Launching on Product Hunt After a Year of Building
I didn't sleep for 26 hours. I was terrified. And then Glitter AI won Product of the Day. Here's what that felt like.
Update: Glitter AI won Product of the Day! Thank you so much to everyone who supported us.
I didn't sleep for 26 hours during launch day. I was nervous the entire time. But it paid off.
I'm still processing everything that happened. But for now, I just want to say thank you.
I couldn't have done this without all of you. The comments, the upvotes, even some paying customers on day one. I'm genuinely touched.
Had to rest for a couple of days after the launch, but I'm getting back to everyone now.
The backstory
I've been coding since I was 8 years old. For as long as I can remember, I've dreamed of building a product that people would actually find useful.
My last startup was venture-backed. This time, I wanted to do it differently. No VCs. Just me, building something I believed in, and seeing if people would pay for it.
So today I launched.
I'm both excited and terrified.
The actual timeline
Someone on Reddit asked me why I spent a year building before launching. Fair question. "Perfection is the enemy of progress" and all that. But my journey was a bit more nuanced:
May 2023 - July 2023: I worked on a different project called Conversational Demos and launched it on Product Hunt. But I realized it would require a complex sale- talking to different stakeholders, jumping on live demos, the whole enterprise dance. That wasn't the life I wanted, so I dropped it.
July 2023 - December 2023: I used some of the source code from Conversational Demos for Glitter AI. That's why I count the earlier period in the "1 year of development."
January 2024: I started reaching out to beta testers and found a ton of bugs, crashes, and things that just weren't clear to users. So I iterated and improved.
Around March 2024: I started feeling the early signs of burnout. I was working 12-15 hour days, and it wasn't sustainable. I spoke with my coach and decided to switch to 4-5 hours a day. That slowed things down a lot, but it was the right move for my mental health.
May 15th, 2024: Launch day.
So I didn't sit on the product for a full year. But I probably could have launched with fewer features much earlier. There's a saying: "If you're not embarrassed by your MVP, you launched too late." I'm still working on that one.
Why I built Glitter AI
The story behind Glitter AI is personal.
I hated being CEO of my last startup.
A lot of it came down to being a perfectionist and not knowing how to delegate. I wanted to make sure things were done "right," so I just... did them myself.
Over time, I learned this was a bad idea. The correct approach was to document first, then delegate.
But creating documentation takes forever.
With Glitter AI, I'm trying to fix that. I wish I had this years ago.
How it works
Pretty simple:
- Go through your process normally, but explain what you're doing out loud
- Glitter AI listens to you, takes screenshots, and turns everything into a written guide
- You can then edit and share this guide with your co-workers, customers, or whoever needs it
In my opinion, this is better than video-based tools like Loom for documentation. Here's why:
- No need to start over five times before you "get it right"
- When a process changes, you just edit it in seconds
- The person reading the guide doesn't need to constantly pause and resume a video
I seriously hope this hits home for busy teams who don't have time to create documentation from scratch. It does for me.
The response
One of the things that surprised me most was how many people reached out saying they'd been looking for something like this. Someone mentioned they'd just spent hours updating help desk articles that morning- exactly the pain I was trying to solve.
People asked about integrations with Confluence, Groove, and other platforms. That's now one of the most requested features, and I'm working on it.
A few fellow founders reached out too. One said they were also launching after a year of development (which flies in the face of the usual "ship early" advice). I get it. Sometimes you need that time.
Thank you
I was honestly so nervous leading up to this launch. Building something for a year and then putting it out there for everyone to judge is terrifying.
But you all showed up. And I'm grateful.
More to come.
Product Hunt Launch Q&A
What tech stack did you use to build Glitter AI?
Pretty standard stuff: Node.js with Next.js on the server, React on the client. The desktop app is built with Electron, but to make it actually do its magic (screen recording, audio capture), I had to write a couple of native Node modules in C.
Why did you spend a year building before launching?
The honest answer is I pivoted partway through. I started with a different product, realized it wasn't right for me, then reused some code for Glitter. Add in beta testing, bug fixing, and a deliberate slowdown to avoid burnout, and you get about a year. Could I have launched earlier? Probably. But I'm glad I took the time I needed.
How is Glitter AI different from Loom or other video tools?
With video, you often have to record multiple takes to get it right. And when a process changes, you have to re-record the whole thing. With Glitter, you get a written guide that you can edit in seconds. The person reading it doesn't have to pause and resume a video- they can just follow along at their own pace.
Do you have integrations with tools like Confluence or help desk platforms?
Integrations are one of the most requested features. I'm actively working on them. If you have a specific platform you'd like to see, reach out and I'll prioritize accordingly.
Any tips for launching on Product Hunt?
Make a list of friends, family, and former colleagues ahead of time. Then reach out to all of them asking for support. It's critical to get to a relatively high position as close to 00:01 PST on launch day as possible- that's how you get organic attention from Product Hunt users.
Turn any process into a step-by-step guide
Recent Posts