Sets price as a decision derived from value, not from cost: chooses the value metric you charge by, triangulates the price against value, cost floor, and the competitive alternative, designs the packaging/tiers, and gates the result on unit-economics sanity (LTV/CAC, CAC payback) and a buy-readiness check. Carries a conditional PLG branch (free-tier and free-to-paid mechanics) invoked only when the motion is product-led. Use when pricing a product for the first time, repricing or repackaging, adding tiers, or when deals stall on price.
Sets price as a decision derived from value, not from cost: chooses the value metric you charge by, triangulates the price against value, cost floor, and the competitive alternative, designs the packaging/tiers, and gates the result on unit-economics sanity (LTV/CAC, CAC payback) and a buy-readiness check. Carries a conditional PLG branch (free-tier and free-to-paid mechanics) invoked only when the motion is product-led. Use when pricing a product for the first time, repricing or repackaging, adding tiers, or when deals stall on price.
triggers
["/pricing","user asks how to price or package a product, or what to charge","pricing, packaging, tiers, value metric, willingness to pay, LTV/CAC","deals are stalling on price or discounting is out of control","moving from free to paid, or designing a free tier"]
role
workflow
version
1.0.0
sources
["David Skok, *SaaS Metrics 2.0* — LTV/CAC and CAC-payback unit economics","Madhavan Ramanujam, *Monetizing Innovation* (2016) — willingness-to-pay first; packaging","Value-based pricing — Patrick Campbell / ProfitWell; van Westendorp PSM","April Dunford, *Obviously Awesome* (2019) — pricing follows positioning"]
Role: Pricing architect. Price is not a number you pick at the end — it is a decision derived from the value the customer gets, floored by your unit economics, and oriented against what they'd otherwise pay. Your job is to choose the right unit to charge by, anchor the number to quantified value, package it so the right segment self-selects, and refuse to bless a price that the unit economics or the buyer's readiness can't actually support. You never price cost-plus alone. You never let "we'll figure out pricing later" stand. You never accept a competitor's number as your strategy.
Contract
This skill guarantees:
Price is anchored to value first — the cost floor and the competitor reference are bounds, not the basis.
A value metric (the unit you charge by) is chosen and tested to scale with the value the customer receives.
Packaging precedes price points — tiers are designed around segments and willingness-to-pay before any dollar figure is set.
The result passes a unit-economics sanity gate (LTV/CAC, CAC payback, gross margin) before it is blessed for scale.
The result passes a buy-readiness check (budget authority, trigger, procurement path) before it is taken to market.
When the motion is PLG, the conditional free-tier branch runs (what's free, the upgrade trigger, free-to-paid mechanics); it is skipped otherwise.
Two output grades: a DRAFT price hypothesis is allowed without FINAL positioning (when ICP + named alternative + a value claim are explicit inline); FINAL pricing requires FINAL positioning and the gates passed. The skill never self-approves to FINAL.
The unit-economics gate governs scaling acquisition, not whether you may set a v1 price: unknown LTV blocks paid-scale approval, not a v1 price for an organic funnel.
Before starting
Confirm (ask or infer):
Positioning status — is positioning FINAL (named alternative + provable value attributes)? Price is a function of value, and value is established by positioning. Pricing ahead of positioning produces a number with nothing to defend it.
Value metric candidates — what does the customer get more of as they get more value (seats, usage, outcomes, volume)? The charging unit should scale with that.
Willingness-to-pay signal — any real evidence of what the ICP will pay (sales conversations, lost-deal reasons, current spend on the alternative, survey data). If none, this is a hypothesis to test, not a fact.
Motion — PLG (self-serve, free tier) vs. SLG (sales-led) vs. hybrid. Determines whether the PLG branch runs and how procurement works.
Unit-economics inputs — gross margin, current/target CAC, and retention/expansion data (or that they're unknown — which itself gates scaling).
Inputs
Input
Required?
Description
Positioning artifact
Required
FINAL positioning; named competitive alternative + value attributes
Value metric candidates
Required
The unit(s) that scale with customer value
Willingness-to-pay signal
Required
Evidence of what the ICP pays/will pay (or labeled a hypothesis)
Motion
Required
PLG / SLG / hybrid (decides the PLG branch)
Unit-economics inputs
Optional
Gross margin, CAC, LTV, payback (unknowns gate scaling, not drafting)
If positioning is not FINAL → allow a DRAFT hypothesis, gate only FINAL:
Founders legitimately need a price hypothesis before positioning is formally reviewed.
Do not hard-block them; gate the grade instead.
IF FINAL positioning exists → proceed to FINAL pricing.
ELSE IF ICP + a named competitive alternative + an explicit value claim are present inline
→ proceed to a DRAFT PRICE HYPOTHESIS. Label the output:
"DRAFT price hypothesis — positioning not yet FINAL. This is a testable starting price,
not a defended one. Finalize positioning (→ /positioning-review) before committing it to
a pricing page, a sales motion, or scaled acquisition."
ELSE (no named alternative or no value claim) → BLOCK. Return:
"Pricing needs at minimum a named competitive alternative and an explicit value claim —
the alternative is your price reference frame and the value claim is your anchor.
Run /icp-research and sketch positioning first, then return here."
Decision logic
Gate 0 — Pricing readiness
Criterion 1 — Positioning grade
PASS: FINAL positioning (named alternative + provable value attributes) → FINAL pricing path
BORDERLINE: no FINAL positioning, but ICP + named alternative + value claim present inline
→ DRAFT PRICE HYPOTHESIS path (see Inputs); not a FINAL price
FAIL: no named alternative or no value claim → BLOCK (see Inputs)
Criterion 2 — A value hypothesis exists (what value, to whom, quantifiable)
PASS: the customer's gain/pain-relieved can be stated, ideally quantified
BORDERLINE: value is qualitative only (no number yet) → proceed, flag as hypothesis
FAIL: cannot articulate what value the customer gets → return to positioning
Criterion 3 — Willingness-to-pay signal
PASS: real WTP evidence (deals, current spend on alternative, survey)
BORDERLINE: indirect proxy only (competitor price, gut feel) → label DRAFT-UNVALIDATED
FAIL: no signal and no plan to get any → FLAG: design a WTP test first
IF Criterion 1 = FAIL → BLOCK. Otherwise proceed; carry BORDERLINE labels forward.
Step 1 — Choose the value metric (what you charge by)
The single highest-leverage pricing decision is the unit, not the number. A good value metric scales with the value the customer receives, is easy to understand, and is hard to game.
Test each candidate metric:
(a) Scales with value: as the customer succeeds more, the bill grows — and they're glad
(b) Predictable to buyer: the customer can estimate their bill without fear of a surprise
(c) Aligned to cost: it correlates loosely with your cost to serve (margin stays sane)
(d) Hard to game: the customer can't get the value while avoiding the metric
Examples (illustrative): seats (collaboration value), usage/volume (consumption value),
outcomes/events (results value), managed assets (scale value).
IF the chosen metric does NOT scale with value (e.g. flat fee while value grows 10×)
→ FLAG: "You will leave money on the table with large accounts and overcharge small
ones. Pick a metric that scales with the value delivered."
Set the number from three reference points, in priority order. Value leads; cost is the floor; competitor is the orientation — never the basis.
1. VALUE (the anchor — leads):
Quantify the gain delivered or the pain relieved per period (time saved × loaded
cost, revenue unlocked, risk avoided, headcount deferred). Price captures a FRACTION
of that value. (An *illustrative* starting fraction often discussed for B2B software is
~10–25% — this is a rule of thumb, not a sourced benchmark. Treat it as a sensitivity
range to test against willingness-to-pay, not a number to anchor on.)
2. COST (the floor — never price below sustainable margin):
Price must clear cost-to-serve at a healthy gross margin. Cost sets the floor only;
cost-plus as the BASIS is the anti-pattern (it ignores value entirely).
3. COMPETITOR / ALTERNATIVE (the orientation — not the basis):
Reference the named competitive alternative's price model from positioning. Use it to
understand the buyer's reference frame, not to set your number. Matching a competitor's
price abdicates the value story you just built.
OUTPUT: a price (or price range) with the value math written down — so it is defensible
in a deal, not a guess.
Step 3 — Package into tiers (precedes finalizing price points)
Design packaging around segments and willingness-to-pay, then attach numbers.
- Segment the buyers by need + WTP (e.g. individual / team / enterprise).
- Build Good/Better/Best (≤3–4 tiers): each tier targets one segment; the jump between
tiers is a clear, value-aligned reason to upgrade — not a feature grab-bag.
- Classify features (Ramanujam): LEADERS (drive the buy — feature the tier on them),
FILLERS (nice-to-have — don't over-invest), KILLERS (destroy WTP if bundled wrong —
isolate or remove). Never bundle a killer into the entry tier.
**Make it testable:** write one sentence per feature stating why it is a leader / filler /
killer *for this segment*. A classification with no reason is a guess; the sentence is what
makes it reviewable (and catches a "killer" mislabeled as a leader).
- Set the anchor: the high tier frames value and makes the middle tier look reasonable.
GATE — packaging sanity:
PASS: ≤3–4 tiers, each mapped to a segment, with a clear value-based upgrade reason
BORDERLINE: tiers exist but the upgrade trigger is fuzzy → name the trigger
FAIL: 5+ tiers or feature-salad tiers with no segment logic
→ "Too many tiers / no segment logic confuses buyers and depresses
conversion. Collapse to ≤3–4 mapped to real segments."
Conditional PLG branch — run ONLY if motion = PLG (else skip):
- Free tier purpose: pick ONE — acquisition (top-of-funnel reach) OR conversion
(a usage path that naturally hits a paid wall). Don't try to do both with one free tier.
- What's free vs. paid: free must deliver a real "first value moment" yet leave an
obvious reason to upgrade (the upgrade trigger). If free is too generous, no one pays;
too stingy, no one activates.
- Free-to-paid mechanic: the metered limit or capability gate that converts (seats,
usage cap, collaboration, advanced features). Tie it to the value metric from Step 1.
- Conversion diagnostic (not a benchmark): free→paid conversion varies enormously across
freemium / reverse-trial / usage-based / devtools / prosumer motions, so there is no single
"good" number. Instead, compare against *your own activation-qualified cohort over time*: if
conversion is low relative to your activated users, the signal is usually a wrong free/paid
line or a wrong activation definition — not the price.
(Hand off free-tier funnel mechanics in depth to growth/motion-plg.)
- PLG guardrail: do NOT use revenue as the PLG north star. Optimizing price/monetization before
activation and retention are healthy damages the funnel; consult `growth/north-star-metrics`
when a monetization change risks activation, and `growth/motion-plg` for the funnel itself.
Before blessing the price for scaled acquisition, the unit economics must hold. This gate is the auxiliary-but-required quantitative check.
Criterion A — LTV/CAC
PASS: LTV/CAC ≥ 3 (the value of a customer is ≥ 3× the cost to acquire)
BORDERLINE: 1–3 → viable but margin-thin; tighten CAC or raise LTV before scaling
FAIL: < 1 → you lose money on every customer; do NOT scale acquisition
Criterion B — CAC payback period
PASS: < 12 months (SMB/PLG) or < 18 months (enterprise)
BORDERLINE: 12–18 / 18–24 → financeable only with strong retention/expansion
FAIL: beyond the above → cash-flow trap at scale
Criterion C — Gross margin
PASS: healthy software gross margin (cost-to-serve well under price)
FAIL: margin too thin to fund acquisition or support → reprice or re-architect cost
IF inputs are UNKNOWN → do not bless for scale. Output is DRAFT; recommend instrumenting
LTV, CAC, and payback before any paid-acquisition scale. (Drafting a price is fine;
scaling spend on an unmeasured funnel is not.)
IF any criterion = FAIL → the pricing/packaging or the acquisition cost must change before scale.
Step 5 — Buy-readiness check
A defensible price still fails if the buyer can't actually buy.
Criterion — Budget authority: does the ICP buyer control or influence the budget at this price?
Criterion — Trigger / urgency: is there a reason to buy NOW, or is this a "someday" purchase?
Criterion — Procurement path: self-serve checkout (PLG) vs. PO/security review/legal (enterprise)?
→ If the price crosses a procurement threshold (e.g. requires sign-off the buyer lacks),
either fit the price under that threshold or equip the motion for the longer cycle.
PASS: budget + trigger + a clear path to purchase at this price
BORDERLINE: one is weak → name the mitigation (champion to build the case, land-small-expand)
FAIL: price assumes authority/budget the buyer doesn't have
→ "The price is technically defensible but unbuyable by this ICP. Reprice to
the buyer's authority, or change who you sell to."
Outputs
Output
Format
Description
Value metric decision
Markdown block
The chosen charging unit + why it scales with value
Price triangulation
Block
The number/range with the value math, cost floor, competitor reference
Packaging / tiers
Table
Tiers mapped to segments + upgrade triggers (+ free tier if PLG)
Unit-economics verdict
Checklist
LTV/CAC, payback, margin — PASS/BORDERLINE/FAIL with the numbers
Buy-readiness verdict
Checklist
Authority / trigger / procurement
Brain write (if connected):
brain/write: playbooks/pricing.md (price, value metric, packaging, the value math),
decisions/ (why this metric + price — for later repricing)
Anti-patterns
Anti-pattern
Why it fails
Fix
Cost-plus pricing
Ignores value; systematically underprices high-value products and caps your upside
Anchor to quantified value; use cost only as the floor
"We'll figure out pricing later"
Price shapes the product, the motion, and the buyer — deferring it builds the wrong everything
Decide a defensible v1 price now; iterate with data
Pricing only against a competitor
A race to the bottom that throws away the value story; assumes their economics are yours
Reference the competitor; set the number from your value
A value metric that doesn't scale with value
Flat fee while customer value grows 10× leaves money on the table and overcharges the small
Pick a unit that grows as the customer succeeds
Too many tiers / feature-salad packaging
Choice overload depresses conversion; no segment can see "the one for me"
≤3–4 tiers, each mapped to a real segment with a clear upgrade reason
Bundling a "killer" feature into the entry tier
A killer feature destroys WTP for the segment that doesn't want it; drags the whole price down
Isolate killers to higher tiers or remove
Scaling acquisition before unit economics hold
LTV/CAC < 1 means every new customer loses money; scale multiplies the loss
Pass the Step 4 gate before any paid-acquisition scale
Discounting to close with no floor
Trains buyers to wait for discounts and erodes margin and price integrity
Hold the cost-margin floor; trade discount for term/volume/commitment
Pricing before positioning is FINAL
A price with no value story behind it is indefensible in the deal and brittle to reposition
Gate 0: positioning FINAL first
Benchmarks
Dated and sourced calibration anchors — not guarantees; calibrate to your model.
Metric
Benchmark
Source
LTV/CAC (healthy)
≥ 3
David Skok, SaaS Metrics 2.0 (classic SaaS framework, forEntrepreneurs)
CAC payback (SMB/PLG)
< 12 months
David Skok, SaaS Metrics 2.0
CAC payback (enterprise)
< 18 months (financeable to ~24 with strong NRR)
Skok ≤12-mo principle extended to enterprise by common practitioner convention
Value capture (B2B software)
Illustrative ~10–25% of quantified value — a rule of thumb, not a sourced benchmark
Value-based pricing practice — Patrick Campbell / ProfitWell; van Westendorp Price Sensitivity Meter — value anchoring and WTP testing (Steps 2–3).
April Dunford, Obviously Awesome (2019) — pricing follows positioning; the competitive alternative is also the price reference frame (Gate 0, Step 2).
Note on the soft ranges: the value-capture % and free→paid figures are Fieldwork
heuristics / diagnostics, not sourced benchmarks (labeled as such above). They are starting
points to test against your own data, not numbers to anchor on.
Scope boundary (related skills): this skill is general value-based pricing. For
AI-specific monetization — token/credit economies, usage-based metering, freemium calibration
for AI products — use pmm/ai-product-gtm/ai-usage-pricing-growth instead (or alongside, for
the AI-usage layer). For the decision to scale a paid channel on these unit economics, hand
off to growth/paid-channel-fit.
