We launched as an app. Almost nobody installed it.
We started with an app because it seemed closer to what people already used. Hardly anyone would download it, so we moved to the web.
A market that lived on Facebook
In 2020, most people in Cambodia were on Facebook but hardly used the open web. Plenty had never typed a URL into a browser. Buying a car meant going from one showroom to the next, which was slow, but it was how everyone did it.
Meet users where they are
So we launched a used-car marketplace app, built in React Native for iOS and Android. An app felt closer to what people already used than a website. The business model, the app design and the admin side took about a year.
Curious, but not enough to install
Whenever I showed the app to buyers and dealers, the reaction was the same. They were interested, but not enough to download it. Sometimes I installed it on their phone myself. Nobody was searching for Carmate, so nobody found the app on their own. And with two of us, we couldn't run an app and a website at the same time.
Each extra step loses people who are only curious, and an install is a big step. A link doesn't need installing. It can be shared on Facebook and Telegram, and people can find it on Google.
Three ways forward
One product, on the web
We moved Carmate to the web and built it to be found on Google and shared as a link. That became today's marketplace for new and used cars, with Seller Central for official brands.
Ready when the market moved
People here are moving online, slowly. More of them now look a car up before they visit a showroom, and the website is where they find us.
What I'd do differently
Before building the app, I'd check whether people would install one at all. A simple web page with a download link, shared on Facebook, would have shown us quickly, for far less than a year of building.