovation for price pain, or obviously-awesome for category confusion
get leads or close deals	Route to 100m-leads for lead volume, or spin-selling for qualified pipeline stalls

Decision tree: symptom to reference

Route on the dominant symptom. If two apply, take the earliest funnel stage: an upstream positioning problem usually causes the downstream messaging or pricing problem. Read the named reference, then act.

• You do not know what to build, or whether anyone wants it -> references/mom-test.md (interview for evidence, not flattery). Follow with references/four-steps.md only after interviews identify a repeatable buyer but you still lack first-customer motion.
• You can build it but have not proven demand -> references/lean-startup.md (Build-Measure-Learn, smallest MVP that tests the riskiest assumption).
• You cannot describe who the product is for, or prospects say "interesting" but do not act -> references/obviously-awesome.md (positioning and category choice).
• You are entering a crowded market and need a wedge -> references/crossing-the-chasm.md (beachhead). Follow with references/blue-ocean-strategy.md only when no beachhead can produce a defensible category position.
• People want it but balk at the price, or you are guessing the number -> references/monetizing-innovation.md (price before you build, willingness-to-pay).
• The offer itself feels weak -> references/100m-offers.md (stack value so the price feels trivial).
• Traffic is fine but no leads -> references/100m-leads.md (lead generation), and re-check messaging.
• The words on the page fall flat -> references/storybrand.md (make the customer the hero). Follow with references/made-to-stick.md only when the message is clear but not remembered or repeated.
• Leads stall in the pipeline, or you lose at the close -> references/spin-selling.md (situation, problem, implication, need-payoff). Follow with references/influence.md only when discovery is sound but decision dynamics still block action.
• You picked product-market fit but have no growth channel -> references/traction.md (test 19 channels, find the one that works).

Process (progressive disclosure)

1. Name the symptom in one concrete sentence with a number ("20 trials, 1 paid"). A vague symptom cannot be routed.
2. Walk the decision tree to the earliest broken stage. Weak close is often weak positioning; low conversion is often the wrong audience. Stop at the root, not the symptom.
3. Read ONLY the primary reference the tree points to first. Each one is self-contained.
4. Apply that reference's scored checklist. If the checklist passes and the route names a conditional follow-up, load that follow-up next. If no follow-up is named or the follow-up also passes, re-diagnose: the funnel breaks in more than one place at once.

Self-assessment: is the router the right entry point?

Score each yes (2 pts for the first two, 1 pt each for the rest):

• I can state my symptom in one concrete sentence with a number. (2)
• I have NOT already named the framework I need. (2)
• My problem touches more than one funnel stage, or I am unsure which. (1)
• I have customer or revenue data to walk upstream against. (1)
• I am at discovery, not mid-execution on a chosen plan. (1)
• I am willing to be routed away from the fix I assumed I needed. (1)

Total 8. Pass threshold 4. Below 4, you likely already know your stage: skip the router and read the named reference. At or above 4, walk the decision tree.

Honest limitations

• Garbage in, garbage out. A symptom stated as "we need to grow" routes nowhere useful; the method needs a concrete, numeric symptom.
• Single-bottleneck bias. The funnel model assumes one dominant broken stage; real businesses often break in two places at once. Re-diagnose after each fix.
• Founder-context bias. The source books target early-stage software founders. A regulated, hardware, or enterprise business will find some routes mis-weighted.
• This router points; it does not execute. Each reference carries its own caveats.

Verification

You used this skill correctly when:

• The symptom was stated in one concrete sentence with a number before routing.
• Exactly one reference was read first (the earliest broken funnel stage), not all 14.
• That reference's scored checklist was applied and produced a pass/fail with a number.
• A falsifiable success metric was set, so the founder knows when to stop or re-diagnose.

Anti-Patterns

Avoid	Why	Instead
Reading all 14 references up front	Defeats progressive disclosure; burns context	Route to one, read it, act
Routing a vague symptom ("we need to grow")	The method needs a concrete numeric symptom	Force a one-sentence symptom with a number first
Fixing the symptom stage, not the root	Weak close is often weak positioning	Walk upstream to the earliest broken stage
Treating a framework as truth	Each book has honest limitations and conflicts	Read the reference's limitations and conflicts sections

References

Fourteen books, one file each, all under references/:

• references/mom-test.md - The Mom Test (Fitzpatrick): customer interviews.
• references/four-steps.md - The Four Steps to the Epiphany (Blank): first customers.
• references/lean-startup.md - The Lean Startup (Ries): Build-Measure-Learn, MVPs.
• references/obviously-awesome.md - Obviously Awesome (Dunford): positioning.
• references/crossing-the-chasm.md - Crossing the Chasm (Moore): tech adoption.
• references/blue-ocean-strategy.md - Blue Ocean Strategy (Kim/Mauborgne): category creation.
• references/monetizing-innovation.md - Monetizing Innovation (Ramanujam/Tacke): pricing.
• references/100m-offers.md - $100M Offers (Hormozi): offer design.
• references/100m-leads.md - $100M Leads (Hormozi): lead generation.
• references/spin-selling.md - SPIN Selling (Rackham): B2B sales.
• references/influence.md - Influence (Cialdini): persuasion principles.
• references/traction.md - Traction (Weinberg/Mares): growth channels.
• references/storybrand.md - Building a StoryBrand (Miller): brand messaging.
• references/made-to-stick.md - Made to Stick (Heath/Heath): message memorability.

Pattern credit: the books-to-skills structure (decision trees, scored checklists, honest limitations, cross-book conflict resolution) follows the MIT-licensed Founder Playbook. The distillations here are original. The npx ai-agents init --pack business installer flag is tracked separately (CLI milestone, not in this skill).