# Competitive Advantage > Moat analysis for products, companies, open-source projects, careers, and communities. Use when asking why something wins, whether an edge survives imitation or substitution (including AI commoditization), which advantages form a coherent strategy stack, or when another skill needs competitive-advantage vocabulary. - Skill: `wkentaro/competitive-advantage` (Agent Skill, multi-file: 3 files) - Install (CLI): `npx skillmds@latest add wkentaro/competitive-advantage` - Raw SKILL.md: https://api.skillmd.com/api/skills/wkentaro/competitive-advantage/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: wkentaro (https://skillmd.com/u/wkentaro) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/wkentaro/competitive-advantage --- # Competitive Advantage Reason from evidence about why something wins and whether that advantage can last. Treat Steph Ango's [How things win](https://stephango.com/moats) as a catalog of strategic prompts; use the moat test to classify each candidate. ## Modes **Explain.** Teach the vocabulary, distinguish an edge from a moat, and use a concrete example. **Diagnose.** Identify the advantages an existing thing actually demonstrates and the mechanism behind each one. **Design.** Propose a small, coherent strategy stack and the reinforcing loop it should create. **Audit.** Challenge a moat claim against evidence, imitation, substitution, decay, and value capture. ## Glossary Use these terms consistently. **Arena** — the environment in which alternatives compete over a meaningful time horizon. **Contested resource** — what the alternatives are trying to accumulate: attention, money, users, contributors, trust, territory, survival, or something else scarce. **Edge** — a current advantage that helps win the contested resource. An edge may be temporary or easy to copy. **Capability** — a repeatable ability to produce an outcome. A capability becomes a moat only when reproducing or substituting for it is difficult enough to matter. **Moat** — an edge whose benefit resists imitation or substitution over the relevant time horizon. **Accumulation** — a stock built through time or use, such as trust, data, distribution, reputation, community, or operational knowledge. **Strategy stack** — two to four compatible advantages that reinforce one another through explicit causal links. **Flywheel** — a causal loop in which exercising one advantage strengthens another and feeds back into the first. **Counterstrategy** — the cheapest credible way an alternative can neutralize, bypass, or reverse an advantage. **Decay** — a change in technology, regulation, taste, scale, ownership, or behavior that weakens an advantage over time. ## The moat test Test each claimed advantage on six dimensions: 1. **Value** — Show how it helps win the contested resource. 2. **Evidence** — Point to observed behavior or results. 3. **Resistance** — Explain why imitation or substitution is costly, slow, structurally blocked, or self-defeating. 4. **Accumulation** — Identify what gets stronger through use, scale, or elapsed time. 5. **Capture** — Identify who retains the resulting value and what prevents it from leaking away. 6. **Durability** — State the time horizon and the changes that could end the advantage. Call something an **edge** when value and evidence are present but resistance is weak or unknown. Call it a **capability** when the organization can produce it repeatedly. Reserve **moat** for advantages with demonstrated resistance over the stated horizon. The same mechanism can be both a capability and a moat. ## Analysis workflow ### 1. Define the arena Name the thing being analyzed, its contested resource, the real alternatives, and the time horizon. Include non-consumption, manual work, and adjacent substitutes where relevant. Complete this sentence: > Over **[time horizon]**, **[thing]** competes with **[alternatives]** for **[contested resource]**. Complete this step when the sentence names one contested resource, a time horizon, and every material class of alternative. ### 2. Gather evidence Inspect behavior that the proposed advantage should cause: retention, referrals, switching, contribution, price tolerance, cost curves, uptime, distribution, ecosystem growth, or repeated outcomes. Mark unsupported claims as unknowns. For current markets, competitors, or metrics, verify facts against current primary sources before treating them as evidence. Complete this step when every proposed advantage has supporting evidence or an explicit unknown. ### 3. Inventory candidates Read [references/strategy-catalog.md](references/strategy-catalog.md) when broad ideation, a full moat inventory, or Ango's taxonomy is useful. Select only candidates that can be connected to the arena and evidence. State every candidate as a mechanism: > Because **[cause]**, each **[use, user, transaction, or elapsed period]** creates **[accumulation or constraint]**, making **[imitation or substitution]** harder. If the sentence cannot be completed, keep the candidate as an edge or hypothesis. Complete this step when every candidate has a mechanism sentence or has been classified as an edge or hypothesis. ### 4. Build the strategy stack Choose two to four compatible advantages. Draw the causal arrows between them and close the loop where the evidence supports one. Prefer a small stack with a believable mechanism over exhaustive category coverage. Check tensions explicitly. Affordability and luxury, specialization and versatility, predictability and surprise, or centralization and decentralization can each work, but combining opposites requires a clear separation by customer, layer, or phase. Complete this step when two to four advantages have causal arrows between them and every tension is resolved by customer, layer, or phase. ### 5. Attack the stack Find the cheapest counterstrategy, the strongest substitute, the accumulation that could reset, the dependency that could withhold value, and the environmental change most likely to cause decay. Apply legal and ethical constraints before recommending any strategy; coercive and deceptive catalog entries are often useful only for diagnosis and defense. Complete this step when each attack surface has a named threat or is explicitly unsupported by available evidence. ### 6. Deliver the result Lead with the verdict, then report: - Arena and contested resource - Strongest strategy stack - Flywheel or causal mechanism - Evidence for each claimed advantage - Edge, capability, or moat classification - Counterstrategies and decay risks - Unknowns and the cheapest next test Separate observed advantages from proposed ones. A design proposal is a hypothesis until behavior demonstrates it. Complete the analysis when every reported moat has evidence, resistance, and a stated time horizon, and every proposal is labeled as a hypothesis. ## Common reasoning failures **Feature moat.** A visible feature is usually an edge unless copying it fails to reproduce an accumulation, system, or constraint behind it. **Label collecting.** Applying many catalog names without causal mechanisms creates no strategy. **Scale as explanation.** Current size is evidence of past success; name the mechanism by which scale changes future competition. **Internal excellence without capture.** A capability that benefits customers but cannot retain users, margin, learning, or bargaining power may create value without defending the creator. **Static analysis.** A moat is relative to alternatives, an environment, and a time horizon. Change any of the three and reassess.