BelongHub: Turning Community Research Into A Browsable Product
BelongHub is a public product from Urbano DX: a catalog of 141 verified community-building groups, apps, games, and traditions from around the world, organized around one pattern that repeats across centuries and continents. Competition between geographic areas, support within them.
In Catalonia the pattern is called a colla: the neighborhood teams that build human towers in town squares. The same shape appears in samba schools in Rio, dragon boat crews in Hong Kong, choirs in Wales, and makerspaces in Berlin. BelongHub collects those examples, verifies them, and makes them searchable.
You can open the live product here: communities.urbanodx.com.
What It Is
A catalog. 141 verified entries, each with a deep page and a related-entries panel, browsable as cards, by angle, or as a table.
A pattern library. 11 design patterns extracted from the catalog, each cross-linked to the entries that use it.
An explorer. A short "tell me about you" quiz that pre-fills filters, plus a full faceted browser for people who already know what they want.
A history. The long story of the pattern, from 18th century Catalan colles to crypto-era bounty programs.
A starting kit. The 7 open gaps in the landscape and 9 chapter models you could start on a weekend.
How The Dataset Is Built
Every entry is enriched on three orthogonal axes: activity domain, geographic unit, and mechanic family. On top of that, each entry carries a forkability score (could you copy this in a week, in a month, or does it need procurement) and a language family, so filtering behaves consistently on every page.
The site is data first. The catalog is generated from an upstream research workflow, and the product is a typed loader plus filter taxonomies on top of it. When the dataset grows, the site grows with it; no page needs to be rewritten by hand.
That is the same discipline we apply to client work: research becomes a structured dataset, the dataset becomes a product, and the product stays cheap to operate.
What It Demonstrates
Data curation as product. The value is not the code; it is the verified, consistently classified dataset and the editorial judgment in the 11 patterns. The interface makes that value browsable.
Lean engineering. A static React app (React 18, Vite, Tailwind) with prerendered pages for search engines, no backend, served by nginx. Fast to ship, almost free to run.
Editorial design. Illustrated rotating hero scenes, count-up stats, and a warm visual language. A research product does not have to look like a database dump.
If you have research, a spreadsheet, or an internal dataset that deserves to be a product, this is the shape of the work. A fixed-scope sprint from our packages can take it from raw data to a working, public web product, the same way BelongHub was built.