I did sunset an AI product and launched a new one.
Let's learn from our mistakes, to avoid repeating them in the future.
The last couple of months have been crazy.
The Twitter API drama impacted my businesses Hivoe and Inboxs, and I had to quickly find a solution (but I’ll talk about this in the next issue).
As you probably know, I have a full-time job as a software engineer, and I’m trying to run my own digital business to find my freedom.
Both in terms of time and finances.
Considering the uncertainties of the Twitter (now X) platform, that started early this year, I looked for new product ideas.
I tried to create an AI product but decided to stop after a few months, and here’s how it went.
The new “brilliant” idea.
A friend of mine mentioned using AI for writing business emails, sparking my interest in creating something similar.
What if I could create a ubiquitous browser extension, to provide writing and replying experiences using AI?
I even found the name the same day OmniWrite.ai
I started building it, thinking that there are many businesses making tons of money with something similar using AI, so why not even me? I’m just looking for €10k MRR to leave my job and live my dream.
Similar products are CopyAI, Jasper AI, and WriteSonic.
I spent weeks building an extension that brings AI writing to every browser tab. I hired a designer to speed up the process and get good-quality visuals.
After a couple of months, the product was ready, launched on Twitter, and started getting some traffic, so I kept waiting for the first subscriber. I even created a Slack community for early adopters.
Feedback was good, but not exceptional, and the engagement in the Slack community was not that great.
Days were passing, and no subscribers.
I finally decided to stop working on it after a couple of months, these are the reasons:
Getting good writing results with OpenAI / GPT is not trivial. It works great for writing business emails, but not for many other use cases (writing and replying to tweets for example).
There’s high competition, and some of those companies got millions of dollars in investments. Plus, they have a ton of small tools to tackle different use cases, like copy for ads, writing emails, landing page copy, google ads headlines, blog post writing, Facebook ads, brand voice, and so on.
Plus, I didn’t get any paying customers with 1.8k unique website visitors, and 65 Chrome extension installs.
All these elements convinced me that, as an indie hacker, this was not the best product to continue working on. I’m looking for good growth products because I don’t have much time to spend and I’m not that good at marketing, in general.
I can’t live without a product
This is what I realized when (spoiler) I sold my two main products, Hivoe and Inboxs.
I can’t live without having my own business.
So after the experience with OmmiWrite, I started looking for new opportunities.
I’ve always done customer support for my products, and I’m excited about the opportunities of automating part of that job with ChatGPT-like custom products.
So I did a prelaunch of a new product idea (validating a new product, is always a good idea)
I sold 30 lifetime deals, at $69 each, in 24 hours. This is what I call validation 🥳
And it gave me the confidence to start working on the product, that’s already live and available to the public at userdesk.io.
That’s all I have for today’s issue.
Was it a good idea to stop working on OmniWrite?
Will I succeed with Userdesk?
The future will tell.
If you have any questions, feel free to reach out to me on Twitter (X)
See you next time 🙌
Good insights before building AI products. Thanks Luca
Great Job!