Install
$ agentstack add skill-dirknicol-pmm-skills-positioning ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Positioning
What this skill does
Helps the user develop, refine, or audit product positioning using April Dunford's five-component framework, optionally extended with Jonah Berger's contagious-idea principles for how positioning gets adopted internally and externally.
The big idea: in fast-moving markets, static positioning is a liability. This skill treats positioning as a living system — built on evidence, validated continuously, repositioned when signals demand it.
When to invoke
- The user is launching, relaunching, or repositioning a product or category.
- The user is preparing positioning for a fundraise, board review, or sales kickoff.
- The user senses positioning is "off" — sales objections changing, win rates dropping, analyst category shifting.
- A competitor has materially moved and a positioning response is needed.
Prerequisites
- Read
pmm-context.mdfirst when available. If it is missing, offer to runpmm-context. If that skill is unavailable or the user wants to continue, collect a minimum brief covering product, target customer, competitive alternatives, proof points, constraints, and current positioning. Mark unsupported inputs as assumptions; do not block the task. - Ask the user for any additional inputs: recent win/loss notes, sales call themes, analyst reports, competitor positioning teardowns, customer interview transcripts. Positioning grounded in evidence is the goal — guesswork makes generic output.
The framework (Dunford's 5 components)
For full methodology see DUNFORD_FRAMEWORK.md. Quick reference:
| # | Component | The question | |---|---|---| | 1 | Competitive alternatives | What would customers do if you didn't exist? (Not just direct competitors — spreadsheets, agencies, doing nothing.) | | 2 | Unique attributes | What do you have that the alternatives don't? Capabilities, not benefits. | | 3 | Value | So what? What benefit does each attribute enable for the customer? | | 4 | Best-fit customers | Who cares about that value the most? Who is willing to pay/switch/champion? | | 5 | Market category | What context do you put the product in so the value is obvious? |
Optional extension: Market trends — what shifts in the world make this positioning the right one now?
Workflow
Step 1 — Diagnose
Read pmm-context.md. State back the current positioning as you understand it. Ask the user: "Is this still right, or are we repositioning?" Capture the answer before proceeding.
Step 2 — Generate alternatives
Don't start with what the product is. Start with what customers do instead. List 5-10 competitive alternatives, including:
- Direct competitors
- Adjacent tools used in combination
- Manual workflows (spreadsheets, agency, intern)
- Doing nothing / status quo
For each, capture: what it does well, what it doesn't, why customers stick with it.
Step 3 — Identify unique attributes
Compare the product against the alternatives. List capabilities/features only this product has, or has materially better. Be ruthless — if the alternative has it too, it's not unique. Group attributes into 3-5 themes.
Step 4 — Translate attributes to value
For each unique attribute, ask "so what?" three times until you reach a customer-meaningful outcome (saved time, reduced risk, increased revenue, status, peace of mind).
Step 5 — Identify best-fit customer
Of all possible buyers, which segment cares most about this value? Define them by:
- Firmographics (company type, stage, size)
- Trigger events (what makes them start looking)
- Buying behavior (who decides, who blocks)
- Willingness to pay / switch
Step 6 — Choose a market category
The category frames the value. Same product, different category = different positioning. Test 2-3 category options:
- For each, list assumed competitors, assumed value drivers, assumed buyer
- Pick the category where your unique attributes most clearly win
Step 7 — Synthesize and output
Produce the artifacts below.
Outputs
A. Positioning canvas
A single-page table with all five components filled in. Include 2-3 trade-offs you're making (what you're NOT, who you're NOT for).
B. Positioning statement (Dunford format)
> For [best-fit customer] who [trigger / pain], [product] is a [market category] that [unique value]. Unlike [primary alternative], [product] [unique attribute].
C. Sales pitch outline
A 5-slide outline that walks a prospect from problem → market shift → category → value → why us. Use the positioning to anchor each slide.
D. Validation plan
3-5 ways to pressure-test this positioning before fully committing:
- Sales call A/B (use new positioning on 10 calls, compare conversion)
- Win/loss interview question: "What category did you put us in?"
- Analyst briefing: see how they recall the positioning a week later
- Homepage scroll-depth test
- Internal team blind test: do 5 employees describe the product the same way?
E. Repositioning triggers
List 3-5 signals that would tell the user it's time to revisit:
- Win rate drops below X
- Analyst category changes (e.g., Gartner renames the space)
- New competitive alternative emerges at scale
- ICP buyer title shifts
- Buyer language in calls changes materially
AI-powered positioning agents (advanced)
For users ready to operationalize, treat positioning as a continuous monitoring practice. See EXAMPLES.md for a setup pattern: define signals to track, choose monitoring tools, set up a synthesis layer, schedule human-in-the-loop reviews, socialize insights.
This is the difference between "we positioned the product in Q1" and "we have a positioning system that flags when reality drifts from our assumptions."
Quality bar
- No buzzwords in unique attributes. "AI-powered" is not unique. "Trained on 100M cited URLs" is unique.
- Trade-offs are mandatory. If the positioning doesn't say what the product isn't, it's not positioning, it's marketing.
- Evidence over assertion. Every claim about competitors, customers, or category should have a source (call, review, analyst quote, data point) or be marked
_assumption_. - The user's voice, not the model's. Match the user's vocabulary and tone from
pmm-context.md. Do not introduce phrases like "in today's fast-paced landscape."
Anti-patterns to refuse
- Generating a positioning statement without either shared context or a task-specific brief.
- Positioning that lists 8 unique attributes (force pruning to 2-4).
- Positioning that has no "what we're not" section.
- Treating positioning as a marketing-copy exercise rather than a strategic anchor.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: dirknicol
- Source: dirknicol/pmm-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.