关于「升单棘轮:首单 RM 19.90 抵扣进 RM 199 的期望值与退款风险」,《弃单召回与升单棘轮》给出的定量结论是什么?
结论:λ 为 0.5 时,情况完全不同。中间档定在 RM 49,阶梯收入升到 RM 36.793,比单级多 +13.501,首单买家中有 0.090 读完中间档后升到顶级。中间档定在 RM 59,阶梯多 +11.355,走完全程者占 0.079。中间档定在 RM 79,阶梯多 +11.796。λ 为 1.0 时,RM 49 一档多 +27.399,走完全程者占 0.182。
数据口径:按本书的示意模型,第一波召回放在一分钟时只挽回同意离开者的 0.00609,放在十五分钟时挽回 0.03732;把三波压在一小时之内,每位离开者净值 RM 1.1790,按十五分钟、二十四小时、七十二小时发送时为 RM 1.9419;第五波的净值已是 -0.0121。 书中网站被当作付费报告产品讨论,不宣称任何预测准确性,不用恐惧招揽;所有访客与订单数据均为合成或示意。
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: 0.079, 0.090, 0.182, 0.5, 1.0, 11.355
| single credited step to RM 199: RM 23.292 per entry buyer | |||||||||
| middle | share | lam | RM/buyer | vs single | stop | mid only | full | mid->full | |
| RM | 49 | 0.35 | 0.0 | 23.353 | +0.061 | 0.608 | 0.313 | 0.080 | 0.000 |
| RM | 49 | 0.35 | 0.5 | 36.793 | +13.501 | 0.608 | 0.223 | 0.080 | 0.090 |
| RM | 49 | 0.35 | 1.0 | 50.691 | +27.399 | 0.608 | 0.130 | 0.080 | 0.182 |
| RM | 59 | 0.45 | 0.0 | 23.657 | +0.365 | 0.626 | 0.309 | 0.065 | 0.000 |
| RM | 59 | 0.45 | 0.5 | 34.647 | +11.355 | 0.626 | 0.230 | 0.065 | 0.079 |
| RM | 59 | 0.45 | 1.0 | 46.477 | +23.185 | 0.626 | 0.146 | 0.065 | 0.163 |
| RM | 79 | 0.45 | 0.0 | 23.549 | +0.258 | 0.783 | 0.128 | 0.089 | 0.000 |
| RM | 79 | 0.45 | 0.5 | 35.087 | +11.796 | 0.783 | 0.032 | 0.089 | 0.096 |
| RM | 79 | 0.45 | 1.0 | 38.909 | +15.618 | 0.783 | 0.000 | 0.089 | 0.128 |
查看完整验算程序
# Chapter 5. p52: a ratchet ladder against a single upgrade step.
# Every rung credits all money already paid, and a buyer can only move up.
# Entry buyers value the full volume at W, lognormal, median RM 65, log-sd
# 0.9 (as p51). A middle rung (three of the eight parts) is worth share s
# of W. At each offer the buyer takes the option with the largest surplus
# value - price, if positive. A buyer who took the middle rung may value
# the rest more after reading it: factor (1 + lam), lam = 0, 0.5 or 1.0
# (an assumption to be tested, not a finding). Integration over 20,000
# quantiles of W (deterministic). Prices and shares are illustrative.
import math
from statistics import NormalDist
ND = NormalDist()
MED, SD = 65.0, 0.9
ENTRY, FULL = 19.90, 199.0
QS = [MED * math.exp(SD * ND.inv_cdf((i + 0.5) / 20000)) for i in range(20000)]
def single(w):
return FULL - ENTRY if w >= FULL - ENTRY else 0.0
def ladder(w, mid, share, lam):
# first offer: stop, middle rung (mid - ENTRY) or full (FULL - ENTRY)
s_mid = share * w - (mid - ENTRY)
s_full = w - (FULL - ENTRY)
if max(s_mid, s_full) <= 0:
return 0.0, "stop"
if s_full >= s_mid:
return FULL - ENTRY, "full"
# second offer, credited: the rest of the volume for FULL - mid
if (1.0 - share) * w * (1.0 + lam) >= FULL - mid:
return FULL - ENTRY, "mid-then-full"
return mid - ENTRY, "mid"
if __name__ == "__main__":
base = math.fsum(single(w) for w in QS) / len(QS)
print("single credited step to RM 199: RM {:.3f} per entry buyer".format(base))
print("middle share lam RM/buyer vs single stop mid only full mid->full")
for mid, share in ((49.0, 0.35), (59.0, 0.45), (79.0, 0.45)):
for lam in (0.0, 0.5, 1.0):
res = [ladder(w, mid, share, lam) for w in QS]
rev = math.fsum(r[0] for r in res) / len(res)
cnt = {k: len([1 for r in res if r[1] == k]) / len(res)
for k in ("stop", "mid", "full", "mid-then-full")}
fmt = "RM {:>3.0f} {:.2f} {:.1f} {:.3f} {:+.3f} {:.3f} {:.3f} {:.3f} {:.3f}"
print(fmt.format(mid, share, lam, rev, rev - base, cnt["stop"], cnt["mid"],
cnt["full"], cnt["mid-then-full"]))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《诊断收费站弃单挽回时序、个性化悬念召回与高客单升单棘轮母典》第 5 章「升单棘轮:首单 RM 19.90 抵扣进 RM 199 的期望值与退款风险」(版本 1.0.0). 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-349 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026349,
author = {Fan, Yafu},
title = {Paywall Abandonment Recovery Timing, Personalised Curiosity Recall and High-Ticket Upsell Ratchets},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}