The neobank map is written from London. The chapters aren't.
Here is the uncomfortable shape of this industry: 817 million of the people who use the neobanks we track are in Asia. 354 million are in Latin America. 105 million are in Africa. Every neobank user in the United States, Canada, the United Kingdom, Europe and Australia combined comes to about 112 million — roughly a fourteenth of the total.
Now count the coverage. Almost all of it is written from London, New York or San Francisco, about the seventh of the market that happens to sit nearby. We are not innocent here: this site verifies licences and custody models for 381 neobanks from a laptop in Europe, and for the other six-sevenths of the world that means checking registers, filings and app stores from a great distance. It is honest work, and it is not the same as being there.
So we are opening the map. neobankbeat now has regional chapters — seven of them, one per region, each with its own page, its own running radar, and a seat for someone who knows that market better than the fintech press does.
What a chapter actually is
Two jobs, deliberately paired.
A radar. A running log of the developments that matter in that region — licences granted and revoked, rails launched, funding that changes the map, the shutdown nobody in London noticed. Entries arrive as pull requests to a JSON file in the open repo, get reviewed like any other change, and render on the chapter page with the contributor's byline. Every entry carries a source link. The build literally fails on an entry without one, which is a nicer way of enforcing a standard than an argument.
Data stewardship. The chapter holder owns their region's rows in data.json: auditing what is there, correcting what is wrong, and reviewing regional additions before they merge. This is the part that compounds. A wrong custody model or a dead licence sitting in the dataset is worse than a missing one, and the person most likely to catch it lives in the market.
The seven
| Chapter | Tracked | Reported users | Status |
|---|---|---|---|
| Asia | 68 | 817M | seat open |
| Europe | 79 | 31M | seat open |
| North America | 61 | 81M | seat open |
| Latin America | 49 | 354M | seat open |
| Africa | 29 | 105M | steward in trial |
| MENA | 22 | — | seat open |
| Oceania | 9 | — | seat open |
Each page carries the same furniture, computed from the dataset at build time: a dotted map of the region with every tracked headquarters lit up, the category mix, a country breakdown, and the full roster ordered by scale. When the dataset changes, the pages change. Nobody has to remember to update a number.
How you take a seat
The seat is earned rather than assigned, and the bar is deliberately concrete: four sourced radar entries and one full audit pass of the region's rows. Do that and your name and link go on the chapter page as its steward. It is roughly a few evenings of work by someone who already follows their market — and it is the same bar we hold ourselves to, which is the point.
There is no payment and no exclusivity, because there is also no advertising, no sponsorship of listings and nothing to sell you. What you get is a byline on an open dataset that AI assistants, journalists and analysts cite, and editorial control over how your region is represented in it.
If that appeals: open an issue and make the case. The contributing guide has the format.