关于「实践与证伪方案:客服收口技能包与预注册检验」,《异议柔道与定金锁单》给出的定量结论是什么?
结论:第八章给出客服收口技能包、红线与转人工条件,示意模拟中技能包的净毛利是施压剧本的 1.809 倍,并给出每组 1,251 条线索的预注册检验方案。书中以标注框区分已发表的证据、可运行的 Python 计算(19 段程序)、原典与作者的假说。
数据口径:书中的顾客群、概率与金额均为示意;全部计算由书中程序生成;本书不构成个人化的法律、财务或投资建议,只讨论守规矩的说服。 第一章的判别树说明,在示意先验下,说「太贵了」的人有 0.734 并不缺钱。
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: 1.809
| style | kept | cancel | complaint | block | after-no | net margin |
|---|---|---|---|---|---|---|
| JUDO | 313.2 | 15.2 | 1.4 | 12.8 | 0.0 | 135.55 |
| PUSH | 300.2 | 51.6 | 30.3 | 76.0 | 72.2 | 74.92 |
| JUDO net margin is 1.809 times PUSH |
查看完整验算程序
# Chapter 8. p81: The skill pack against a pressure playbook, on 20,000
# synthetic objection leads (shares 0.46 / 0.37 / 0.17 for "too
# expensive" / "let me think" / "ask my partner"; 0.03 of leads show a
# vulnerability signal). Path parameters are illustrative and echo
# Chapters 2 to 6. JUDO: probe then matched reply, a 3-day refundable
# deposit, at most 2 follow-ups in daytime, stop on "no", a brief for
# both partners, a human for vulnerable leads. PUSH: reflexive 10 % cut,
# 5 follow-ups at any hour, keeps writing after a "no" half the time,
# presses one partner to sign, treats every lead the same.
# Net = margin of kept sales - 0.05 per cancellation - 0.60 per
# complaint - 0.30 per block. Figures per 1,000 leads.
import math
import random
# close margin cancel complain|cancel
PATH = {
("JUDO", "price"): (0.37, 0.45, 0.04, 0.10),
("JUDO", "think"): (0.30, 0.45, 0.05, 0.10),
("JUDO", "partner"): (0.26, 0.45, 0.05, 0.10),
("JUDO", "human"): (0.10, 0.45, 0.00, 0.00),
("PUSH", "price"): (0.40, 0.35, 0.08, 0.30),
("PUSH", "think"): (0.34, 0.45, 0.15, 0.30),
("PUSH", "partner"): (0.27, 0.45, 0.40, 0.30),
}
FOLLOW = {"JUDO": (2, 0.02, 0.0), "PUSH": (5, 0.05, 0.5)}
N = 20000
def run(style, rng):
t = {"kept": 0, "cancel": 0, "complaint": 0, "block": 0,
"after_no": 0, "margin": []}
for _ in range(N):
u = rng.random()
kind = "price" if u < 0.46 else "think" if u < 0.83 else "partner"
vulnerable = rng.random() < 0.03
path = "human" if (vulnerable and style == "JUDO") else kind
close, m, cancel, comp = PATH[(style, path)]
if vulnerable and style == "PUSH":
comp = min(1.0, comp * 3)
if path == "think":
cap, block, ignore_no = FOLLOW[style]
for _ in range(cap):
if rng.random() < block:
t["block"] += 1
break
if rng.random() < 0.10: # the lead says no
if rng.random() < ignore_no:
t["after_no"] += 1
if rng.random() < 0.20:
t["complaint"] += 1
continue
break
if rng.random() < close:
if rng.random() < cancel:
t["cancel"] += 1
if rng.random() < comp:
t["complaint"] += 1
else:
t["kept"] += 1
t["margin"].append(m)
net = math.fsum(t["margin"]) - 0.05 * t["cancel"] \
- 0.60 * t["complaint"] - 0.30 * t["block"]
return t, net
print("style kept cancel complaint block after-no net margin")
rng = random.Random(20261112)
res = {}
for style in ("JUDO", "PUSH"):
t, net = run(style, rng)
k = 1000.0 / N
res[style] = net * k
print("{:<5}{:>6.1f}{:>8.1f}{:>11.1f}{:>7.1f}{:>10.1f}{:>13.2f}".format(
style, t["kept"] * k, t["cancel"] * k, t["complaint"] * k,
t["block"] * k, t["after_no"] * k, net * k))
print("JUDO net margin is {:.3f} times PUSH".format(
res["JUDO"] / res["PUSH"]))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《顾问式销售异议柔道、价格嫌贵贝叶斯拆解与定金锁单停时协议》第 8 章「实践与证伪方案:客服收口技能包与预注册检验」(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-146 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026146,
author = {Fan, Yafu},
title = {Consultative Objection Judo, a Bayesian Decomposition of "Too Expensive" and Deposit-Lock Stopping-Time Protocols},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}