关于「触达是串联乘积:「三天触达数万」的上下界」,《异业联营拆账与频次保护》给出的定量结论是什么?

结论:模拟里抽 100 人,活跃 48 人,比例 0.480,威尔逊区间 0.385 到 0.577,有效规模在 4,616 到 6,922 之间,区间很宽。抽 385 人,活跃 154 人,比例 0.400,区间 0.352 到 0.450,有效规模 4,227 到 5,396。抽 1,068 人,活跃 443 人,比例 0.415,区间 0.386 到 0.445,有效规模 4,627 到 5,335。

数据口径:纳什议价解先还双方的破裂点,再平分合作红利,五五分只是破裂点对称时的特例;示意参数下,渠道方单干 2,000、操盘方单干 6,000 时,渠道方份额是 0.433。 频次方面,示意名单推 8 次比推 3 次成交多不到一倍,流失成员却是 7.453 倍;一次性操盘方的最优推送会让十二个月联合价值损失 0.1468。

English version

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: 0.352, 0.385, 0.386, 0.400, 0.415, 0.445

z1.9600; sample size for margin e (worst case a = 0.5)
e0.10: n=97 e=0.05: n=385 e=0.03: n=1,068
reported N12,000; simulated true active share 0.41
n active share wilson lo wilson hi effective N
10048 0.480 0.385 0.577 4,616-6,922
385154 0.400 0.352 0.450 4,227-5,396
1,068443 0.415 0.386 0.445 4,627-5,335
effective/reported at n1,068: 0.415; upper 0.445 < 0.70 -> YELLOW
查看完整验算程序
# Chapter 4. p42: auditing a self-reported audience size by sampling.
# The channel reports N members. Without handing over any list, the
# channel draws n members at random (by its own tool) and reports how
# many opened anything in the last 30 days. Active share a is estimated
# with a Wilson 95 % interval; "effective size" = N * a. Sample size for
# a margin e at 95 %: n = z^2 a (1 - a) / e^2, worst case a = 0.5.
# Simulated audit with a fixed seed; the true active share is a
# parameter of the simulation, not a measured fact.
import math
import random
from statistics import NormalDist

Z = NormalDist().inv_cdf(0.975)


def wilson(k, n):
    p = k / n
    den = 1.0 + Z * Z / n
    mid = (p + Z * Z / (2 * n)) / den
    half = Z * math.sqrt(p * (1 - p) / n + Z * Z / (4 * n * n)) / den
    return mid - half, mid + half


def need(e, a=0.5):
    return math.ceil(Z * Z * a * (1 - a) / (e * e))


if __name__ == "__main__":
    rng = random.Random(20261120)
    print("z = {:.4f}; sample size for margin e (worst case a = 0.5)"
          .format(Z))
    print("  " + "  ".join("e={:.2f}: n={:,}".format(e, need(e))
                           for e in (0.10, 0.05, 0.03)))
    n_rep, true_a = 12000, 0.41
    print("reported N = {:,}; simulated true active share {:.2f}"
          .format(n_rep, true_a))
    print("   n   active   share   wilson lo  wilson hi   effective N")
    for n in (100, 385, 1068):
        k = 0
        for _ in range(n):
            if rng.random() < true_a:
                k += 1
        lo, hi = wilson(k, n)
        print("{:>5,} {:>7} {:>8.3f} {:>10.3f} {:>10.3f}  {:>6,.0f}-{:,.0f}"
              .format(n, k, k / n, lo, hi, n_rep * lo, n_rep * hi))
    ratio = k / n
    flag = "YELLOW" if hi < 0.70 else "GREEN"
    print("effective/reported at n = {:,}: {:.3f}; upper {:.3f} < 0.70 -> {}"
          .format(n, ratio, hi, flag))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《异业私域联营拆账、技术换流量零本金借船与联合发售脉冲母典》第 4 章「触达是串联乘积:「三天触达数万」的上下界」(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-371 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026371,
  author    = {Fan, Yafu},
  title     = {Joint-Venture Audience Monetisation, Tech-for-Traffic Barter Structures and Co-Launch Revenue Waterfalls},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论