Markets don't make themselves.

Thousands of markets never form: willing buyers and sellers who cannot find each other, trust each other, or close a deal. We call them thin markets. Eighty years of economics explains why they stay that way, and the frictions that keep them thin can now be engineered away.

  1. 1 The problem
  2. 2 The solution
  3. 3 Examples
  4. 4 Cost
  5. 5 Documentation
🔓 Open source · MIT · contributors welcome
The owner of a small machine shop stands beside an idle milling machine, looking out at an empty yard.
AI-generated illustration
1 · The problem · Why Markets Don’t Form

Why don’t some markets form?

Buyers and sellers who cannot find, trust or deal with each other. A capable shop with an idle machine and a buyer who needs exactly that machine can sit an hour apart and never meet.

The frictions are known, documented over eighty years, and they repeat across thousands of markets.

A domain expert marks a printed page while a developer beside him works at a laptop.
AI-generated illustration
2 · The solution · What We’re Building

What are you building?

A free, open-source market engine, built once and dressed for each market: semantic matching, confidential disclosure, and a Deal Brief that hands off to the parties’ own professionals.

A market is built on it in one direction. First a wiki of what the market knows, written with someone who knows it. Then a twin, filled with synthetic participants, to watch the market behave. Then the marketplace itself.

A regional landscape with a harbour, a grain elevator, a factory, glasshouses and a timber yard.
AI-generated illustration
3 · Examples · Catalog

Show me the markets

Hundreds of prospective thin markets across 29 sectors, each tagged with the challenges that bind it and the interventions that would release it.

A shop supervisor studies a grid of cards on a laptop, with machines running behind a glass wall.
AI-generated illustration
3 · Examples · Demo

Can I see it work?

A sandbox simulation of the matching engine on a three-sided precision-machining market, populated and matchable.

Two business owners shake hands over a crate of machined parts at a loading dock while an adviser looks on.
AI-generated illustration
3 · Examples · Blog

What does a deal look like?

Narrative scenarios, one market at a time, showing what it looks like once the friction is engineered away.

A planner arranges a row of blank sticky notes into a timeline on an office wall.
AI-generated illustration
4 · Cost

What does a build cost?

The software is free. The cost is people’s time, and most of that is the judgment of someone who knows the market.

For a market of moderate complexity, a new wiki and a first demonstration twin come to roughly 215 to 355 hours of skilled work, and about US$100 to US$1,200 in AI fees and tools.

First estimates, from a small number of builds. The page says how small.

A reader works through a bound report at a table stacked with reports and binders.
AI-generated illustration
5 · Documentation · Library

What should I read, and where does it apply?

The whitepaper for the argument, the matrix for reference, the ebook for the Canadian case, and the worked examples: teardowns of real ventures, scenario stories and a diagnostic checklist.

The build guides are public too: the sponsor’s guide and the wiki-to-twin recipe, in the MarketForge repository.

A well-used desk with old notebooks, an engineering drawing, a slide rule and an open laptop.
AI-generated illustration
Who · About

Who is behind this?

A self-funded, pay-it-forward project by a founder with five decades of professional work behind it, with no investors and no board.

Five people of different ages work around a table in a room whose door stands open; one chair is empty.
AI-generated illustration
Take part · Contribute

Can I help?

Yes. The frictions that keep markets thin are common across thousands of markets, so the layer that fights them can be built once and reused. DeeperPoint is building that layer as free, MIT-licensed, open-source software: no licence fee, no usage tier, and no upsell behind the demo.

Code, knowledge of a market, a market to test on, and honest critique all have a place to land.