Concepts

The automation band

The automation band is the range, 30 to 70%, of a vertical's work that AI can automate in which buying companies beats building one. Below it, AI changes too little to pay for the acquisition premium; above it, you would pay for people you are about to replace, so you build instead. Read more →

• Below 30%: AI changes too little to pay for the acquisition premium.
• 30–70%: the band. The rest is real operations and real customer relationships, worth buying.
• Above 70%: build instead. You would pay for people you are about to replace.

The data exception: above 70%, buying can still make sense when the companies' client relationships produce data you cannot get any other way. Then you are buying the panel, not the work.

The prediction economy

The prediction economy is the phase in which the winners in vertical markets are no longer the companies that automate the work, but the ones that see the market through the work. It is the third level of value creation in software: after selling the tool and selling the outcome, companies sell what their distribution lets them see. Read more →

First set out in The Prediction Economy, 2026

The panel

The panel is a network of service providers, such as tax advisors, insurance brokers or IT providers, whose daily work produces private data about their clients before it becomes public. Read together and calibrated over years, it shows where a market is moving months before any statistic does. Read more →

Telemetry moat

A telemetry moat is the advantage a company builds from the behavioral data it collects as a by-product of serving its customers. A model can be copied and a feature rebuilt in a weekend, but that data cannot be bought from public sources, and it grows with every customer. Read more →

Distribution as an asset class

Distribution as an asset class is the thesis that a customer base, and what its behavior reveals, can be valued, bought and sold on its own, separate from the product or company that built it. As software becomes cheap to build, access to customers becomes the scarce asset, and buyers price it directly. Read more →

First set out in Dis·tribution

Distribution arbitrage

Distribution arbitrage is the gap between what it costs to win a customer segment yourself and what an acquirer pays to get it by buying a company that already serves it. When buying a competitor's customers becomes cheaper than winning them, M&A turns into a game of customer-acquisition cost. Read more →

First set out in How AI Companies Will Get Acquired, 2026

Revenue reclassification

Revenue reclassification is the repricing of a company's revenue because AI changed how likely it is to stay, not how large it is. Buyers still pay for sticky revenue; AI changes what counts as sticky, so, as in a credit downgrade, the revenue stays the same, the label changes and the price follows the label. Read more →