5D Library › Affiliate Network Power Laws, Influencer Portfolio Optimization and Incentive Stability Bounds · 中文

Preface: Sharing Out a Heavy Tail

Note

Please read this first. This book studies three questions in affiliate and creator marketing: why affiliates’ output is so concentrated; how a budget should be split between head creators and micro creators; and how a commission plan can give incentives without crossing the legal line of a pyramid scheme. Its numbers of affiliates, sales distributions, conversion rates, customer counts, commission rates and recruitment speeds are illustrative or synthetic models, meant to show structure; they are not data from any real programme or platform, and all data are generated by the programs printed in the book. Statutes are stated as found in public texts at the time of writing. The book is not legal, financial or investment advice; any commission plan must be reviewed by a locally qualified lawyer before it is put into effect.

Several claims about affiliate and creator marketing are popular: that in a team of ten thousand the best hundred sell half; that a first-sale bonus wakes the dormant; that putting 20% of the budget on head creators and 80% on micro creators cuts the variance of acquisition cost by 60%; that multi-level commission lets everyone be a boss. This book writes them as arithmetic that can be run, then asks when they hold.

What the book covers

Chapter 1 computes where Price’s law applies under heavy-tailed output, and the traps of dormant registrants and tail-index estimates. Chapter 2 computes the deadweight in first-sale activation bonuses and compares sales contests with pay by volume. Chapter 3 tests the “20% head, 80% micro” claim with mean-variance methods. Chapter 4 compares rules for learning while spending and sets out three ways of computing CAC. Chapter 5 computes the geometry of multi-level commission and its saturation in a finite market, and sets out the legal line in Malaysia and the United States. Chapter 6 follows the decay of an affiliate cohort, compares how three channels divide a sale, and gives a compliance monitoring measure. Chapter 7 reads Tao Zhu Gong in the Shiji’s “Biographies of the Money-Makers”, the labour shares of the Shanxi merchants and the rewards and punishments of the Wei Liaozi. Chapter 8 gives a year’s ledger, concentration monitoring and a pre-registered activation experiment.

Claims corrected after checking

  1. Price’s law is not a constant. At a tail index of 1.5, the share of the top N\sqrt{N} falls from 0.584 for a hundred people to 0.299 for ten thousand (Chapter 1, p01).
  2. Dormant registrants make concentration look higher. In the same network the top hundred hold 0.460 when the dormant are counted and the top fifty 0.356 when only the active are; Hill’s estimator makes the tail look too heavy when kk is too large (p02).
  3. Most of a first-sale bonus is deadweight. Of a 20-ringgit bonus, 0.867 goes to people who would have sold anyway, and large bonuses clearly lose money (Chapter 2, p03).
  4. Sales contests are no more efficient than pay by volume. With the same budget, the contest yields 2,570 and pay by volume 3,023 (p04).
  5. “Variance down 60%” holds only when creators are weakly correlated. The cut is 0.833 at a correlation of 0 and only 0.333 at 0.3 (Chapter 3, p05).
  6. CAC has three definitions. Average per campaign 723 ringgit, median per campaign 500, pooled 304; only pooled CAC matches profit and loss (Chapter 4, p08).
  7. Most people in a multi-level plan must lose. A full twelve-level tree with five recruits each needs 305,175,781 people; once a finite market saturates, 0.724 of members have a net loss (Chapter 5, p09, p10).
  8. Three sources. Tao Zhu Gong “three times made a thousand in gold, and twice gave it away”, making three fortunes and giving away two; the Wei Liaozi says “rewards are prized when they reach the small”, and the popular line “reward one man and ten thousand rejoice” is not in that chapter; the labour shares of the Shanxi merchants rest on Wei Juxian’s study, and the book quotes no original text it could not check (Chapter 7).

How to read the boxes

The boxes separate published evidence, runnable Python computations (16 programs, standard library only, fixed random seeds, each actually run at build time with its real output printed beneath it), classical sources, notes and the author’s hypotheses; every hypothesis states what result would show it wrong. Programs and their output use ASCII characters only.

Disclosure

The author uses the works of 5D Library (including this book) as the methodological basis of community courses, the 5D Ascension Cards (Dimensional Shift Card Decks) and the Yafu AI knowledge base, and runs several websites, paid online calculators, subscriptions, reports, courses, business licences and advisory services, as well as a creator marketing agency and affiliate programmes whose creator campaigns, commission design and licensed partnerships are exactly what this book studies; the author therefore has a direct commercial interest in its subject. This book promotes no software, course, card deck, affiliate programme or service, uses no transaction, creator or user data from the author’s businesses, and accepts no sponsorship from any platform or brand. All parameters in the book are synthetic or illustrative. The experiment in Chapter 8 must follow a pre-registered protocol, and its results will be published whether positive or negative.

Fan, Y. (2026). Affiliate Network Power Laws, Influencer Portfolio Optimization and Incentive Stability Bounds: Heavy-Tailed Productivity, Markowitz Traffic Allocation and Single-Tier Commission Equilibria [分销网络幂律激活、网红矩阵组合优化与激励稳定边界] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090229 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).