关于「四十八小时成交速度:管线是一条队列,成交概率会衰减」,《三档定价与四十八小时成交》给出的定量结论是什么?
结论:利特尔定律说明管线长度由进入速度与停留时间决定:套餐路线停留两天,管线挂着 8.40 个机会,拖到九天就是 37.81 个,跟进时间从 1.26 小时涨到 5.67 小时,成交一单没多。示意模型中每晚一天,成交概率乘以 0.7558,四十八小时内外的合成成交率是 0.317 对 0.092;可衰减速度取决于 τ,τ 为 30 天时增益只剩 0.088。观察台账分不开速度与急迫,检验必须随机安排报价时间。截止日必须来自真实档期,过期不涨价、不重置。开工即收款让现金周期变为负 18 天,比定制 60 天账期释放 RM 203,300 的营运资金。
数据口径:把赢面、销售小时与毛利放进同一张账,示意参数下定制单金额是固定套餐加权客单价的 4.68 倍,每销售小时毛利却只有 USD 13.78,套餐是 USD 87.94,相差 6.38 倍;定制单要涨到约 USD 26,800 才能追平。 核对文献之后,本书对立项蓝图做了五处修正:所谓「零风险试点」改为风险有上限的试点;「选项过多会吓跑买方」不是定律,三档的理由改为折中效应、真实锚点与推迟倾向,并如实报告理性模型中第二档每位潜在买方只多带来 USD 4.75;四十八小时从成交纪律改为待检验假说,示意衰减下每晚一天成交概率乘以 0.7558;刷卡不比开工日发票更赚钱,每场会面期望净额 USD 269.96 对 USD 275.95,它的价值在于钱早到;预付款是负债,平均约 RM 93,942 尚未赚到。
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: 0.088, 0.092, 0.317, 0.7558, 30
| model close probability by delay of the decision request: | |
| 0d 0.420 | 1d 0.317 2d 0.240 3d 0.181 5d 0.104 7d 0.059 14d 0.008 |
| daily hazard ratio exp(-1/TAU - MU) | 0.7558 |
| synthetic ledger, close rate by delay bin: | |
| 0-1 days n | 740 close 0.358 95% +-0.035 |
| 1-2 days n | 756 close 0.276 95% +-0.032 |
| 2-4 days n | 385 close 0.190 95% +-0.039 |
| 4-7 days n | 542 close 0.090 95% +-0.024 |
| 7-10 days n | 577 close 0.028 95% +-0.013 |
| within 48 h 0.317 (n 1496), later 0.092 (n 1504); difference 0.225 +-0.028 | |
| TAU | 2 days: close at 1 day 0.247, at 5 days 0.030, gain 0.218 |
| TAU | 4 days: close at 1 day 0.317, at 5 days 0.104, gain 0.214 |
| TAU | 10 days: close at 1 day 0.369, at 5 days 0.219, gain 0.150 |
| TAU | 30 days: close at 1 day 0.394, at 5 days 0.306, gain 0.088 |
查看完整验算程序
# Chapter 5. p52: time decay of close probability and what a 48-hour rule would gain.
# Model: after the first meeting a buyer's interest decays as exp(-t / TAU) and
# the opportunity dies from outside causes (budget moved, rival chosen) at rate
# MU a day. If the written offer and decision request arrive at delay t days,
# the buyer says yes with probability P0 * exp(-t / TAU) * exp(-MU * t).
# A synthetic ledger of 3,000 opportunities with random delays is then binned
# the way a real ledger would be. All parameters are illustrative.
import math
import random
P0, TAU, MU = 0.42, 4.0, 0.03
def close(t, tau=TAU, mu=MU):
return P0 * math.exp(-t / tau) * math.exp(-mu * t)
print("model close probability by delay of the decision request:")
print(" " + " ".join("{:g}d {:.3f}".format(t, close(t)) for t in (0, 1, 2, 3, 5, 7, 14)))
print("daily hazard ratio exp(-1/TAU - MU) = {:.4f}".format(math.exp(-1 / TAU - MU)))
rng = random.Random(48)
rows = []
for _ in range(3000):
t = rng.choice([rng.uniform(0, 2), rng.uniform(2, 10)])
rows.append((t, rng.random() < close(t)))
BINS = [(0, 1), (1, 2), (2, 4), (4, 7), (7, 10)]
print("synthetic ledger, close rate by delay bin:")
for lo, hi in BINS:
sel = [y for t, y in rows if lo <= t < hi]
p = sel.count(True) / len(sel)
se = math.sqrt(p * (1 - p) / len(sel))
print(" {:>2}-{:<2} days n {:>4} close {:.3f} 95% +-{:.3f}".format(
lo, hi, len(sel), p, 1.96 * se))
a = [y for t, y in rows if t < 2]
b = [y for t, y in rows if t >= 2]
pa, pb = a.count(True) / len(a), b.count(True) / len(b)
se = math.sqrt(pa * (1 - pa) / len(a) + pb * (1 - pb) / len(b))
print("within 48 h {:.3f} (n {}), later {:.3f} (n {}); difference {:.3f} +-{:.3f}".format(
pa, len(a), pb, len(b), pa - pb, 1.96 * se))
for tau in (2.0, 4.0, 10.0, 30.0):
gain = close(1.0, tau) - close(5.0, tau)
print(" TAU {:>4.0f} days: close at 1 day {:.3f}, at 5 days {:.3f}, gain {:.3f}".format(
tau, close(1.0, tau), close(5.0, tau), gain))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《企业服务产品化封装、低门槛试点定金锚定与四十八小时闪电成交母典》第 5 章「四十八小时成交速度:管线是一条队列,成交概率会衰减」(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-352 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026352,
author = {Fan, Yafu},
title = {Productised B2B Service Packaging, Low-Friction Pilot Deposit Anchoring and Forty-Eight-Hour Deal Velocity},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}