What does Paid-Ad Scaling for Paywalls conclude, in numbers, about “Stopping at Zero: the Rule of Three, Gamma-Poisson Posteriors and Spend Red Lines”?
Answer: Part B gives the 90 % red lines. At 0.45 per click the line needs 111 clicks, a spend of 49.95 ringgit or 1.88 multiples of target. At 0.90 per click it needs 43 clicks, 38.70 ringgit, 1.46 multiples. At 1.80 per click it needs only 9 clicks, 16.20 ringgit, 0.61 multiples. With the same zero unlocks, a creative with expensive clicks should be stopped earlier, because it demands a higher unlock rate.
Scope of the numbers: Under illustrative parameters one icebreaker buyer on tarotmalaysia is worth 2.68 times the icebreaker price, and a first-order break-even stop rule captures only 0.222 of the attainable profit across forty synthetic creatives. Parameters are illustrative and all calculations are generated by the book's programs.
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: 0.45, 0.61, 0.90, 1.46, 1.80, 1.88
| A. P(CPA > T) after n clicks and zero unlocks | |
| clicks cpc 0.45 cpc 0.90 cpc 1.80 | |
| 0 | 0.3369 0.5635 0.8150 |
| 10 | 0.4413 0.6910 0.9084 |
| 20 | 0.5292 0.7812 0.9547 |
| 40 | 0.6657 0.8904 0.9889 |
| 60 | 0.7626 0.9450 0.9973 |
| 100 | 0.8803 0.9862 0.9998 |
| 150 | 0.9492 0.9975 1.0000 |
| B. clicks and spend before P(CPA > T) > | 0.90 |
| cpc 0.45 unlocks 0 | 111 clicks, spend RM 49.95 = 1.88 T |
| cpc 0.45 unlocks 1 | 204 clicks, spend RM 91.80 = 3.46 T |
| cpc 0.45 unlocks 2 | 288 clicks, spend RM 129.60 = 4.89 T |
| cpc 0.90 unlocks 0 | 43 clicks, spend RM 38.70 = 1.46 T |
| cpc 0.90 unlocks 1 | 90 clicks, spend RM 81.00 = 3.06 T |
| cpc 0.90 unlocks 2 | 132 clicks, spend RM 118.80 = 4.48 T |
| cpc 1.80 unlocks 0 | 9 clicks, spend RM 16.20 = 0.61 T |
| cpc 1.80 unlocks 1 | 32 clicks, spend RM 57.60 = 2.17 T |
| cpc 1.80 unlocks 2 | 53 clicks, spend RM 95.40 = 3.60 T |
Show the full program
# Chapter 2. p23: two-stage Beta-Binomial red line (click -> unlock).
# A creative buys clicks at cost c each; a click unlocks the paid report
# with probability p, so CPA = c / p and "CPA > T" means p < c / T.
# Prior p ~ Beta(1, 24) (mean 0.04, illustrative site history). After n
# clicks and k unlocks the posterior is Beta(1 + k, 24 + n - k). For an
# integer first parameter the Beta CDF is a binomial tail:
# I_x(a, b) = P(Binomial(a + b - 1, x) >= a).
# T = 26.51 RM (p11). Three click prices show why clicks and spend must
# be read together: the same spend means different evidence.
import math
T = 26.51
A0, B0 = 1, 24
def beta_cdf(x, a, b):
n = a + b - 1
terms = [math.comb(n, j) * x ** j * (1 - x) ** (n - j)
for j in range(a, n + 1)]
return math.fsum(terms)
def p_bad(n, k, cpc):
return beta_cdf(cpc / T, A0 + k, B0 + n - k)
def kill_clicks(cpc, q, k=0):
n = k
while p_bad(n, k, cpc) < q and n < 5000:
n += 1
return n
if __name__ == "__main__":
print("A. P(CPA > T) after n clicks and zero unlocks")
print(" clicks cpc 0.45 cpc 0.90 cpc 1.80")
for n in (0, 10, 20, 40, 60, 100, 150):
row = [p_bad(n, 0, c) for c in (0.45, 0.90, 1.80)]
print(" {:>6d}".format(n) +
"".join("{:>10.4f}".format(x) for x in row))
print("B. clicks and spend before P(CPA > T) >= 0.90")
for c in (0.45, 0.90, 1.80):
for k in (0, 1, 2):
n = kill_clicks(c, 0.90, k)
print(" cpc {:.2f} unlocks {:d}: {:>4d} clicks, spend RM"
" {:>7.2f} = {:.2f} T".format(c, k, n, n * c,
n * c / T))Single source and citation
Fan, Y. (2026). Paid-Ad Scaling for Diagnostic Paywalls: Meta and TikTok Stopping Rules, Server-Side CAPI Calibration and Hundred-Order Daily Unit Economics (Version 1.0.0), Chapter 2: Stopping at Zero: the Rule of Three, Gamma-Poisson Posteriors and Spend Red Lines. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-345 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026345,
author = {Fan, Yafu},
title = {Paid-Ad Scaling for Diagnostic Paywalls: Meta and TikTok Stopping Rules, Server-Side CAPI Calibration and Hundred-Order Daily Unit Economics},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 1 · Funnel Unit Economics: Why First-Order Break-Even Is Not a Stop Line
- Chapter 3 · Signal, Noise and Sequential Tests: Small Budgets Can Only Call "Clearly Bad"
- Chapter 4 · Many Creatives on a Small Budget: Thompson Sampling, Even Splits and Screening Speed
- Chapter 5 · Calibrating Server-Side Events: Deduplication, Net-Value Signals and the Report Inflation Factor
- Chapter 6 · Control Laws for Scaling: Saturation, Carry-Over and Marginal Returns
- Chapter 7 · Classical Sources: the Sunzi on Waging War and Hopkins's Scientific Advertising
- Chapter 8 · Closing the Loop: a Three-Site Ad Ledger, the Distilled Tool and a Pre-Registered Refutation