I have been part of three launches: the Airtel TV streaming app for the African market and the Reach Mobile SIM service in the US, both at Datami, and the Nexter Finance prediction market in decentralised finance. Three industries, three very different teams. A few things held true every time.
Ship the smallest true thing
At zero, the temptation is to design the whole vision so that nothing looks missing on launch day. The result is usually a product that launches late, or never.
At Nexter the scope was broad, predictions across crypto, stocks, commodities and real-world events, and it was still moving. Instead of waiting for it to settle, I proposed launching one mode first, Lossy Predictions, and releasing the rest in phases. The team shipped a real product while the rest of the scope was still being decided.
The first version doesn't need to be complete. It needs to be true to the idea, and real enough to learn from.
Design for a scope that will move
On a 0→1 product the requirements are guesses, however confident they sound. Designing as if they were fixed produces work that has to be thrown away the moment the first real users arrive.
What survives is structure: a small set of patterns that can absorb new features without a redesign, and flows built around the few decisions the product can't change. The rest you design to be replaced.
Put it in front of a few real people early
You don't need a large study to find out whether a first product makes sense. At Nexter I ran usability testing with six users. Six people will not tell you what the market thinks, but they will show you, very quickly, where the product doesn't make sense to someone who isn't in the room when it was designed.
For a product in a new category, that is most of the risk.
Design for people who aren't like you
All three were for people whose context wasn't the team's. A streaming app for a market I didn't live in. A US SIM service for people used to buying one in a store. A DeFi product that had to serve people who already knew web3 without shutting out everyone else.
The habit is the same each time: find out what the person already knows and fears, and start from there, not from what the team finds obvious.
Let the launch be the beginning
The day a 0→1 product ships is the day you finally get real evidence. The teams that do well treat launch as the start of the design process, not the end of it: watch what people actually do, keep what works, and change the rest quickly.
At Reach Mobile, launch is where we learned the most. Customer feedback and analytics showed people dropping out of the purchase flow, while a short flow we had built for social and ad campaigns was the one people finished, so it became the only one. Soon after, the audience shifted from people in their twenties to people over 35, and the site was redesigned for them.
Three launches in, that is the part I would tell anyone starting their first. Getting to launch is hard. What you do in the weeks after it decides whether the product lasts.