Month 1-4 Recap – A Good Start

This post is part of the transparency series

One major goal when transitioning to Insomnia full-time was to share as much about the process as possible. As an independent developer with little time, success heavily relies on the effective use of shared knowledge, tools, resources, and peers. Without things like blog posts, open source, and friends, Insomnia would not be where it is today. This post is an attempt to give back.

This is the first post in a series titled Transparency, where I share numbers and talk about what I've done, what I'm thinking, and the mistakes I've made. If there is anything you think that's missing from these posts, feel free to reach out and let me know. I'll try to accommodate as many requests as possible.

Alright, let's dive in.

User Growth and Other Stats

I'd like to start this post by going over some interesting statistics on how Insomnia has been doing so far. Since Insomnia is a desktop app, these stats will be divided into application stats and website stats.

Active App Users

Here is a screenshot from Mixpanel on daily active app users. I've chosen to define an active user as someone who has sent at least one API request from the app in a given day.

Insomnia Active Users

If you're curious, the lines marked on the graph are the following.

As you can see, most of the user base can be attributed to being featured on Hacker News in early September. Since then, retention has been high and growth has been steady but slow. High retention is good. Slow growth either means I'm doing a poor job at getting the word out, or the target market is too small. I have a suspicion it's the former, because similar applications exist with millions of users.

Website Traffic

In my opinion, one of the biggest mistakes made by independent developers is not having a proper website. A website is one of the only ways of gaining users organically (unless you are in an app store), so a website should be a huge priority.

Here are some website stats taken from Google Analytics over the same time period as the one above. Notice the huge spike from Hacker News.

Insomnia Web Traffic

It's harder to see in this graph, but if you ignore the spikes, there is also a steady growth here. Another good sign.

Miscellaneous Stats

In order to keep this post relatively short, I've condensed the rest of the stats into tables. If there is anything else you want to see here, just let me know.

App Stats

Value Description
2,000 daily active users
5,000 weekly active users

Website Stats

Value Description
110,000 sessions
00:01:22 average session length
54% average bounce rate

General Stats

Value Description
29 App Releases
320 Help Scout email conversations
10 months Financial runway

Financial Spend

Cost Description
$300 S3 for download hosting
$150 SSL wildcard certificate
$70 Business registration
$50 Heroku
$40 Sentry

Financial Earnings

Cost Description
$0 :(

Insomnia currently has no premium features, so has been unable to generate revenue. However, A beta version of the first paid feature (sync) is almost complete, and additional team collaboration features are in the works. This number should be greater than zero in the next post.

Time Allocation, Priorities, and Roadmap

The last four month have been crazy. Full-time work on Insomnia began on July 5th, and the app launched on July 24th. Effectively, the last four months looked like this.

  • July – Finish development for initial public launch
  • August – Implement remaining core features (environments, and cookie support)
  • September – Bug fixes, website, small improvements, and talking to users
  • October – Research and implement cloud sync features

I don't have the next four months planned out, so I can't tell you exactly what will happen, but I do have some general goals.

  • Have sync and team features available for purchase by December
  • Begin open-sourcing parts of the codebase
  • Implement highly requested features

    • Binary file support
    • Request chaining and/or scripting
    • More import sources (Postman, Swagger, etc)
  • Write more content like this

Mistakes and Lessons Learned

Here are the main things that I have learned so far. Each one of these could probably have a separate blog post written about it.

  • Talking to users takes is extremely valuable, but takes a lot of time

  • Code quality usually only matters for the core components

    • Because those are the hardest parts to refactor
    • For Insomnia, it's the data model and storage system
  • Hosting Electron apps (50MB each) on S3 is expensive

    • $200/mo after Hacker News post
    • Used Cloudflare to cut costs (now $10/mo)
  • People don't like receiving app updates every day (who knew!?)
  • Working at home makes you feel like a slob

    • Have transitioned to coffee shops
  • Cloud sync is more complicated than expected

Wrap Up

Overall, Insomnia is doing well. The user base is growing, the app is becoming more stable, and the people using it are amazing to interact with. However, it's not generating revenue yet, and user growth is slow. Considering the sync beta should be ready to go live this week, the next few weeks should be very interesting.

I plan to try and write one of these posts every month, so keep your eyes out for that. Again, if there is anything else that you would like to see included in these posts, just let me know.

This post is part of the transparency series

Need a better way to test APIs?