Created an AI app as a non-dev. Here's my journey so far
Three weeks ago, I launched FireTexts: never waste time thinking of the perfect text message again.
Here are some key numbers from the creation and launch:
Design and development cost: $1,900
I mapped out the idea in a bunch of bullet points, gave it to a designer, took the design and gave it to a developer with the requirements laid out (also in bullet points).
To find these freelancers, I messaged a bunch of the top people on Fiverr in the category I was looking for, and went with the most competent sounding people.
Marketing spend: $56 (website + hosting)
I used Webflow which costs $17/month. I may switch this to WordPress or something else that is free.
Some more notes from pre-launch:
It took a few tries to get approved by Apple (Google was easier), but the developer didn't charge much extra to help get it submitted. Once it was approved I asked a few friends to test it out and provide a review (helped jumpstart it a bit).
Metrics for the first 3 weeks:
Downloads: 960
Subscriptions: 20 monthly, 1 yearly
Revenue from subscriptions: $124
Current ARR (I know it's too early): $1,010
App store ratings: iOS: 5/5, Google Play: 4.3/5
I also encountered a few mistakes and learned valuable lessons:
Testing:
Unfortunately, I didn't thoroughly test the final iteration before going live, resulting in a major setback. The registration flow for premium subscribers, which allowed them to save texts and access more features, had a bug that failed to authenticate subscriptions for users who registered via Google or Apple. This led to a poor experience for a significant number of subscribers and resulted in a negative reviews. Luckily, I fixed it pretty quickly and reached out to the reviewer to give her lifetime premium. She removed her review the next week.
Attention to Detail:
When I launched on Product Hunt and garnered attention from blogs and PR, I realized a couple of days later that I hadn't registered the email that I associated with the app. I scrambled to rectify the situation, but it was likely too late as the initial surge in traffic had subsided, potentially causing me to miss out on other opportunities. The lesson here is to double-check everything before launch.
Attention from Others:
The app has already attracted some attention, with one acquisition offer, one offer for free work on the app, and even an old friend reaching out to discuss the app.
Next Steps:
The product at its current iteration is not too useful. It takes a user signing up for premium to really get some good messages of of it (adding custom inputs).
Also, the ability to craft replies based on previous messages is vital. Perhaps too, letting users add their own API and learning their voice.
I've attached a poll to see what you think.
