关于「两部制:保底月费、按量计费、阶梯与封顶」,《托管月费与价值锚定抽成》给出的定量结论是什么?
结论:两部制把月费拆成两个数,各管一件事:入门费覆盖固定成本、分配双方都确定的那部分价值;单价分配增长,也分配萎缩的风险。五份在 1,500 单时都收 RM 1,900 的合同,到了 4,500 单,价值占比从 0.114 散到 0.341。Oi 的迪士尼困境说明,客户规模不同时单价要高于边际成本,而统一的两部制会诱使供应商放弃小客户;解法是菜单。保底有三种形状,差别在淡季与边际价格。计量单位要与价值同涨同落,座位数恰恰相反。「衡」档的递减阶梯就是两份两部制的下包络,加一个保底与一个封顶。RM 4,800 的封顶只让桥算在快速增长的路径上放弃 0.033 的收入,却把最大的单月跳升从 RM 3,487.30 压到 RM 2,127.96。客户自选方案时会偏向包月,一年一次的最优账单追溯是成本最低的修正。下一章接着讨论按服务费抽成的混合合同,以及激励在其中怎样对齐、又怎样错位。
数据口径:书中的公司与数字全部虚构,合同与法规只写马来西亚与台湾的一般框架,不是法律、税务或投资意见。 书中以标注框区分已发表的证据、可运行的计算、原典与作者的假说,每条假说都写明什么结果会推翻它。
对照数据表
下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.033, 1900, 2127.96, 3487.30, 4800
| budget line RM 2,375.00 | |||||
| plan | mean CV | months over budget | largest monthly jump | revenue 36 mo | in +4% paths |
| flat 1,900 | 0.000 | 0.0000 | 0.00 | 68,400 | 68,400 |
| pure 1.28/order | 0.311 | 0.4126 | 6,974.61 | 97,822 | 154,762 |
| 1,000 + 0.60/order | 0.180 | 0.2827 | 3,269.35 | 81,854 | 108,545 |
| hybrid B, no cap | 0.188 | 0.3621 | 3,487.30 | 90,550 | 123,695 |
| hybrid B, cap 4,800 | 0.173 | 0.3621 | 2,127.96 | 89,451 | 119,580 |
| hybrid B, cap 3,500 | 0.142 | 0.3621 | 1,534.14 | 85,908 | 107,313 |
| hybrid B, cap 4,800 gives up RM 4,115 over 36 months in +4% paths (0.033 of the uncapped bill) | |||||
| hybrid B, cap 3,500 gives up RM 16,382 over 36 months in +4% paths (0.132 of the uncapped bill) |
查看完整验算程序
# Chapter 3. p35: How predictable each bill is for Lanyu, and what a cap costs Bridgecalc.
# 2,000 simulated 36-month volume paths, seed 135 (random.Random instance): the p32 seasonality x
# log-normal month noise (sigma 0.12) x one growth rate per path drawn from -1%, 0%, +2%, +4% a month with
# equal weight. Budget line = 1.25 x RM 1,900. Plans from p12 and p32; volumes and noise are ASSUMED.
import math
import random
RNG = random.Random(135)
SEASON = [0.85, 0.80, 0.95, 1.00, 0.95, 0.90, 0.95, 1.00, 1.05, 1.10, 1.35, 1.30]
RUNS, MONTHS, BUDGET = 2000, 36, 1.25 * 1900.0
def hybrid(q, cap):
return min(cap, 1900.0 + 0.96 * max(0.0, min(q, 3000) - 1500) + 0.64 * max(0.0, q - 3000))
PLANS = {
"flat 1,900": lambda q: 1900.0,
"pure 1.28/order": lambda q: 1.28 * q,
"1,000 + 0.60/order": lambda q: 1000.0 + 0.60 * q,
"hybrid B, no cap": lambda q: hybrid(q, math.inf),
"hybrid B, cap 4,800": lambda q: hybrid(q, 4800.0),
"hybrid B, cap 3,500": lambda q: hybrid(q, 3500.0),
}
def cv(xs):
m = math.fsum(xs) / len(xs)
return math.sqrt(math.fsum((x - m) ** 2 for x in xs) / (len(xs) - 1)) / m
if __name__ == "__main__":
paths = []
for _ in range(RUNS):
g = RNG.choice((-0.01, 0.0, 0.02, 0.04))
paths.append((g, [1500 * (1 + g) ** t * SEASON[t % 12] * math.exp(RNG.gauss(0, 0.12) - 0.0072)
for t in range(MONTHS)]))
print(" budget line RM {:,.2f}".format(BUDGET))
print(" plan mean CV months over budget largest monthly jump revenue 36 mo"
" in +4% paths")
rev4 = {}
for name, f in PLANS.items():
cvs, over, jump, rev, r4 = [], 0, 0.0, [], []
for g, qs in paths:
bills = [f(x) for x in qs]
cvs.append(cv(bills))
over += sum(1 for b in bills if b > BUDGET)
jump = max(jump, max(b - a for a, b in zip(bills, bills[1:])))
rev.append(math.fsum(bills))
if g == 0.04:
r4.append(math.fsum(bills))
rev4[name] = math.fsum(r4) / len(r4)
print(" {:19} {:8.3f} {:19.4f} {:21,.2f} {:14,.0f} {:13,.0f}".format(
name, math.fsum(cvs) / RUNS, over / (RUNS * MONTHS), jump, math.fsum(rev) / RUNS, rev4[name]))
for cap in ("hybrid B, cap 4,800", "hybrid B, cap 3,500"):
lost = rev4["hybrid B, no cap"] - rev4[cap]
print(" {} gives up RM {:,.0f} over 36 months in +4% paths ({:.3f} of the uncapped bill)".format(
cap, lost, lost / rev4["hybrid B, no cap"]))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《技术底座托管月费、价值锚定抽成混合契约与客户终身毛利锁定母典》第 3 章「两部制:保底月费、按量计费、阶梯与封顶」(版本 0.1.0). 5D Library. DOI 待发布
DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-135 · v0.1.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026135,
author = {Fan, Yafu},
title = {Hosted-Platform Retainers, Value-Anchored Revenue-Share Hybrids and Customer Lifetime Gross-Margin Lock-In},
year = {2026},
publisher = {5D Library},
edition = {0.1.0},
note = {DOI pending}
}