A marketplace finished weeks before the world closed

Tourism is one of the pillars of the Mauritian economy, and in 2018 the operators who actually deliver it, the boat crews and guides and drivers, were easy enough to find and hard to book. Most had a phone number and a Facebook page, a few had a website. What none of them had was live availability, payment on the spot, or a single place where a traveller could compare them and book in one sitting. Discovery worked. The transaction did not. Island Experience was my answer to that: one place where a traveller could find, book and pay a Mauritian operator directly.

Mine, set by me, funded by me. Build the entire platform, sign the operators, and launch it.

I was the only shareholder and the only engineer. The specifications, the architecture, the interfaces for travellers, operators and administrators, the booking engine, payments, in-app chat between merchants and customers, search, the reporting and the operational dashboards, the containerised development environment, the pipelines, the deployment, the documentation. Eighteen months of it, paid for out of my own account.

By early 2020 it was finished. Not a prototype and not a demo: the working platform, with signed agreements from local operators waiting to be switched on.

Then the borders closed. Tourism in Mauritius did not decline, it stopped, and it stayed stopped for two years. The platform never took a booking. I wound the company down and paid for that decision myself.

BUILT — THREE INTERFACES · BOOKING ENGINE · PAYMENTS · MERCHANT AND CUSTOMER CHAT · SEARCH · OPERATIONAL AND SALES REPORTING · CI/CD · PRODUCTION DEPLOYMENT · SIGNED LOCAL OPERATORS

The engineering was never the thing that killed it, and that is exactly the lesson. I had built the whole system and I had not built enough room for the world to be wrong.

So this is the case study that changed how I read other people’s platforms. When I assess one now, I am not only asking whether the architecture is sound. I am asking what it costs to keep alive for a year with no revenue, what happens to the plan if the market moves underneath it, and which decisions become irreversible before the first customer pays. I ask because I have paid that bill with my own money, which is a different kind of knowing than reading it in a book.

Investors and owners do not need an advisor who has only shipped things that worked. They need one who knows precisely how a good system dies.