What does Paid-Ad Scaling for Paywalls conclude, in numbers, about “Funnel Unit Economics: Why First-Order Break-Even Is Not a Stop Line”?
Answer: Of the forty creatives, 23 truly make money, and keeping exactly those earns 17,241 ringgit. Rule A kills 15 good creatives and keeps 2 bad ones, and earns only 3,822, or 0.222 of the attainable profit. Rule B does much better: it kills 7 good creatives but keeps 13 bad ones and earns 12,295, a share of 0.713. Rule C makes no mistakes and takes the full 1.000.
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: 0.222, 0.713, 1.000, 12295, 13, 15
| site-average funnel: u 0.085 refund 0.151 V 19.88 | ||||
| good creatives 23 of 40; best attainable profit 17,241 | ||||
| rule | kill good | keep bad | profit | share |
| A first ROAS>=1 | 15 | 2 | 3,822 | 0.222 |
| B site-average V | 7 | 13 | 12,295 | 0.713 |
| C own back-fill | 0 | 0 | 17,241 | 1.000 |
Show the full program
# Chapter 1. p13: a first-order ROAS stop rule misjudges creatives.
# 40 synthetic creatives for one assessment paywall (RM, illustrative).
# Each creative has a true CPA, upsell rate and refund rate. "Bait"
# creatives buy cheap icebreakers from people who rarely upsell and
# often refund; "fit" creatives cost more per buyer but bring readers
# who buy the deep report. Truth: keep a creative when V_i > CPA_i.
# Rule A keeps it when first-order ROAS = P1 / CPA >= 1. Rule B keeps it
# when CPA <= V computed with site-average funnel rates. Rule C uses the
# creative's own back-filled upsell and refund rates (week 4 onwards).
# Profit is counted on RM 1,000 of spend per creative.
import math
import random
P1, P2, CB, CBF, RATE, FIX, TAX, C1, C2, REP = \
9.90, 129.0, 0.005, 15.0, 0.034, 0.50, 0.06, 0.40, 1.50, 6.00
def order_net(price, refund, cost):
kept = price * (1 - refund - CB) * (1 - TAX)
return kept - (price * RATE + FIX) - CB * CBF - cost
def value(u, rf, rep):
return math.fsum([order_net(P1, rf, C1), u * order_net(P2, 0.04, C2),
rep])
rng = random.Random(20261120)
ads = []
for i in range(40):
if i % 2 == 0: # bait
cpa = rng.uniform(5.0, 14.0)
u = rng.uniform(0.00, 0.04)
rf = rng.uniform(0.15, 0.35)
rep = rng.uniform(0.5, 2.0)
else: # fit
cpa = rng.uniform(9.0, 40.0)
u = rng.uniform(0.10, 0.22)
rf = rng.uniform(0.02, 0.08)
rep = rng.uniform(4.0, 9.0)
ads.append((cpa, u, rf, rep))
avg_u = math.fsum(a[1] for a in ads) / len(ads)
avg_rf = math.fsum(a[2] for a in ads) / len(ads)
avg_rep = math.fsum(a[3] for a in ads) / len(ads)
v_avg = value(avg_u, avg_rf, avg_rep)
rules = {"A first ROAS>=1": lambda c, v: P1 / c >= 1.0,
"B site-average V": lambda c, v: c <= v_avg,
"C own back-fill": lambda c, v: c <= v}
print("site-average funnel: u {:.3f} refund {:.3f} V {:.2f}".format(
avg_u, avg_rf, v_avg))
truth_profit = []
for cpa, u, rf, rep in ads:
v = value(u, rf, rep)
truth_profit.append(1000.0 * (v / cpa - 1.0))
best = math.fsum(p for p in truth_profit if p > 0)
print("good creatives {:d} of 40; best attainable profit {:,.0f}".format(
len([p for p in truth_profit if p > 0]), best))
print(" rule kill good keep bad profit share")
for name in rules:
kg = kb = 0
prof = []
for (cpa, u, rf, rep), p in zip(ads, truth_profit):
keep = rules[name](cpa, value(u, rf, rep))
if keep:
prof.append(p)
if p > 0 and not keep:
kg += 1
if p <= 0 and keep:
kb += 1
tot = math.fsum(prof)
print(" {:<18} {:>9d} {:>9d} {:>8,.0f} {:>6.3f}".format(
name, kg, kb, tot, tot / best))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 1: Funnel Unit Economics: Why First-Order Break-Even Is Not a Stop Line. 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 2 · Stopping at Zero: the Rule of Three, Gamma-Poisson Posteriors and Spend Red Lines
- 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