| name | clarity-pricing |
| description | Stage 5 of the stranger test — audit the pricing page against the behavioral-economics literature (anchoring, decoy/compromise effects, loss aversion, mental accounting, pain of paying) and the catalog of SaaS pricing failures. Distinguishes ethical architecture from dark-pattern manipulation. Scored 0-10. |
clarity-pricing — "What does it cost me, and will I regret it?"
The pricing page is where the visitor switches from evaluating the product
to forecasting their own regret. The behavioral literature explains both
how prices are perceived (anchors, frames, mental accounts) and how pricing
pages destroy trust (opaque units, redefined limits, decoy manipulation).
This stage audits for clarity and predictability first — the 2025–26
agent-tool repricing crises made predictability the scarcest asset a pricing
page can display — and treats manipulative architecture as a defect, not
a technique to recommend.
Consume the inventory's pricing sections and number census.
Rubric
| Check | Weight | What to verify |
|---|
| The 10-second answer | ×3 | A visitor with a use case answers: what will I pay, what do I get, what is never charged? Without a calculator, a glossary, or fine print. |
| Anchor honesty (Kahneman) | ×2 | The first number seen frames all others. Is the anchor a real, relevant price — not an inflated strike-through or a decoy nobody buys? |
| Unit legibility | ×2.5 | The billing unit maps to something the buyer can count in their own world (seats, projects). Invented units (credits, ACUs, "tasks", compute points) that require conversion are the #1 modern failure — audit whether a buyer can forecast next month's bill within ±20%. |
| Value-metric alignment | ×2 | The meter charges what the customer values more of, not what they fear doing. Metering the action the product exists to encourage (usage of the core loop) taxes adoption. |
| Tier architecture (Simonson) | ×1.5 | Tiers map to buyer situations, not feature-count ladders. Compromise/decoy effects present? Flag decoys as manipulation. 3–4 real options max (choice overload beyond that — Schwartz/Iyengar, with replication caveats). Upgrade triggers statable in one sentence each. |
| Frame & fairness (Thaler) | ×1.5 | Gains framed as gains, losses not disguised; no penalty-framed downgrades; mental accounting respected (the bill lands in the account the buyer budgets from). Dual entitlement: price changes justified by cost/value, never silently redefined mid-subscription. |
| Predictability | ×2 | Fine print stable, limits defined in the open, "unlimited" meaning unlimited, no quarterly redefinitions. Any usage-based element must show worst-case exposure, not just best-case. |
Anchored bands
- 9–10 — 10-second answer passes; units countable in the buyer's world;
no decoys; worst-case bill statable; the page reads as a commitment, not
an opening bid.
- 7–8 — Answer within 30 seconds; one conversion required (annual/
monthly, per-seat math); architecture honest.
- 5–6 — Invented unit with a conversion table; or tiers as feature
ladders forcing a matrix comparison; or one manipulative element
(decoy tier, fake strike-through anchor).
- 3–4 — Bill not forecastable; "contact us" hiding all real prices for
self-serve-sized buyers; limits defined only in fine print.
- 0–2 — Terms contradict the page's own claims, or historical
redefinitions visible (the trust-destroying pattern).
Findings
Directions: unit-countable, anchor-honest, decoy-out, worst-case-shown,
trigger-sentence, meter-off-core-loop, fine-print-surfaced. Evidence =
quoted pricing copy. Never advise adding a decoy, artificial scarcity, or
strike-through theater — this stage's ethics are part of its rubric: what
Ariely documents as effective is audited here as costly, because
detected manipulation converts to churn and reputation damage in
subscription businesses where the relationship is repeated, not one-shot.
Doctrine suppression: a site whose pricing doctrine forbids comparison cards
or toggles is audited for clarity within its chosen genre.
Full citations and the failure catalog: references/pricing-psychology.md.
