swot-analysis

SWOT Analysis (Strategy-Grade)

Most SWOT outputs are useless because they're vague ("strong brand," "limited resources") and disconnected from action. This skill enforces evidence-anchoring and produces a ranked strategic-priorities list as the actual deliverable — the four S/W/O/T quadrants are inputs to the priorities, not the output.

When to invoke

• Annual or quarterly strategic planning
• Pre-fundraise strategic narrative
• M&A diligence (target's strategic position)
• After a major competitive move (new entrant, competitor pivot)
• After a major macro shift (regulation, economic downturn, AI disruption)

Don't use for sales-collateral SWOTs — those want a flatter, more flattering format.

Inputs

• Brand: name, industry, value proposition, differentiators
• Brand intelligence: target audience, products/services, market size, growth trends
• Recommended channels (if known)
• Competitor intelligence: at least 3-5 direct competitors with their positioning + differentiators (use competitor-discovery-cot first if missing)
• Market trends affecting the category

If competitor data is missing, surface that in dataQuality — without it the SWOT degrades to a generic blurb.

The four quadrants — what each means

Quadrant	Source	Question	Trap to avoid
Strengths	Internal	What does this brand do better than alternatives?	"Strong team" — too vague. Be specific: "20+ years selling into mid-market healthcare"
Weaknesses	Internal	What internal limitations constrain growth?	"Limited resources" — generic. Be specific: "No retention program; CAC payback 18 months"
Opportunities	External	What favorable market conditions could the brand pursue?	Restating market trends. The opportunity is what this brand specifically can do with the trend
Threats	External	What unfavorable external forces could damage the brand?	"Competition" — bare. Be specific: "Hubspot bundling AI features at zero marginal cost compresses our category's pricing"

Method

Work in this order:

1. Establish data sufficiency. What's known? What's assumed? What's missing? Set dataQuality accordingly: excellent (all sources), good (most), fair (some), poor (mostly inference).

2. Pass through each quadrant once with evidence. Each item must cite the source: brand intelligence, competitor analysis, market data, observed channel performance. Items without evidence are downgraded or dropped.

3. Identify cross-quadrant tensions. Where does a Strength enable an Opportunity? Where does a Weakness magnify a Threat? These are the strategic levers.

4. Rank strategic priorities. From the cross-quadrant analysis, produce 5-8 candidate priorities ranked by impact × feasibility. Each priority must answer: what to do, why, and what to stop doing. Include estimated hours_per_week cost per priority.

5. Optimize priority selection (Z3 solver). Use the knapsack template from skills/solver-patterns (Template 2) to select the optimal subset of priorities under founder capacity constraints.

   Model inputs (derived from Step 4 candidates):

   • item_ids: each candidate priority's ID
   • values[i]: impact_score × swot_coverage_weight — where impact_score is {high=3, medium=2, low=1} and swot_coverage_weight counts how many S/W/O/T items the priority addresses
   • costs[i]: estimated hours_per_week per priority (from Step 4)
   • total_capacity: founder's weekly hours available for strategic work (ask if not provided; default 15)
   • categories[i]: list of SWOT quadrant tags per priority (e.g. ["threat_pricing_compression", "weakness_no_retention"])
   • must_cover_categories: all critical Threats (items scored "high" in threats quadrant) — at least one selected priority must address each
   • max_concurrent: default 3 (avoids context-switching for a solo founder)

   Solver flow (per skills/solver-patterns §2 and skills/gtm-output-schemas §8):

   1. clear_model — fresh session
   2. Build knapsack model from candidate priorities
   3. add_item calls per Template 2
   4. solve_model with 10s timeout
   5. On SAT: the selected priorities are the optimal set under founder capacity
   6. On UNSAT: founder capacity is too low to address all critical threats — report which threats are uncovered and recommend capacity expansion or threat acceptance
   7. clear_model — cleanup

   Output enrichment. For selected priorities: show hours allocated and remaining capacity. For dropped priorities: show explicit "freed hours" and which SWOT items go unaddressed. Flag uncovered SWOT items as accepted risks.

6. Set confidence. Not all SWOTs are equal. State your confidence (0-1) and why.

Output schema

{
  "strengths": [
    "Specific internal strength backed by evidence (3-7 items)"
  ],
  "weaknesses": [
    "Specific internal weakness backed by evidence (3-7 items)"
  ],
  "opportunities": [
    "Specific external opportunity tied to a market trend (3-7 items)"
  ],
  "threats": [
    "Specific external threat tied to a market force or competitor move (3-7 items)"
  ],
  "competitorInsights": {
    "competitive_advantages": ["Where this brand beats the named competitors, specifically"],
    "competitive_gaps": ["Where named competitors are stronger"],
    "market_opportunities": ["Unmet customer needs surfaced by competitor analysis"]
  },
  "marketInsights": {
    "industry_trends": ["Trends actively reshaping this category (12-24 months)"],
    "market_forces": ["Macro forces — economic, technological, regulatory"],
    "growth_areas": ["Sub-segments with above-average growth"]
  },
  "strategicPriorities": [
    {
      "priority": "Specific action — what to do",
      "rationale": "Why this priority is ranked here, citing the SWOT evidence",
      "impact": "high | medium | low",
      "effort": "high | medium | low",
      "hoursPerWeek": 5,
      "swotItemsAddressed": ["threat_pricing_compression", "weakness_no_retention"],
      "stopDoing": "What this priority explicitly de-prioritizes (a SWOT without trade-offs is a wishlist)"
    }
  ],
  "optimizedSelection": {
    "selectedPriorities": ["priority_id_1", "priority_id_2", "priority_id_3"],
    "droppedPriorities": [
      { "id": "priority_id_4", "freedHours": 8, "uncoveredSwotItems": ["opportunity_ai_tooling"] }
    ],
    "totalHoursAllocated": 14,
    "remainingCapacity": 1,
    "uncoveredThreats": [],
    "acceptedRisks": ["opportunity_ai_tooling"],
    "solverStatus": "optimal"
  },
  "confidence": 0.85,
  "dataQuality": "excellent | good | fair | poor"
}

Quality bar

• 3-7 items per quadrant. Fewer = thin analysis; more = unranked dump.
• Every item is evidence-anchored. No bare assertions.
• Strategic priorities have a stopDoing field. A SWOT without an explicit trade-off is wishful thinking.
• dataQuality: 'fair' or worse triggers a recommendation to gather more intel before acting. Bad data + confident SWOT = expensive strategy mistakes.
• Solver selection respects critical threat coverage. Every high-impact Threat must be addressed by at least one selected priority. If the solver returns UNSAT, surface which threats are uncoverable at current capacity.
• Dropped priorities are explicit. The droppedPriorities array isn't a graveyard — it's a conscious decision log with freed hours and accepted risk items.

Common pitfalls

• Strengths that read like marketing copy. "Best-in-class customer support" is sales material. "Sub-2hr median first response on enterprise tier" is a Strength.
• Opportunities that are just trends. "Market is growing" is not an opportunity. "Market growing 18% CAGR + our highest-LTV segment is healthcare which is growing 24%" is.
• No internal/external split. S/W must be internal; O/T must be external. If a "Strength" is "growing market," it's an Opportunity.
• Strategic priorities that aren't actually trade-offs. "Invest in marketing AND product AND sales" is not a priority list.