Technology Evolution & Disruption
Modern Monopolies: What It Takes to Dominate the 21st Century Economy
Moazed and Johnson's argument: the companies that dominate the internet economy do not make or stock what they sell — they run the connection between producers and consumers. For investors, whether a business is a pipe or a platform changes how its value scales with users, and the difference shows up in a handful of marketplace metrics.
The big picture
Moazed, founder of the platform consultancy Applico, and Johnson split businesses into two models. A linear business (often called a pipe) buys or makes something and sells it down a chain to customers; value is added step by step and each new customer adds roughly the same amount. A platform owns no inventory of its own. It builds an audience on two sides — producers and consumers — and makes money from the exchanges between them. Its value depends on how many users sit on each side and how reliably they find each other, so growth on one side makes the platform more useful to the other. The authors' bet is that this model, not better products, explains why a small number of internet companies ended up with dominant market shares, and that the same logic keeps spreading into industries that are still run as pipes. The book was published by St. Martin's Press in 2016.
Why it matters now: in 2026 the fight in AI is partly a fight over who becomes the platform — model marketplaces, agent app stores, cloud consoles that sit between model builders and the companies deploying them, and network services that charge AI crawlers for access to publishers' content. The book gives a vocabulary for asking which of those businesses actually connect two sides and which are pipes with a platform label.
The 3 strategic pillars
-
Own the connection, not the production
A platform creates value by making exchanges between outside producers and consumers possible, not by producing goods itself. The unit to study is the core transaction — the single repeatable exchange the platform exists to enable, such as one ride, one booking, one sale or one app install.
Because producers bring the supply, the platform's marginal cost of adding inventory is close to zero and its revenue is a cut of the exchange. In the numbers: gross merchandise value (GMV, the total value transacted on the platform) grows much faster than the platform's own cost base, and revenue is GMV times the take rate (the share of each transaction the platform keeps).
-
Four jobs make the core transaction work
The authors describe the platform's work as four functions: building an audience on both sides, matching producers with the right consumers, providing the core tools that make the exchange cheap to do, and setting the rules that keep quality up and bad actors out.
Each function leaves a trace in reported metrics. Audience shows as active producers and consumers; matching shows as the match rate (requests that end in a transaction) and time to match; tools show as the share of transactions completed without leaving the platform; rules show as repeat use and cohort retention (the share of a joining group still active months later). When matching is weak, users multi-home — they use a rival platform at the same time — and the platform loses pricing power.
-
Network effects, and the chicken-and-egg start
A platform becomes more valuable to each user as more users join, especially on the opposite side (a cross-side network effect: more drivers make the app better for riders and the reverse). The hard part is the start, when neither side has a reason to join without the other.
Platforms solve the start by seeding one side first — subsidising it, producing some of the supply themselves, or winning one small market (a single city, campus or niche) completely before expanding. Once both sides are dense enough, growth feeds itself and the leader's advantage compounds, which is why the authors expect platform markets to tip toward one or two winners and to draw regulators' attention.
What a Closelooknet reader does with it
The working use is a two-step check. First, classify the business: does it own the goods or services it sells, or does it take a cut of exchanges between outside parties? The answer sets which growth model is reasonable — per-customer (linear) value for a pipe, value that rises faster than users for a platform. Second, check whether the network effect is real in the reported data rather than in the pitch: a rising or stable take rate without losing producers, a match rate that improves as the platform grows, falling multi-homing and improving cohort retention. The mistake it prevents is paying a platform multiple for a pipe, or assuming that a large user count alone produces network effects when the users do not transact with each other. The value laws in the pack are outside stand-ins — the book argues the point in words and gives no formula for what a network is worth.
The bridge to the Closelooknet approach
The book is the business-strategy version of ideas that Closelooknet covers from the economics side. Shapiro and Varian's Information Rules formalised positive feedback, lock-in and standards wars two decades earlier; read together, the two books separate a network effect (value to users rises with users) from ordinary scale (cost per unit falls with volume). Moore's Crossing the Chasm describes the same start-up problem for a single product — a platform has to win a beachhead on both sides at once, which is why many start in one city or one niche. On the site, Cloudflare as Agentic Toll Booth is a live case of a network service turning into a two-sided platform between publishers and AI crawlers, and SaaSpocalypse asks whether AI agents turn seat-based software into commodity supply on someone else's platform. The glossary entries for moat, cohort analysis, churn and net revenue retention define the pack's health inputs. The pack runs on the reader's own data; its output is a research note, not a signal.
Action-Kit — from theory to practice
Tooling & data
| What you need | Where to get it | Cost |
|---|---|---|
| Company filings and shareholder letters GMV or gross bookings, net revenue, active buyers and sellers, and any disclosed retention or frequency figures | SEC EDGAR (US) and company investor-relations pages Marketplaces define GMV, bookings and 'active' users differently — record the definition next to every number and keep it constant across quarters. | Free |
| Web and app traffic estimates Audience overlap between rival platforms as a rough proxy for multi-homing, plus visit trends on each side | Similarweb Panel-based estimates; the free tier shows top-line traffic, audience overlap needs the paid tier. Treat as direction, not a measured share. | Freemium |
| Standardised multi-year financials Revenue and cost history to check whether revenue grows faster than the cost base as GMV scales | stockanalysis.com | Freemium |
The formulas
-
Value multiplier under four stand-in laws
Linear ×m · Metcalfe ×m² · Odlyzko ×m·ln(m·n0)/ln(n0) · all calibrated to 1× at today's users- n0: active users today (both sides)
- m: growth multiple of users (2 = users double)
Metcalfe's n² and the Briscoe–Odlyzko–Tilly n·ln n are outside models, not claims of the book. n² is usually treated as an upper bound, n·ln n as a conservative case; neither models saturation or congestion.
-
Two-sided cross-side multiplier
V ∝ P·C + s·(P² + C²)/2; multiplier = (mP·mC·p(1−p) + s/2·(mP²p² + mC²(1−p)²)) / (p(1−p) + s/2·(p² + (1−p)²))- p: producer share of users today
- mP, mC: growth multiples of producers and consumers (pack default mP = mC^e, e = 0.7)
- s: same-side weight (0 = users value only the other side)
A Closelooknet illustration of cross-side effects. It shows the chicken-and-egg problem in numbers: growing one side alone adds far less than growing both.
-
Take rate and match rate
Take rate = net revenue / GMV; Match rate = matched transactions / requests- GMV: total value transacted on the platform
- Net revenue: what the platform keeps
- Requests: searches, ride requests or orders placed (one definition, kept constant)
- Matched transactions: requests that ended in an exchange
A take rate that rises while producers leave or multi-home more is extraction, not strength. The pack flags both ends of a reader-set range.
-
Cohort retention
Retention(k) = active users of cohort c in month k / users who joined in cohort c- Cohort: users who joined in the same month or quarter
- k: months since joining (the pack uses month 12)
Improving retention for newer cohorts is one of the clearest traces of a working network effect — later joiners find a denser network than early ones did.
Applied Pack · free members
Moazed Applied Pack
The network-effect multiplier model: compare how value scales with users under linear, n², n·ln n and a two-sided cross-side law, then score your own marketplace KPIs — take rate, match rate, multi-homing, cohort retention — against thresholds you set.
- Moazed_Network_Effect_Model.xlsx — READ ME; Value Curves (amber inputs: users today, value today, producer share, producer growth elasticity, same-side weight, growth multiples → live multipliers and values under four laws); Two-Sided Grid (producer × consumer growth → cross-side multiplier); Platform Health (EXAMPLE_ quarterly KPIs you overwrite → take rate, match rate, consumers per producer, GMV per consumer, QoQ growth and four threshold checks)
- moazed_network.py — stdlib-only: reads a CSV of your own platform KPIs and prints the health table, or with --curves prints the value-multiplier table for your user count
- platform_kpis_sample.csv — EXAMPLE_ rows showing the input format
- README.txt — which model is whose, input definitions, formulas, how to run, and the educational-use disclaimer
Pack security
Macro-free Excel · plain-text Python you can read before you run it · no installers, no network access — the code works only on files you provide. Served only from closelook.net; we never distribute through download portals or email attachments. How to verify in 30 seconds →
SHA-256 bc5aa4f58aa560a8e502757b22d1b30852b0bf3f7c619426e1a7e43e01207725
The pack is a free-membership download — no card, free forever. Membership also delivers the Daily Pulse and Weekly Signal to your inbox.
Join the Look — free → Already a member? Sign in →
Educational templates — a research diary companion, not investment advice.
Closelooknet publishes a market diary, not investment advice. This condensed read restates the book's ideas in our own words for education — for the author's full argument, go to the source.