Roads Audio

The biggest area of confusion with the app was what to do once it was downloaded. This was a challenge for new users especially, the app expected them to start recording audio, but people weren't.

After asking people why, they seemed to avoid recording anything because they didn't know what to record. Most wanted to see how the app worked and expected a social feed of content to listen to before contributing their own recordings.

In response to this I redesigned the onboarding with three guided audio messages which both show people how the app's conversation flow works, and describes the core features of the app.

The key differentiator of Roads is the ability to comment on audio with audio, however in the initial version you could only comment one level deep.

This was a major rewrite to the codebase because initially the app had two types of audio. Episodes (the main conversation) and comments.

With this update all audio is now considered a Post, and there are several Post types (Root Post, Comment Post, Response Post, etc.). The new setup is much more flexible and will allow for future post types, but more importantly this setup allows for conversations to continue naturally within the context of the audio you listen to.

This was a highly requested feature from several of the early users, and was developed in response to that feedback.

I posted a link to Roads on Show Hacker News which received 72 points and 56 comments! This lead to about 100 new accounts created, and was the first big push of traction.

Seeing people use the app was a great feeling, but the feedback in the comments was huge. One feature that was requested by several people was the ability to have nested comments. Which was later released in December.

https://news.ycombinator.com/item?id=37758099

Since the app was built with Flutter I was able to launch on both iOS and Android, and do to the ~5 months of beta testing there were no significant bugs/ errors.

Launching the landing page with the waitlist did collect some emails, about 50, which was better than none but ultimately the launch on the app stores was pretty quiet.

As a start to my marketing efforts, I launched a landing page. The page explains the main features of the app and allows people to sign up to be notified when it's live.

My goal is to get a few hundred emails within the next month or two which will make the actual launch so much more successful. The landing page can be viewed at https://roadsaudio.com

After several iterations of the design and adding all the core features I have launched a beta version of the app to friends and family.

This is a huge step and now I can get some feedback on what I have built so far. Honestly, I should have launched the beta much earlier and had people testing out features along the way but now is better than later.

In the next two months, I plan to fix major bugs and start building the go to market plan. Then my goal is to launch the app to the public some time in June.

I've been building the app with Flutter and Firebase for the past few months. Today I made the decision to move off Firebase and build a custom backend with Django Rest Framework.

This wasn't an easy decision, Firebase would definitely allow me to launch a first version quicker, but I am already seeing limitations with it. Mainly, I have complex relationships that can be modeled much simpler with a relational database like Postgres. I'm noticing to make this with with Firestore's NoSQL setup I have to copy data and keep it in-sync or read tons of extra documents which will not scale.

I have some experience with Python and Django but definitely will be learning a lot. It's exciting to me, at this point I'm digging deeper into the tech and "building this for the long term" which I realize could be a mistake because it will delay the launch.

Finally I started actually building Roads. I've had the idea for this project for several years and have always been to busy to actually start. I also know how much work this will actually take so to start this is committing to at least a year of full time work.

I'm ready to commit fully and see where this goes. I think when you have an idea that sticks in your mind for awhile and you keep coming back to it, you have to at least put some time aside to give it a try.

Back in 2017 while I was traveling in South East Asia, me and a few friends who were all in different parts of the world wanted to keep in touch. However, the different timezones and various busy schedules made this difficult.

So we started a "podcast", at least that's what we called it. Really what we did was record 10-15 minute long audio recordings and upload them to a shared google drive. It was a whole new way to keep in touch asynchronously and it worked for the most part.

One major thing that kept happening in our "podcast" was one person would talk about something, then when another person responded they would summarize what was talked about and add their comment. There was a lot of repeating of context and you also needed to remember what people said so you can respond to it later.

This is the major problem Roads is setting out to solve. Allow people to comment directly on audio with a new branch of audio. Here's an example; Someone records a 10 minute update of there day, then the next person listens and they have a comment, say at the 3 minute mark, they can now record their comment directly on the original 10 minute audio. When a third person listens and gets to the 3 minute mark they have the option to listen to the comment or "go down the road".

Overall the goal is to let people have deeper conversations and respond to each other within the context of the audio discussion happening.