How I grew my AI Saas to >$100/month in 1 week with no audience

Hello everyone,

I launched https://www.agenthost.ai last week and today it's at $126/m. Agenthost.ai is a hosting and payments platform for GPTs that helps you generate revenue from your OpenAI GPT on your website.

Compared to the $1k-3k/month crowd, $126/m isn't really that impressive, but given that the AI SaaS market is getting crowded, I wanted to share some tips on how I was able to stand out and start generating revenue on day one without an existing audience.

Go where people are desperately searching for solutions

Immediately after OpenAI's Dev Day, GPTs went viral on Twitter, and thousands if not millions of GPTs were created. It was great to see and I made one myself for engineers that today over 300 people use and chat with. Pretty quickly though, it became clear that requiring a ChatGPT plus account for people to even try a GPT was a high hurdle but even so, there were GPTs with hundreds of thousands of users. I began looking at the community forum when I had a question and noticed there were a ton of questions from people who wanted to "talk" to their GPT via an API. Day after day the same question kept getting asked so I wrote a blog on how to do it and started sharing it. The next thing I knew, hundreds of people started visiting the website I built every single day. This is when I realized there was an opportunity here.

Let people try for free, charge for everything else

I used to be one of those VC-obsessed builders who never wanted to charge (see Silicon Valley scene on charging lol) but now that I'm focused on becoming an indie dev, setting up Stripe and our pro plan was the first thing I did for the MVP. If people were not willing to pay for this, I had no interest in building a product for it. I created a free tier with just enough limit for people to verify it works but too small for anyone who actually wanted people to use the tool. This worked great in converting our "curious if it works" people into "Okay it works, I'll pay to actually use it". So now our MRR is growing linearly with the number of free users which feels great.

Track everything early on and be flexible

I was able to glean so much insight about how users were actually using the product through fullstory. I was also able to find bugs and fix them proactively which impressed a bunch of our users. I can't stress enough that early on you'll have no idea how people will actually use your product. You just have assumptions and it's important to validate those assumptions with cold hard data. I've been using fullstory for all of my apps when starting out and it works great. Highly recommend!

Anyways, that's it for now. If folks are interested, I can share learnings/updates once we hit 1 month!