Aggregation Theory: How Free Distribution Rewrote the Rules of Monopoly

Consider two businesses that, on paper, do the same job.

A newspaper in 1985 connected readers with news, and advertisers with readers. To do it, the paper needed printing presses worth tens of crores, fleets of delivery trucks, and a distribution network built over decades. Those assets were the moat. A competitor could hire better journalists tomorrow; they could not replicate the distribution machine for years, if ever.

Now consider the news aggregator on your phone this morning. It connects the same readers to the same journalism. It owns no presses, no trucks, no newsprint contracts. By every rule of twentieth-century strategy it should be defenceless — an intermediary with nothing proprietary underneath it. Instead, it captures a larger share of news-industry profit than most publishers earn, and publishers reorganise their entire operations around its ranking decisions.

Something structural changed between those two worlds. Aggregation Theory — articulated by the analyst Ben Thompson — is the clearest account of what.

The old physics: distribution as the choke point

Every value chain has three parts: suppliers who make things, distributors who move them, and consumers who buy them. Profit does not spread evenly across a chain. It pools at the choke point — the stage where capacity is scarcest and hardest to replicate.

For most of the twentieth century, that stage was distribution. Newsprint was expensive to print and truck; broadcast spectrum was licensed and finite; retail shelf space was physically limited. Whoever controlled a distribution channel could extract rent from both sides: suppliers had to pay (or accept poor terms) to reach consumers, and consumers had little choice but to accept whatever the channel carried.

This is why the giants of that era look the way they do. They were, underneath the branding, distribution machines: broadcast networks, newspaper chains, retail conglomerates. Their supplier relationships were adversarial and their consumer relationships were captive, and both facts flowed from the same source — ownership of the scarce middle.

The inversion

The internet did something no previous technology had done: it made the marginal cost of distribution effectively zero. Copying and delivering a digital good to one more person costs nothing measurable. Even for physical goods, the internet collapsed the discovery and transaction layer — the part of distribution where the old chokehold actually lived.

Follow the consequences in order.

First, supply explodes. When anyone can publish, list, or sell without paying a distribution gatekeeper, they do. The number of news sources, sellers, drivers, and hosts grows by orders of magnitude.

Second, abundance creates a new scarcity. A consumer facing effectively infinite supply cannot evaluate it. The binding constraint is no longer access to goods — it is the consumer's finite attention and their unwillingness to make a hundred small decisions. Whoever solves that problem best — the cleanest interface, the best defaults, the least friction — becomes the place consumers simply start.

Third, and this is the step that produces monopoly-like outcomes: once consumers start in one place by habit, suppliers have no choice but to be there. And every supplier who joins makes the aggregator's catalogue more complete, which makes it a better starting point for consumers, which makes it more unavoidable for suppliers. The loop feeds itself. This is a demand-side network effect, and unlike a factory or a truck fleet, it strengthens with scale rather than merely growing.

What the aggregator deliberately refuses to own

Here is the counterintuitive core of the theory. The aggregator's power comes from what it doesn't own.

The old distributor owned physical capacity, which meant capital intensity, which meant limits to scale. The aggregator owns only the user relationship — an interface, a habit, a default — and sources supply from an open ecosystem it never has to finance. A hotel chain must build hotels to grow; a lodging aggregator adds a million rooms by changing nothing but a database. A taxi company buys cars; a ride-hailing aggregator's fleet is other people's cars.

This has a brutal consequence for suppliers: they are commoditised — not by malice, but by structure. When the consumer's loyalty attaches to the aggregator's interface rather than to any individual supplier, suppliers become interchangeable line items in someone else's catalogue, competing against each other on price and ranking. The commission the aggregator charges is not really a fee for services. It is rent on access to aggregated demand — the modern descendant of what the newspaper charged advertisers, with one difference: the newspaper's chokehold was limited by geography and print capacity. The aggregator's is not.

Second-order implications

Competition changes shape. You cannot out-compete an aggregator by building a marginally better version of its suppliers' product — the suppliers already did that, and it made the aggregator stronger. Successful challenges historically come from a different direction: a new interface on a new platform shift (desktop → mobile → whatever is next), or a sub-segment of demand served so differently that users change their starting point. The competitive unit is the habit, not the product.

Regulation aims at the wrong layer. Antitrust doctrine built for the distribution era looks for the old signatures of monopoly — ownership of scarce infrastructure, prices rising above competitive levels. Aggregators own no scarce infrastructure and typically charge consumers nothing. The power sits in the demand relationship, which existing tools were never designed to measure — which is why decade-long cases against aggregators so often end in stalemates.

The theory has limits, and they are instructive. Aggregation works where supply is abundant, digitisable or near-commodity, and where transaction costs were the main barrier. It works poorly where supply is genuinely scarce and differentiated — which is why the model that flattened news and taxis has struggled to flatten, say, high-end manufacturing or specialist medicine. Whenever you want to predict whether an industry can be aggregated, ask one question: if distribution were free, would supply be interchangeable? If yes, an aggregator is coming. If no, the choke point — and the profit — stays where it is.

That single question, applied industry by industry, explains more of the last twenty years of business history than any other framework of its size. See also the Theory of Constraints — aggregation is what happens when the binding constraint of an entire economy moves from supply to demand — and the Bullwhip Effect, for what happens to chains where information, rather than attention, is the distorted resource.