关于「排名、点击与直接输入:地域精确名的护城河从哪里来」,《域名池续费剪枝》给出的定量结论是什么?

结论:第五章用级联模型拆解排名、点击与直接输入,检出点击意愿提升两成每组需要 3,837 次展示。

数据口径:示意参数下,触发倍数为 2.893,续费 15 时下边界为 599、上边界为 16,849;42 枚示意名字里有 23 枚应当弃续,期权部分占最优价值的 0.861,合理续费只需全额 1,340 中的一小部分。 核对原文之后,本书做了蓝图修正:2012 年的精确匹配域名更新针对的是低质量网站,名字不带排名加成只是稳健性假定;林登塔尔的重复交易指数只反映成交过的名字;不依赖排名的访问在示意参数下约占 0.309,是可靠的地板而不是流量的大头。

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: 3837

A. impressions per arm, no clustering
u0.05: p1 = 0.1051, n = 57,699
u0.10: p1 = 0.1101, n = 14,735
u0.20: p1 = 0.1201, n = 3,837
u0.40: p1 = 0.1401, n = 1,034
B. cities per arm, 2,400 impressions per cell, u0.20
ICC 0.000: design effect1.00, cells 2
ICC 0.002: design effect5.80, cells 10
ICC 0.010: design effect24.99, cells 40
ICC 0.050: design effect 120.95, cells194
查看完整验算程序
# Chapter 5. p53: how many impressions it takes to see a match uplift.
# Test: does a query-matching domain lift click-through at the same
# rank? Compare CTR p0 of non-matching sites with p1 = p0 (1 + u) of
# matching sites at rank 3 (p0 = 0.1001 from p51). Two-sided alpha
# 0.05, power 0.80, normal approximation for two proportions. Clicks
# cluster by city and query, so the design effect 1 + (m - 1) ICC
# inflates the count; m = impressions per city-query cell. Values
# illustrative.
import math
from statistics import NormalDist

Z = NormalDist()


def n_two_props(p0, p1, alpha=0.05, power=0.80):
    za = Z.inv_cdf(1 - alpha / 2)
    zb = Z.inv_cdf(power)
    pb = (p0 + p1) / 2
    num = (za * math.sqrt(2 * pb * (1 - pb))
           + zb * math.sqrt(p0 * (1 - p0) + p1 * (1 - p1))) ** 2
    return math.ceil(num / (p1 - p0) ** 2)


if __name__ == "__main__":
    p0 = 0.1001
    print("A. impressions per arm, no clustering")
    for u in (0.05, 0.10, 0.20, 0.40):
        n = n_two_props(p0, p0 * (1 + u))
        print("   u = {:.2f}: p1 = {:.4f}, n = {:>7,}".format(
            u, p0 * (1 + u), n))
    print("B. cities per arm, 2,400 impressions per cell, u = 0.20")
    n = n_two_props(p0, p0 * 1.2)
    for icc in (0.0, 0.002, 0.01, 0.05):
        deff = 1 + (2400 - 1) * icc
        cells = math.ceil(n * deff / 2400)
        print("   ICC {:.3f}: design effect {:>6.2f}, cells {:>4}".format(
            icc, deff, cells))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《精确地域域名实物期权估值、续费剪枝最优停时与自然搜索垄断护城河母典》第 5 章「排名、点击与直接输入:地域精确名的护城河从哪里来」(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-341 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026341,
  author    = {Fan, Yafu},
  title     = {Real-Options Valuation of Geo-Exact Domain Portfolios, Optimal Renewal Pruning and the Moats of Organic Search},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论