PRO · Positioning & GTM
Derive the five components in sequence, each feeding the next. Reversing it — starting from the category you wish you were in — is how teams end up positioned against a market that doesn't want them. Worked here on a metrics product default-positioned as "yet another BI dashboard tool." illustrative
Dunford's 5 components · derive in order, each feeds the next
What they use if you don't exist
Tableau / Looker head-to-head, and "do nothing" — analysts hand-write SQL per request, metric definitions live in 40 conflicting dashboards. The status quo is the real incumbent.
Capabilities the alternatives lack
Metric definitions are git-versioned, code-reviewed, and compiled to one semantic layer every tool queries. No BI tool does this. (A specific mechanism — "easier to use" is not unique.)
What those attributes let them do
Consistent, governed numbers — "revenue" means the same thing in every dashboard, board deck, and Slack message; ends the "whose number is right?" fire drill; audit-ready.
Who cares most about that value
Mid-to-large data teams (≥5 analysts, illustrative) where conflicting metrics already caused a visible, expensive mistake. NOT the solo analyst at a 10-person startup — no pain yet.
The frame that makes the value obvious
Reframe from "BI dashboard tool" (where git-versioning reads as a quirky checkbox) to "metrics / semantic governance layer" — a frame where versioned definitions are the whole point. Same product, opposite win rate.
The category choice (step 5) is where the work pays off — but you can only reach it by deriving 1→4 first.
Section 1 · deliberate, not inherited
Misnaming positioning as "a tagline problem" sends you to copywriting when the frame itself is wrong. Three distinct jobs, three owners.
PM owns
Positioning
What it IS and for whom. The frame of reference — this whole module.
Marketing owns
Messaging
How to say it. The words that carry the positioning to market.
Sales owns
Narrative
How to tell it in context — the pitch shaped for the deal in the room.
Positioning is a choice, not a description.
You named the product against the alternatives that existed the week you started, and never revisited it. Pre-PMF, treat the statement as a living hypothesis you test in sales calls — not a monument you defend.
Each component feeds the next.
Alternatives → attributes → value → best-fit → category. Fail a validation test and the component above it is wrong — re-derive, don't patch the words.
Indecision, not a rival.
40–60% of B2B deals end in no-decision. A sharp frame reduces the buyer's fear of messing up — position against the status quo, not only competitors.
Section 3 · attributes → value
April Dunford · the "so what?" ladder
Ask "so what?" of each unique attribute until you land on a business outcome, not a feature:
Stop laddering when the next "so what?" is a business outcome, not a feature.
Gibson Biddle (DHM) · Christensen (JTBD)
DHM hard-to-copy test: a differentiator a competitor can ship next quarter is a feature, not positioning. Ask: does it make customers Delighted, is it Hard to copy, is it Margin-enhancing? Position on attributes that survive a rival's roadmap.
Sell what's on the truck: position the product you ship today, not the roadmap. AI capability is increasingly table stakes — frame on shipped value; address roadmap on the roadmap, not in the pitch.
Section 4 · the validation battery
Run all three falsification tests before you lock the frame (Dunford). Fail any → the component above it is wrong; go back, don't patch the words.
| Test | How to run it | Pass / Fail signal |
|---|---|---|
| Customer-mirror | Read the value statement to 5–8 best-fit customers cold. Ask them to describe, in their words, what it does and who it's for. | Pass: they echo your value (often better than you wrote it). Fail: they reach for the old category ("oh, a dashboard thing") → your category didn't land. |
| Substitutability swap-test | Swap your product name for the top competitor's in the statement. Re-read. | Pass: it now reads false (the differentiator is yours alone). Fail: still reads true → you described the category. Most generic positioning dies here. |
| Durability | War-game the next 2–3 competitor moves and the next model / tech shift. Does the frame survive them? | Pass: the differentiator is structural / hard-to-copy (DHM). Fail: a rival or base-model upgrade neutralizes it next quarter → it's a feature; re-derive. |
Worked swap-test: "Tableau compiles every metric from one code-reviewed source" reads false → the attribute is genuinely ours. Cadence: re-run customer-mirror every quarter and after any major competitor launch — positioning rots silently as the market moves around you.
Section 5 · pick the category
| Strategy | What it is | Cost / when to pick it |
|---|---|---|
| Head-to-head | Win an existing market as defined | Cheapest demand (buyers already shopping) — pick when you can out-execute incumbents on their own terms. Don't pick if your edge is invisible in their feature-compare grid. |
| Big-fish-small-pond | Sub-segment a market you can dominate | Best default for challengers — own a niche convincingly, then widen. Pick when a clear sub-segment feels your value most. |
| Create-a-new-game | Define a new category | Highest cost — you must build the demand and educate the buyer. Pick ONLY when no frame fits AND you can fund years of education. Most "new category" plays should have been big-fish. |
Geoffrey Moore · beachhead & the chasm
Crossing from early adopters (buy vision) to the early majority (buy references) requires a beachhead: one narrow segment you dominate completely before expanding. Do not spread thin pre-chasm — the early majority wants a safe, proven choice for their exact use case, not a flexible tool for everyone.
Our worked example sits between big-fish ("governance layer for data teams") and create-a-new-game ("metrics layer") — pick big-fish first unless you can fund category education.
Section 6 · the one-line statement
The single-sentence output of the derivation, in Moore's template.
Filled (worked): "For data teams who can't trust that 'revenue' means the same thing twice, [Product] is a metrics governance layer that gives every dashboard one audited, version-controlled definition. Unlike BI tools bolted onto ad-hoc SQL, our product compiles every metric from one code-reviewed source."
When buyer ≠ user (PLG / bottom-up B2B)
| End-user statement (drives adoption) | Economic-buyer statement (drives purchase) | |
|---|---|---|
| Cares about | getting the job done today, low friction | ROI, risk reduction, governance, total cost |
| Lands on | the product (in-app, docs, free tier) | the sales deck, security review, the order form |
| Value frame | "ship a trustworthy metric in minutes — no waiting on the data team" | "end the 'whose number is right?' fire drill org-wide; audit-ready governance, fewer board-deck errors" |
Failure mode: a bottom-up product that only writes the user statement gets viral adoption and then stalls at procurement because no one armed the buyer's ROI case (and vice versa: an ROI-only frame gets shelfware). Pressure-test each — hostile read, real alternative incl. status quo, benefit not feature, target narrow enough to fall in love — then run the §4 battery on it.
Section 7 · the sales pitch build
A two-act arc: setup reframes the buying criteria toward your strengths; follow-through shows your product satisfying the criteria you just reset.
Act 1 · Setup — reset the buying criteria
Insight / setup — the non-obvious truth about the buyer's problem. "Every team you query trusts a different 'revenue' number, and you only find out in the board meeting." If they nod, you've reframed the criteria.
Alternatives & their inherent tradeoffs — name the options including status quo, and the structural weakness each carries. Be fair; fairness earns trust.
"The way we see it" — perfect-world criteria — the 3–4 things a perfect solution would do, stacked toward what only you deliver (governed AND self-serve AND auditable).
Act 2 · Follow-through — satisfy them
Introduce the product — map each capability → a criterion → a value. Not a feature tour; a criteria-satisfaction walk.
Proof — references (the early-majority currency), data, a live demo of the exact JTBD. One on-segment reference beats five generic ones.
The ask + de-risk the decision — the failure mode here is not losing to a rival, it's the deal stalling. That's JOLT ↓
Section 8 · the real B2B enemy is indecision
of B2B deals end in no-decision — driven by the buyer's FOMU (fear of messing up), not a competitor. This is the JOLT Effect — Dixon & McKenna, The JOLT Effect — NOT Dunford. Position against the status quo, not only rivals.
Judge the indecision
Is the buyer stuck on valuation (is it worth it?) or outcome uncertainty (will it work for us?)? Diagnose before you push.
Offer a recommendation
Don't lay out 12 options and "let them decide" — that increases FOMU. Tell them what you'd choose for their situation, and why.
Limit the options
Narrow the configuration / tier choices; choice overload kills more deals than price.
Take risk off the table
Pilots, success guarantees, staged rollouts, opt-out clauses, proof-of-value periods — make "yes" feel safe.
A sharp frame reduces FOMU because the buyer can see this is built for their exact situation. Indecision is a positioning symptom as much as a sales one.
Section 9 · launch = risk reduction, not a date
Valuable
Does it solve a real problem? Evidence, not assertion.
Adoptable
Can users find it and succeed? Onboarding, discoverability.
Viable
Can you support and sustain it? Cost, ops, COGS envelope.
Mechanics: feature-flag to a small cohort first → watch real failure modes → widen (no big-bang). Ship a cohesive bundle — related changes as one coherent narrative, not a dribble of half-features.
Prepare
Beta, sales/support enablement, pricing + positioning locked, success metrics + guardrails defined, launch tier/config narrowed.
Coordinate
Product / marketing / sales / support in sync; internal comms before external; the §6 statement(s) and §7 pitch are the source of truth for all messaging.
Measure
Adoption + retained-value metrics, iterate, retro.
Launch-readiness checklist (fill-in)
Optics-velocity trap
Ship-count and launch velocity are vanity. A launch nobody adopts moved no outcome — measure adoption and retained value, not the press hit. (For the launch spec itself, the 10-section PRD — Kevin Yien (Square), popularized by Lenny Rachitsky, NOT a "ChatPRD standard" — carries the Non-Goals + Narrative sections a launch needs.)
Section 10 · keep positioning alive
Monthly PM↔Sales sync
Sales brings top-5 loss reasons + top-5 customer requests; PM shares upcoming releases + positioning guidance. This loop catches drift (and §4 customer-mirror failures) before they become a rewrite.
Customer-committed roadmap gate
When Sales promises a feature to close a deal, gate on (1) strategy fit, (2) how many other customers need it, (3) true cost incl. maintenance — separate customer-funded builds from the roadmap.
Deal-breaker escalation
When the CRO escalates, make opportunity cost visible: "Here's what we'd deprioritize to do this; want that trade?" Score one-off-vs-generalizable × deal value vs maintenance/opp-cost × reversibility → build / yes-with-conditions / no-with-alternative.
Aumayr split (mature orgs)
Strategic PM owns market strategy, pricing, positioning; Operational PM owns backlog and delivery — don't let positioning become an afterthought squeezed between sprints.
Regulatory gate
In fintech / healthcare, claims in positioning and launch copy are compliance-gated (substantiation, fair-balance, adverse-action rules) — legal review is a mandatory launch gate, and an unsubstantiated benefit claim can block the launch outright.