Flynt is now in your pocket

18 months ago, Flynt started as a web-only challenger to what high-quality networking should mean, and every step we took was in the direction of making that experience better for our community. That mission never changed. If anything, it feels stronger seeing how our members connect with each other through Friday matches, in-person meetups, and the People You Might Like feature for even more tailored connection suggestions. But as the usual tech startup tradition goes, we thought, why not build a mobile app? Jokes aside, we were unsatisfied by how disconnected the core product was from where our members actually stay up to date on the important stuff, their WhatsApp and email, and we wanted to centralise everything under one roof. Well, we thought about a macOS app too, but I'm getting ahead of myself. That's a story for another day.

Choosing Expo

We set off to propose a plan for how we should build our mobile apps, given none of us had previous iOS or Android development experience. I knew a little bit about Kotlin and did write a basic weather app in Swift, but I don't think any of that would have helped get Flynt off the ground. After some long discussions (which actually took a few hours, as that's how fast Flynt changes), we realised that Expo gives us everything we need. As we'd built the web app using TypeScript, building the same feature set in React Native was not a big leap for us. After all, we benefited heavily from AI programming (vibe-coding is under a soft ban), and setting up workflows to analyse how each feature was built on the web and replicate the same for mobile in React Native gave us 80% of what we needed pretty quickly. We were amazed by how quickly we started testing on a physical device. The remaining 20% is always the hardest as buttons not reacting to any taps, pop-ups strongly deciding not to close at all after we open them or us simply realising that we now need to build everything three times (or at least build twice but test on three different platforms).

Android app built & shipped in 2 weeks after iOS app was released

On iOS, our favourite bug was buttons that just would not react to taps for a good few seconds. The culprit? A one-second timer we forgot about that was quietly re-rendering the whole app, well, every second. The other one was worse as signing in could crash the app outright, because tearing down keyboard listeners made React Native poke at the JS engine from the wrong thread (memory corruption, the gift that crashes somewhere different every time). Add white flashes when you overscroll, Apple's pull-to-refresh spinner photobombing our custom one, and a privacy manifest that silently never got generated, and you get a decent picture of our first months on iOS. Once our first version was available on iOS, we immediately started thinking about Android which came with its own personality. Buttons again, except this time they were not slow, they were simply dead, so we ended up wrapping every Pressable in the app. Images blanked out the moment anything about them changed, and the keyboard either covered the input or left a mysterious gap where the tab bar used to be. And just when the app behaved, Google Maps rendered a lovely blank grey rectangle until we gave up and switched to static maps. However, the whole journey from starting development to going live on Android took only two weeks, as we massively benefited from the same infrastructure that powered our iOS app and used the same backend the web already runs on.

Mobile releases are automatically cut every week

After the first versions went live on iOS and Android, we still relied on someone remembering to release new mobile versions each week to keep up with the new feature set, and as a result we sometimes found issues on mobile days after we had fixed them on web. We have now automated it so that a fresh release is built each Sunday from the latest version of the repo and shipped automatically without human input. We still need to remember to build those features in our mobile repo in the first place, which is ripe for automation on its own. We'll tackle that very soon.

More members now have the Flynt app than have WhatsApp connected

We let Flynt members connect their WhatsApp account so they never miss the important moments, like when their match is ready to reveal. It works, but between WhatsApp nudges and the occasional important email, staying in the loop meant juggling apps that are not Flynt, and that never felt very delightful to us. We strongly believe our mobile apps will bridge that gap, so our members burn far fewer calories finding out what is happening across Flynt. The uptake suggests our members agree too. In under three months, more of Flynt community have the Flynt app than have WhatsApp connected.

Moments of delight

We occasionally award exclusive badges to our Flynt members, such as when they come to our in-person meetups or take a challenge (p.s. "Working in your natural way" is my favourite so far). Although we had this idea for a long time, having our app felt like the right opportunity to render 3D badges and add a bit of delight. This is not the only feature where delight mattered as much as functionality, so consider it a "delight hunt" next time you open the Flynt app.

More coming soon..

We are very excited about our mobile apps, and usage has already shifted dramatically away from web. We are also aware we have not fully maximised mobile delightfulness yet, as our first sprint was about getting things functional. That said, we could not stop ourselves from adding a bit of magic where we could, making sure our members get the best experience from their very first install. Our next sprint is about adding that delight where it matters, such as simplifying new registrations and logins, tapping into device-specific advantages to make match days feel more "wow", and making sure our members do not need to juggle various tools to find out what is happening on Flynt, get ready for their upcoming match and tailor their profile to what works best for them. If you have any ideas, please chat to us.