Pricing Power

What this page is / is not
· This page is a conditions-preparation tool: it shows you which start condition is missing, how to fix it, and what the first step is. It is not investment advice, and it is not a pricing plan.
· The thresholds come from an internal methodology doc, 《经济盈利算法·11 定价权算法 v1.0》 (domestic machine /opt/inventory/econ_algorithms/11-定价权算法.md, 162 lines, finalised 2026-09-14, read-only), and have not been through any external empirical testing. The doc's §1 cites literature only for price discrimination / bundling (Pigou 1920, Adams & Yellen 1976, Shapiro & Varian 1999); the numbers 1.5× / 2× / 2.5–5× / 40% are practice lines the methodology set for itself, and §1 further labels freemium and the anchoring effect as “provenance unconfirmed / not verified item by item”.
· Compliance red lines (the doc's §7 caveats + §9 boundaries): this page only handles version / time / quantity separation, never fraudulent pricing where the same thing costs different people different amounts; no fake original prices, no invented discounts, no resale-arbitrage separation; where price controls or price-display rules apply, follow local law.

Evidence chain (the md doc has no line numbers, so these point to sections)
· The three-price method and its three readings → 11-定价权算法.md §4 Step1; the three-tier pricing table and the 2.5–5× advanced edition → §4 Step2
· Bundling test (only negative correlation works) → §4 Step3; the separation-mechanism table and its priority order → §4 Step4
· Price-raise test (any one of three signals → probe a 10–20% rise; close rate down > 40% within 30 days → roll back) → §4 Step5
· G1–G6 and their thresholds → §3; the condition-preparation list and the wording of “first step” → §5.2 / §5.3

Not given by the doc (left blank here rather than invented)
· The doc gives no percentage thresholds for “everyone says they'd buy / half say they'd buy / almost nobody says they'd buy” → this page only lists your actual acceptance ratios at the three price points and draws no line for you.
· A remedy for G5: the checklist in §5.2 has only five lines, G1/G2/G3/G4/G6, with no G5 → when G5 is missing this page simply marks “not given by the doc” and offers no remedy.
· The doc gives no rank for “region” in the priority order (the order stops at identity difference) → flagged by this page.

Where it is not aligned / has been simplified (said plainly)
· Unconfirmed ①: this page takes the mid price you enter to be the standard edition of the three tiers when deriving the advanced-edition range; the doc only says the standard edition should sit at the price where half of people say they'd buy, and that the advanced edition is 2.5–5× the standard, without stating outright that the mid price is the standard edition. That mapping is this page's positioning.
· Simplification ①: the doc's §4 Step3 asks you to “ask 10 random people which one they'd rather have” to judge demand correlation; this page turns the conclusion into one dropdown for you to decide.
· Simplification ②: the three price-raise signals of §4 Step5 are condensed here into “how many you hit”, rather than entered one by one.
· Not done: the doc's §7 suggests “let an LLM generate the edition names and selling points for the three tiers” —— this page is pure front-end and calls no model.
· Not done: external data such as competitors' price structures or platform price-comparison transparency (§6) is not collected here; you enter it by hand.

[Evidence section · where to check each number]
① Source: internal methodology doc 《11-定价权算法.md》v1.0 (domestic machine /opt/inventory/econ_algorithms/11-定价权算法.md, 162 lines, read-only): the start conditions G1–G6 = §3; the three-price method / three-tier pricing / bundling / separation mechanisms / price raises = §4 Step1–Step5; preparation list and first step = §5.2 / §5.3. Literature in §1 (Pigou 1920; Adams & Yellen 1976, DOI 10.2307/1886045; Shapiro & Varian 1999).
② Three categories
· [Verifiable] The G1–G6 thresholds (≥2 groups with a gap ≥1.5× / cannot be resold / ≥2 versions whose difference is function and service / one high-price anchor / unit margin ≥ 2× incremental cost / ≥3 real deals), three-tier pricing with the 2.5–5× advanced edition, 分隔优先顺序 时间差>版本差>数量差>身份差, the 10–20% probe raise and the rollback when the close rate drops >40% within 30 days, and the three readings of the three-price method —— all of these match §3 / §4.
· [This page's presentation choice] The mapping “the mid price you enter = the standard edition of the three tiers” is this page's positioning; the doc only says the standard edition should sit at the price where half of people say they'd buy, and never states that the mid price is the standard edition.
· [No source found] The numbers 1.5× / 2× / 2.5–5× / 40% carry no external citation in the md; the doc itself marks its attribution of the term freemium as “unconfirmed”; the anchoring effect (Kahneman & Tversky 1974) is marked by the doc as “not verified item by item this round” → no source found. The doc also gives no rank for “region” in the priority order.
③ How to check it yourself: three price points 50 / 100 / 200, asking 10 people each, with replies of 9 low / 5 mid / 1 high → the acceptance ratios should come out 90% / 50% / 10%; then enter 2 groups, a 1.5× gap, cannot be resold, 2 versions, an anchor, and 3 deals → G1–G6 should all pass. Whichever one you leave out is the only one that should fail.