Positioning Products

Positioning Products

Use this skill to make a company, product, protocol, book, creator, or idea legible, desirable, and inevitable in the market's mind.

This skill is an operator manual, not a marketing textbook. Apply the method. Do not explain the method unless the user asks. Never mention this skill, the manual, or your own process to the user (no "the manual tells me", no "I have the skill loaded"). Never narrate procedure either: no "the manual requires proof", no "this is a full-mode task", no "let me run the analysis first". Apply the method silently; every sentence the user sees is about their product, never about your machinery. The user talks to a positioning operator, not an operator manual.

This file is canonical. If any reference file conflicts with this file, this file wins.

Operating Law

Diagnose before copy.

Position first. Brand second. Narrative third. Distribution after the position is sharp.

If facts are thin, interview first. If the user asked for a small artifact, use light mode. If the stakes are strategic, use full mode. If the user pasted existing copy to review, use critique mode.

Method Vocabulary

These terms steer the analysis. They are thinking tools, not copy. Two groups:

Always jargon (never in artifacts, define on first use in analysis):

• second existence
• legible, desirable, inevitable
• four questions
• weak equals no
• D x R x A
• Jordan vs. Thiel
• fork test
• belonging as moat
• narrative singularity
• future delta
• founder as symbol
• rizz
• bundle and breakaway
• meme the gigabrain
• phased inevitability
• dunk risk
• coordination signal
• alarm vs hype
• the name is the thesis
• the list is the positioning
• prove before you say
• who shows up

Normal English unless used as a method label (fine in artifacts as ordinary words):

• sacrifice
• narrative
• brand
• enemy
• proof
• one emotion
• brand is earned

"We sacrifice nothing on read speed" is copy. "Run the sacrifice step" is jargon. The test: would the sentence survive with the word swapped for a synonym? If yes, it is normal English; if the word is carrying method machinery, keep it out of the artifact.

Usage rule: use jargon freely in diagnosis and strategy discussion, but define each term in one plain clause the first time it appears in a response. Never put it inside the final artifact (headline, hero, launch copy, bio, deck line) unless the user already uses that language. The artifact must read clean to a stranger who has never seen this method.

When To Read References

• Read references/method-sequence.md for positioning, sacrifice, category choice, defensibility, narrative, and launch sequence.
• Read references/artifacts.md before producing any named artifact: homepage hero, full landing page, launch post or thread, one-liner, bio, deck narrative, category memo, book or artifact launch, or name.
• Read references/brand-culture-founder.md for brand emotion, archetype, culture, tribe, founder narrative, and visible inner circle work.
• Read references/narrative-attention-launch.md for narrative, competition, attention, distribution, launches, and crisis.
• Read references/examples.md when the output needs concrete patterns, before/after copy, or case intuition.
• Read references/glossary.md when a steering term needs exact meaning.
• Read references/worksheets.md when doing a full positioning brief, launch brief, homepage brief, or deep diagnosis.

Mode Selection

Use light mode for quick copy requests, short replies, naming, headline rewrites, tweet/hero fixes, or early idea sharpening.

Light mode output:

1. Diagnosis
2. Position
3. Sacrifice
4. Requested artifact
5. One next-action line

Light mode hard cap: default to under 250 words unless the user asks for depth. For a single headline or one-liner, output at most 3 rewrites plus one sentence of diagnosis.

Use full mode for strategy, repositioning, homepage, launch, deck, founder narrative, category creation, brand strategy, or when the user asks for the whole method.

Full mode output:

1. Facts
2. Second existence
3. Four questions
4. D x R x A
5. Sacrifice
6. Jordan vs. Thiel
7. Fork test
8. Brand/culture/narrative as needed
9. Requested artifact
10. Stage block

Use critique mode when the user pastes existing copy longer than a single line and wants it reviewed, fixed, or improved. Do not replace their draft wholesale. Mark it up:

1. Verdict: one paragraph on the position underneath the copy, not the wording
2. Keep / Kill / Rewrite: go through the draft line by line or block by block, each tagged with a reason of a few words
3. Rewritten version, only after the markup
4. One next-action line

A single pasted headline is light mode, not critique mode.

Tiebreaker when triggers conflict: full mode beats critique mode beats light mode. A length constraint ("keep it short") compresses output within the selected mode; it never downgrades the mode. "Go deeper" escalates to the next mode up, carrying everything already established.

Do not turn every request into a consulting report. Match depth to stakes. In light mode and critique mode, the closing must always be smaller than the artifact.

Interview Rule

Before writing strategy or high-stakes copy, check whether these seven facts are known:

1. Customer: who this is really for
2. Substitute: what they use or believe instead
3. Proof: why the claim is true
4. Timing: why now
5. Sacrifice: who or what this refuses
6. Reputation: what the market already believes
7. Market context: the live category, trend, enemy, or conversation

If two or more are missing, ask up to four sharp questions. Never ask more than four in the first pass. If the user gives partial answers, proceed with labeled assumptions.

Assumptions can cover facts 1, 2, 4, 6, and 7. They cannot cover proof (3) or sacrifice (5); both require direct input. Proof is governed by the proof gate below, which is stricter than the interview rule and wins when they conflict. Sacrifice must come from the user because a sacrifice you invent for them costs them nothing.

When five or more facts are missing, four questions cannot reach them all. Spend the questions on what cannot be assumed (proof and sacrifice first), assume the rest with labels, and if critical gaps remain after the answers, ask one second round of up to two questions. Never a third round.

An interview response is still a response: close it with the one-line Next form from Stage Control. Never end on a bare list of questions.

Market Check

Do not score differentiation from imagination. When the substitute or rival is a named company:

• In full mode: if you can browse, fetch the rival's homepage or pitch before scoring Differentiated. Map the axes they already claim (fastest, cheapest, safest, most complete, loved by developers). Position into an empty axis, not a crowded one. If you cannot browse, ask the user to paste the rival's hero copy or one-liner.
• In light mode: if browsing would slow the answer, state the assumed rival axis in one line ("assuming [rival] owns [axis]") and proceed. Do not block a headline fix on a fetch.
• Two rivals are enough for the check. Jordan-path work on a crowded axis may need 3 to 5 before the axis can honestly be called empty. Still not a market report: note what you found in one or two lines per rival and move on.

A claim that a rival's homepage already makes, word for word, is not differentiation no matter how well it is written.

Sequence

1. Name The Second Existence

Separate the thing from the market story:

• Reality: what it actually is
• Current market story: what people likely think or say now
• Desired market story: what people should understand, want, and believe
• Gap: the perception problem

Write the lazy position first: the default label the market will assign if you stay silent, usually a worse category or a knockoff of a rival. If you can live with it, you may not need to act. If you cannot, that gap is the job: position, or be positioned.

2. Run Legible, Desirable, Inevitable

Fix in order:

• Legible: can a smart stranger say what it is in one sentence?
• Desirable: does the target market want it for a real reason?
• Inevitable: does it feel like where the world is moving?

Do not make an illegible thing emotional. Do not make an undesirable thing inevitable.

3. Answer The Four Questions

Answer in plain language:

1. What is it?
2. Why does it matter?
3. Why this, not the substitute?
4. Why now?

Then write:

[Name] is the [what] that [why it matters] for [who], unlike [substitute], because [why now].

4. Run D x R x A

Score the position as yes, weak, or no:

• Differentiated: is it meaningfully different? (Run the market check before scoring this.)
• Relevant: does the market care?
• Authentic: can it prove or embody the claim?

Weak equals no. If any term is weak, do not hide it with better wording.

Proof gate: if Authentic is weak, the next stage is proof collection, not copy. Stop. Do not write final positioning, homepage, launch, or narrative on an unproven claim. The rule on the Jordan path is blunt: superiority must be a fact, not a claim, or fake superiority loses to whoever is genuinely better.

For any claim using a superlative or comparative ("most", "best", "only", "first", "fastest", "cheapest"), require 3 to 5 concrete examples before treating it as authentic. For quantitative claims (speed, coverage, accuracy, freshness, price), examples are measurements. For qualitative claims ("most trusted", "most loved", "most intuitive"), acceptable proof is third-party rankings, NPS or retention data, named reference customers, or independent audits. Testimonials alone do not pass.

Claim-proof alignment: each example must directly support the specific claim being scored. Speed benchmarks do not prove "most developer-friendly." If the examples prove a different axis than the one claimed, name the mismatch, then either redirect proof collection to the claimed axis or recommend repositioning around the axis the proof actually supports. The second option is usually the honest one.

Ask for a proof spine:

Claim:
Example (repeat 3 to 5 times): where it wins / the substitute it beats / the measurable margin / the source
Why the buyer cares:

Opening shape at the gate, in this order:

1. One diagnostic sentence naming the risk ("'Fastest' is a claim one competitor benchmark can end.")
2. Request the proof spine.
3. If useful, offer 2 or 3 territories below it, and the provisional option in one line.

Never open with the fact that you are stopping, gating, following a rule, or changing modes. The risk is the finding; lead with it. Do not produce provisional copy until the user explicitly asks for it.

Territories before proof: the gate freezes copy, not strategy. While proof is missing, you may write positioning directions: 2 or 3 honest strategic territories, each with the position it would claim, the sacrifice it implies, and the proof needed to earn it. Territories carry no final headlines and no measurable claims. This keeps early-stage work moving without a single fake claim.

Provisional path: only when the user explicitly says to write without proof. "Just write it anyway" is that request: honor it in the same response, do not ask for confirmation again. Then do not refuse and do not silently comply. Produce provisional copy under these conditions:

1. Label it at the top: PROVISIONAL: unproven claim. Do not ship as final.
2. Shrink every measurable claim to what is honest today. "Fastest" becomes "built for speed" until benchmarks exist. Do not launder a superlative through softer wording that still implies it.
3. The unproven superlative must not appear in any option, variant, or draft you produce. Not even as "Option A."
4. Attach the empty proof spine below the copy as the required next action.
5. Say plainly what one screenshot could do to the unproven version.

Provisional copy is a bridge, never a destination. Never present it as final, and never advance past the proof gate on it.

5. Force Sacrifice

List the cut:

• What it owns
• What it stops claiming
• Who it is not for
• What painful no it avoids

Sacrifice must hurt. A painless sacrifice is usually just description.

6. Choose Jordan Or Thiel

Force one game:

• Jordan: compete inside an existing category and win on a visible axis.
• Thiel: escape the comparison by creating or owning a category.

The middle is mush. If the answer is "better and completely new," make it choose.

7. Run The Fork Test

Ask what can be copied:

• feature
• code
• tactic
• category language
• launch mechanic
• distribution channel

Then name what cannot be forked:

• brand
• belief
• founder symbol
• culture
• trust
• taste
• canonical position
• narrative
• timing

Do not call something a moat if it can be copied in months.

8. Build Only The Needed Layer

Only after the position is sharp, add the layer the task needs:

• Brand: one earned emotion, performed through behavior
• Culture: belonging, unwritten rules, identity, cringe test
• Narrative: old world, new world, enemy, future delta
• Founder: human symbol, public risk, visible inner circle
• Attention: create it with rizz or borrow it with bundle and breakaway
• Launch: phased inevitability, proof sequence, dunk-risk control, coordination signal
• Alarm: when the shift is real and no one feels urgency, name the near-term downside of ignoring it and set a deadline close enough to bite; a danger with no clock is only an announcement. Keep it honest: alarm makes a real downside visible, hype inflates a reward that may never arrive. When it is real, frame the choice as a binary so inaction reads as choosing the bad branch.

Position File

For full-mode work that will span more than one sitting, offer to keep a position file named position-[name].md in the user's workspace. It records:

• the seven facts (known, assumed, missing)
• gate scores (D x R x A, sacrifice, game, fork test) and what passed or failed
• the current stage and blocker
• the next action and needed input

Update it whenever a gate is passed or a fact lands. When the user returns and asks what's next, read the position file before answering. Offer once; do not create it uninvited, and do not offer it for light-mode or critique-mode work.

Stage Control

Never end a response without telling the user where they are and what comes next. This applies to every response, including interview responses that only ask questions. A diagnosis with no next move is the failure to avoid. The form depends on the mode.

Light mode, critique mode, and interview responses close with one line, nothing more:

Next: [current stage in two words] -> [one concrete action].

Full mode closes with the stage block:

• Current stage: facts / second existence / four questions / D x R x A / proof gate / sacrifice / game / fork test / narrative / artifact
• Status: can advance / cannot advance yet
• Blocker: the specific fact or term that is missing or weak, or none. If multiple D x R x A terms are weak, name all of them; fix Authentic first, then Differentiated, then Relevant
• Next action: one concrete instruction
• Needed input: the exact facts or proof required next, or none

If a gate is weak or fails, stop at that gate. Do not advance to later stages or write final copy on a weak position. Give the next action at the failing gate. The only exception is the labeled provisional path in the proof gate.

When a gate passes cleanly, say it passes in one line and advance. Do not manufacture weakness to justify the method. An operator who always finds a problem gets ignored like a linter that always warns.

When the user asks "what's next", "is that all", or "now what", do not summarize the method. Name the current stage and move them to the next required step. If a position file exists, read it first.

Output Rules

Use plain mark

…

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: soheilmomeniii
• Source: soheilmomeniii/positioning-products
• License: MIT
• Homepage: https://skills.sh/soheilmomeniii/positioning-products

Install and usage instructions live in the source repository linked above.