Hey IH,
I'm Abhi, and I've officially joined the "building in public" club.
My project is an AI resume builder called Matcha Resume. It's been a tough market to crack, but I just hit a milestone that feels huge: my first 10 paying users.
It's not $10k MRR, but it's 10 real people who believe in the product enough to pay for it. The journey from 0 to 10 has been a grind, and I wanted to share my biggest lessons so far.
My assumption: People needed better templates. The reality: People needed better content.
My first 5 users gave me brutal (and amazing) feedback. They didn't care about fonts; they were desperate to know what to write. They were stuck on "how do I turn my boring job into measurable achievements?"
The lesson: I stopped adding template features and went all-in on AI-driven content analysis. The tool now scores their bullet points (like Challenge-Action-Result) and suggests how to fix them, not just how to format them.
The lesson: Users don't want a black box to replace them. They want a tool to assist them.
I pivoted the AI to be a "co-pilot." It doesn't write for you; it identifies specific weaknesses in your existing resume.
"This bullet point is passive voice."
"You're missing 5 critical keywords from the job description."
"This achievement has no measurable result."
This transparency was the unlock. Users finally trusted it because it was enhancing their own work, not making stuff up.
The lesson: Getting 0 users is way more likely than getting your idea stolen.
The moment I started posting on Reddit and talking to career changers, I got my first 3 signups. They are now my most valuable source of feedback and have shaped the entire product.
I'm still very much at the beginning, but 10 users feels like validation. My next goal is 100.
I'm still figuring out marketing and how to reach job seekers. If anyone has had success marketing a niche B2C SaaS, I'd love any advice.
Thanks for reading!
P.S. If you're job hunting or just want to see the tech, you can try the tool I built to solve this: matcharesume.com
The distinction between 'Intelligent AI' and 'Useful AI' is a brilliant point. I'm facing the exact same challenge with my AI food scanner. At first, I thought users wanted a 'judgment' on their food. But I quickly learned that people hate being judged by an app; they want transparency.
I pivoted the AI to be a 'context provider'—it doesn't tell you 'don't eat this', but it highlights the hidden additives and processing markers so you can decide for yourself. That shift from 'black box' to 'transparency tool' was the real unlock for our user experience. Congrats on the first 10 paying users!