Disruption of the Internet Marketplace

I tried to build an internet marketplace. The vision was to aggregate a fragmented, overlooked, and underserved market into one platform – and ultimately enable more demand. I thought the business model was beautiful. As we approached launch, we realized that the business model won't be around much longer. This is my attempt to articulate my thesis.

Money out of nowhere

In 2019, Bill Gurley wrote an essay called "Money Out of Nowhere: How Internet Marketplaces Unlock Economic Wealth." The argument was that internet marketplaces didn't just move money around – they created it. By connecting buyers and sellers who couldn't find each other, they unlocked economic value that didn't exist before. 

Some of the biggest winners from the last two decades were internet marketplaces. Many of them had a laughably small TAM when they were starting out (e.g. Airbnb, Etsy, Uber). But as they grew, their TAM grew. Today, some of the most capital-efficient and wonderful businesses are internet marketplaces. 

However, in the coming years, many of these companies will face an existential threat. The ‘take rate’ will soon flow back to the seller, as we see a new paradigm of connecting buyers and sellers that don’t require a marketplace.

Not all network effects are created equal

The most defensible characteristic of a marketplace is its network effects. 

‘Heavy’ marketplaces have operational network effects. These marketplaces build operational infrastructure – where stronger network effects means competitive advantage primarily at the unit economics level. For example, as Uber grows, the rides are easier to find, but they are also quicker and cheaper. These companies are better understood as operational platforms with marketplace functionality. Their moat is built around the physical assets they own, and their ability to deliver superior unit economics.

‘Light’ marketplaces have informational network effects. These marketplaces build informational legibility – where stronger network effects means competitive advantage primarily at the customer experience level. For example, as Upwork grows, freelancers are easier to find, but the hourly rate or the quality of the service itself does not change. These companies are pure play marketplaces, and cannot be operational platforms – because they cannot own their supply. As aggregating supply becomes increasingly easier, informational scarcity is no longer a defensible moat.

The business model as structurally undermined

The core value proposition for light marketplaces is discovery + booking. We are heading towards a world where discovery is commoditized and bookings are individualized.

As more efficient solutions to the core value proposition emerge, we will see the light marketplace business model be structurally undermined.

1. Discovery: Before LLMs, supply was fragmented and/or non-existent. This was the opportunity for companies like Airbnb, as they created a category and unique supply base. But supply will follow wherever visibility is highest. We see this with Booking.com being the second-mover in alternative accommodations, but already with the same number of listings as Airbnb. The overlap in their supply base is likely significant. In this context, discovery is already being commoditized. But LLMs change the equation completely – as they can gather information across marketplaces, individual websites, and social media. This doesn't eliminate centralized aggregators, but it does erode their core advantage — being the only place where supply is legible.

2. Booking: Handling booking is essentially sellers letting the marketplace be a part of payment flow because they help with discovery, lower purchase friction, handle compliance, and guarantee trust. But supply will follow wherever revenue opportunity is highest. When discovery becomes commoditized, we will see each part of booking become a core service. Scheduling, payments, compliance, trust… There will be tools that solve this for far cheaper than the marketplace take rate. Each of these will be built for the individual businesses. The booking portion of the value proposition doesn’t lose value; it just becomes bad value.

We are heading towards a world where there will be more efficient ways to connect buyers and sellers. Even if marketplaces integrate further or "adopt AI", the economics don't work. The business model breaks from multiple angles simultaneously. (1) Sellers go wherever visibility and revenue opportunity is best. (2) If their take rate gets squeezed to retain supply, their marketing budget gets squeezed, which leads to lower booking volume.

Eventually, and increasingly, sellers will ask: Why would I pay 15% to the marketplace if I can solve discovery and booking for far cheaper?

The new paradigm

We will see an unbundling of the ‘marketplace stack’. Marketplaces are able to command a strong take rate because they bundle the industry end-to-end, from discovery to payment to relationship. 

As discovery becomes commoditized, the other dominoes will fall, and there will be companies that see the opportunity to provide services directly to sellers – without the marketplace lock-in and cost. 

The atomic unit for sellers is currently a platform listing (e.g. Vrbo listing, Thumbtack profile, etc). Soon, it will be an agentic business profile and toolkit that doesn't belong to any one platform.

A simple way to visualize it is: Imagine if ChatGPT created a new Business Profile. Sellers create an account, add their business information, explain their business needs, connect their payments, and connect their calendar. Sellers get visibility and an agent that optimizes for their business goals; ChatGPT gets supply to enable agentic commerce. The pricing power of the marketplace depends entirely on owning discovery and booking. Once those are unbundled, the take rate has no justification. That is where the business model is undermined. 

I write more about the new value chain of marketplaces here.

Money into the hands of sellers

Ultimately, light marketplaces go from efficiency providers to unnecessary middlemen — as sellers find better visibility and revenue opportunity elsewhere. The change will be gradual. At the highest level, the money out of nowhere becomes money into sellers' hands — at higher margins, through infrastructure and relationships they control.