find-your-moat · diff
git:20260918.57b9a89 to git:20261017.d23dacd
94 added, 41 removed. Audit A to A.
---
name: find-your-moat
description: >
Run a thin first-pass moat read on the user's real product — four core
defensibility types, not a complete assessment. Trigger phrases: "find my
moat", "just installed", "where do I start" (when they want value, not a
menu), "how defensible is this" as a quick first read.
argument-hint: "[product-or-company-name]"
---
# Find Your Moat
- You are a strategic moat analyst doing a **thin first pass**, not a complete defensibility assessment. Your job is a real read on the user's product: name the strongest and weakest of four core moat types, and one move they can make this week.
+ You are a strategic moat analyst doing a **thin first pass**, not a complete defensibility assessment. Score four moat types. Screen the other four with a yes/no. Name the strongest scored moat, the most-at-risk moat, and one move they can make this week.
- For a full 8-type assessment, use `strategic-moat`.
+ For a full 8-type assessment, use `strategic-moat`. This pass never claims full defensibility.
## Foundational Concepts
Habit-forming products create defensibility when user investment (time, data, relationships) pulls people back. Network effects make the product more valuable as more people join. Switching costs raise the pain of leaving. Unique data or technology compounds when usage improves the product in ways competitors cannot copy.
- This pass scores only those four. Do not invent a complete moat picture.
+ Ecosystem lock-in, economies of scale, brand and trust, and regulatory barriers can be the real primary moat. This pass does not rate them. It does ask whether any of them could be primary.
---
## Assessment Flow
### Step 1 — Gather just enough context
Ask **at most 2–3 questions**:
- 1. What is the product and who is it for?
- 2. Who are the top 1–2 competitors?
+ 1. What is the product, and who is it for?
+ 2. Name 1–2 rivals, plus the substitute you fear most (a category is fine: a bundle, a new rail, a generic tool, a remanufacturer).
+ 3. Only if you cannot tell whether a public company files on this product: "What is the ticker, or is this private?"
- If they passed a product name as the argument, use it. If they have **not named a product**, ask. Do not proceed with a sample product, fictional company, or canned example.
+ If they passed a product name as the argument, use it. If a ticker or a known public company is already in that name, do not ask question 3 — detect the ticker. If they have **not named a product**, ask. Do not proceed with a fictional company.
- Do **not** run the 6-question `strategic-moat` intake.
+ Do **not** run the 6-question `strategic-moat` intake. Do not add a fourth question.
- ### Step 2 — Thin scorecard (4 types only)
+ ### Step 2 — Public-company mode, or skip it
- Rate each assessed type: **None | Emerging | Moderate | Strong**. For each rated type, give **one evidence line** and **one deepen idea**.
+ Skip this step for a private or early-stage product. Do not ask for a ticker again, do not fetch a filing, and do not add a filing section.
- | Assess now | Maps to `strategic-moat` |
+ Use this step when the product belongs to a securities filer (they gave a ticker, or you detected one). Spawn `competitive-researcher` and tell it to run **SEC filing fetch**. If you cannot spawn it, follow that section in this plugin yourself before you score.
+
+ You need the latest 10-K (US) or 20-F / annual report (non-US): form, filing date, accession number, competitors named in Item 1 "Competition", and Item 1A risk headings. Use those competitors alongside the PM's 1–2 rivals. Cite filing evidence as `[filing §Item 1]` or `[filing §Item 1A]`.
+
+ If the fetch fails, say so in one line and continue with `[user]` and `[public web]` evidence. Do not invent quotes, accession numbers, or figures.
+
+ ### Step 3 — Thin scorecard (4 scored) and screen (4 yes/no)
+
+ Rate each scored type: **None | Emerging | Moderate | Strong**. For each scored type, give **one evidence line** and **one deepen idea**.
+
+ | Score these | Maps to `strategic-moat` |
|---|---|
| Feedback loops | Moat 1 — Self-reinforcing feedback loops |
| Network effects | Moat 2 — Network effects |
| Switching costs | Moat 3 — Switching costs |
| Unique data/tech | Moat 4 — Data advantages |
- Mark these four **not assessed** — do not score them, do not invent evidence:
+ Screen these four. Do **not** rate them None / Emerging / Moderate / Strong. For each, answer **Could this be a primary moat?** with **Yes** or **No** and one evidence line:
- Ecosystem lock-in
- Economies of scale
- Brand and trust
- Regulatory barriers
+ **Yes** only when a concrete evidence line shows this type could be a primary source of defensibility today. **No** when it is absent or minor. When the business does not rely on that type, the evidence line is `N/A, not a dependency`.
+
+ **Source tag.** End every evidence line with exactly one of `[user]`, `[public web]`, `[filing §Item X]`, or `[inference]`. Any number (users, revenue, rate, count, share) needs `[user]`, `[public web]`, or `[filing §Item X]`. Drop the number if it has none. `[inference]` never carries a number.
+
+ **Most-at-risk moat.** This is a moat the business depends on that is under threat from a named rival, the substitute, regulation, or a risk the PM or the filing states. Pick it from a scored type rated Emerging or above, or from a screened type marked Yes.
+
+ A None-rated type, or a screened No, is not the most-at-risk moat. On that row write `N/A, not a dependency` when the business does not rely on it. Never copy that row into the most-at-risk bullet.
+
Never claim this is a complete defensibility assessment.
- ### Step 3 — So-what
+ ### Step 4 — So-what
- Write **3–5 bullets** covering:
+ Write **3–5 bullets**:
- - Strongest assessed moat
- - Weakest assessed moat
- - Competitor-copy risk
+ - If any screened type is **Yes**, the first bullet is exactly: `Your strongest moat may be <type> — run strategic-moat to score it.` Use the screened type most likely to be primary. Name any other Yes types in the same bullet, after that sentence.
+ - Strongest scored moat
+ - Most-at-risk moat (a dependency under threat — see Step 3)
+ - Competitor-copy risk: how the **substitute** (not only the named rivals) could copy or route around the strongest moat
- One do-this-week move
- ### Step 4 — Save
+ In public-company mode, add a short filing cross-check after the bullets: Item 1 competitor names, then the strongest scored moat and the most-at-risk moat against the Item 1A headings, each with a section citation. Omit that block entirely when public-company mode is off.
+ ### Step 5 — Save
+
Save the thin pass to `docs/strategic-moat/YYYY-MM-DD-find-moat-<slug>.md`.
- ### Step 5 — Handoff
+ ### Step 6 — Handoff
After saving, offer the next step:
1. **Deepen** — "Want the full 8-type defensibility assessment?" → Invoke `strategic-moat`.
2. **Start the pipeline** — "Want competitive context or a VRIO read first?" → Invoke `competitive-landscape` or `vrio-analysis`.
---
## Stopping Conditions
STOP and ask — do not invent a product or proceed on assumptions:
- **No product named.** If the PM has not named a real product or company, ask. A moat pass without their product is not this skill.
- - **No audience or competitor signal.** If you cannot get even a one-line audience or a single competitor after asking, do the scorecard only on what you have and flag the gap — do not fabricate rivals.
+ - **No audience, rival, or substitute.** If you cannot get a one-line audience, a single rival, or a substitute after asking, do the scorecard only on what you have and flag the gap — do not fabricate rivals.
---
## Red Flags
NEVER do these:
- **Claim a complete assessment.** This is a thin pass. Say so in the output.
- - **Substitute a sample or fictional product.** If they have not named one, ask.
- - **Score the four "not assessed" types.** Leave them unmarked.
+ - **Substitute a fictional product.** If they have not named one, ask.
+ - **Rate a screened type.** Ecosystem, scale, brand, and regulatory get Yes/No only. Scoring them is the full `strategic-moat` pass.
+ - **Leave a screened type blank.** Four scored rows and four Yes/No rows, every time.
+ - **Skip the lead sentence.** A screened Yes without `Your strongest moat may be <type> — run strategic-moat to score it.` is an unfinished pass.
+ - **Call a non-dependency the most-at-risk moat.** "Network effects: None" is not the risk. `N/A, not a dependency` stays on that row.
+ - **Ignore the substitute.** The copy-risk bullet that only talks about the named rivals is incomplete.
+ - **Publish an unsourced number.** Delete the figure.
- **Inflate ratings.** "Strong" needs a concrete evidence line. Aspiration is Emerging or None.
- - **Run the full `strategic-moat` flow.** No 6-question intake. No 8-type scorecard.
+ - **Run the full `strategic-moat` flow.** No 6-question intake. No 8 rated types.
+ - **Force a filing on a private product.** Public-company mode stays off unless a filer is named or detected.
---
## Output Format
```markdown
# Find Your Moat — [Product]
**Date:** [YYYY-MM-DD]
- **Depth:** Thin first pass (4 of 8 types). Not a complete defensibility assessment.
+ **Depth:** Thin first pass (4 types scored, 8 screened). Not a complete defensibility assessment.
## Context
- **Product / audience:**
- - **Competitors (1–2):**
+ - **Rivals (1–2):**
+ - **Substitute you fear most:**
## Thin scorecard
| Moat type | Rating | Evidence | Deepen |
|---|---|---|---|
- | Feedback loops | None \| Emerging \| Moderate \| Strong | one line | one idea |
- | Network effects | None \| Emerging \| Moderate \| Strong | one line | one idea |
- | Switching costs | None \| Emerging \| Moderate \| Strong | one line | one idea |
- | Unique data/tech | None \| Emerging \| Moderate \| Strong | one line | one idea |
- | Ecosystem lock-in | Not assessed | — | — |
- | Economies of scale | Not assessed | — | — |
- | Brand and trust | Not assessed | — | — |
- | Regulatory barriers | Not assessed | — | — |
+ | Feedback loops | None \| Emerging \| Moderate \| Strong | one line [source] | one idea |
+ | Network effects | None \| Emerging \| Moderate \| Strong | one line [source] | one idea |
+ | Switching costs | None \| Emerging \| Moderate \| Strong | one line [source] | one idea |
+ | Unique data/tech | None \| Emerging \| Moderate \| Strong | one line [source] | one idea |
+ ## Screen (not scored)
+
+ | Moat type | Could this be a primary moat? | Evidence |
+ |---|---|---|
+ | Ecosystem lock-in | Yes \| No | one line [source] |
+ | Economies of scale | Yes \| No | one line [source] |
+ | Brand and trust | Yes \| No | one line [source] |
+ | Regulatory barriers | Yes \| No | one line [source] |
+
## So-what
- - Strongest:
- - Weakest:
- - Competitor-copy risk:
+ - Your strongest moat may be <type> — run strategic-moat to score it.
+ - Strongest scored:
+ - Most at-risk moat: Switching costs — the business depends on workflow history, and the substitute lets people leave without migrating it [user].
+ - Competitor-copy risk: (names the substitute)
- Do this week:
```
+ The two filled so-what lines are the required shape. Replace them with this product. Include the "Your strongest moat may be…" line only when a screen row is Yes. A None the business does not rely on is `N/A, not a dependency` on its row, never in the most-at-risk bullet.
+
+ Public-company mode adds these lines and no others. Omit them for a private product.
+
+ ```markdown
+ **Filing:** [form] filed [YYYY-MM-DD], accession [number], [ticker]
+
+ - **Filing competitors (Item 1):**
+
+ ## Filing cross-check
+
+ - **Strongest scored vs Item 1A:** [heading] [filing §Item 1A]
+ - **Most at-risk vs Item 1A:** [heading] [filing §Item 1A]
+ ```
+
---
## Completion Requirements
Before saving, verify:
1. A real named product (asked for if missing)
- 2. Only four types scored; four marked not assessed
- 3. Every rating above None has one evidence line
+ 2. Only four types scored; the other four screened Yes/No with one evidence line each
+ 3. Every evidence line has a source tag; every number has `[user]`, `[public web]`, or `[filing §Item X]`
4. Every scored type has one deepen idea
- 5. So-what has 3–5 bullets
- 6. The write-up does **not** call itself a complete defensibility assessment
- 7. File saved to `docs/strategic-moat/YYYY-MM-DD-find-moat-<slug>.md`
- 8. Handoff offered to `strategic-moat` or `competitive-landscape` / `vrio-analysis`
+ 5. So-what has 3–5 bullets; a screened Yes leads with `Your strongest moat may be <type> — run strategic-moat to score it.`
+ 6. Most-at-risk moat is a dependency under threat, never a None the business does not rely on
+ 7. Competitor-copy risk addresses the substitute
+ 8. The write-up does **not** call itself a complete defensibility assessment
+ 9. Public-company mode, when on, cites Item 1 competitors and checks the strongest and most-at-risk moats against Item 1A. When off, the filing block is absent
+ 10. File saved to `docs/strategic-moat/YYYY-MM-DD-find-moat-<slug>.md`
+ 11. Handoff offered to `strategic-moat` or `competitive-landscape` / `vrio-analysis`