What does Long-Video Half-Lives and Deep Trust conclude, in numbers, about “The Description-Box Tollbooth: a Chain of Multiplications, Coupled Elasticities, Sample Sizes and Disclosure”?

Answer: Chapter 6 writes the description box as a chain of multiplications and computes coupled elasticities, sample sizes and the uncertainty of three months of data, 0.967 of which comes from the sales stage.

Scope of the numbers: Parameters are illustrative and all calculations are generated by the book's programs. This is general business and marketing analysis, not marketing or investment advice; platform rules change.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.967

point estimate per 1,000 views14.70 units
posterior P10 10.14 P50 15.73 P90 23.08 P90/P10 2.28
share of log-variance by stage:
expand0.004
click0.029
land0.967
Show the full program
# Chapter 6. p63: how sure is "revenue per 1,000 views" after 3 months?
# Illustrative counts for one video over three months: 30,000 views,
# 2,350 description expansions, 290 link clicks, 9 sales at 49 units.
# Each stage rate gets a Beta(1 + hits, 1 + trials - hits) posterior;
# 20,000 draws (fixed seed) give the band of revenue per 1,000 views
# and the share of log-variance owed to each stage.
import math
import random

RNG = random.Random(20261120)
COUNTS = [("expand", 30000, 2350), ("click", 2350, 290), ("land", 290, 9)]
PRICE = 49.0
N = 20000


def q(xs, p):
    return xs[int(p * (len(xs) - 1))]


if __name__ == "__main__":
    draws = {name: [] for name, _, _ in COUNTS}
    rpm = []
    for _ in range(N):
        prod = 1000.0 * PRICE
        for name, n, x in COUNTS:
            r = RNG.betavariate(1 + x, 1 + n - x)
            draws[name].append(math.log(r))
            prod *= r
        rpm.append(prod)
    point = 1000.0 * PRICE
    for _, n, x in COUNTS:
        point *= x / n
    rpm.sort()
    print("point estimate per 1,000 views: {:.2f} units".format(point))
    print("posterior P10 {:.2f}  P50 {:.2f}  P90 {:.2f}  P90/P10 {:.2f}"
          .format(q(rpm, 0.1), q(rpm, 0.5), q(rpm, 0.9),
                  q(rpm, 0.9) / q(rpm, 0.1)))
    var = {}
    for name in draws:
        xs = draws[name]
        m = math.fsum(xs) / N
        var[name] = math.fsum((v - m) ** 2 for v in xs) / (N - 1)
    tot = math.fsum(var.values())
    print("share of log-variance by stage:")
    for name, _, _ in COUNTS:
        print("  {:<7} {:.3f}".format(name, var[name] / tot))

Single source and citation

Fan, Y. (2026). Long-Form Video Search Half-Lives, Deep-Trust Pre-Selling and Multi-Year Passive Acquisition Assets (Version 1.0.0), Chapter 6: The Description-Box Tollbooth: a Chain of Multiplications, Coupled Elasticities, Sample Sizes and Disclosure. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-370 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026370,
  author    = {Fan, Yafu},
  title     = {Long-Form Video Search Half-Lives, Deep-Trust Pre-Selling and Multi-Year Passive Acquisition Assets},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book