Trying to be a real first-time SaaS founder has been interesting over the past two months or so. I’ve finally decided to take it seriously.

I’m building Vizzly — a better way to visual test your applications. A ton of my career has been around testing. I’ve spoken at JSConf about testing UI and building a test framework. For some reason I’ve always been around testing. So… why not a testing SaaS product? Makes the most sense.

None of this is in any particular order and none of it is that serious — just what I’ve learned so far.

Ship early and often

Stupid cliché. Still true.

I spent two years with a friend working on a different product. We tried to think about everything. Make it perfect! We were engineers who had helped multiple startups get acquired. We knew what we were doing, right?

Wrong. We shipped nothing. We sat on perfection. That product still hasn’t seen the light of day.

With Vizzly, I shipped much faster, but I still had that “make sure everything is perfect” itch. I implemented (arguably) too many features before even telling anyone about it. I behaved like I had one shot at “launching.”

Reality: you get as many launches as you want. No one has heard about your product. They probably don’t care about it either, honestly. So just ship it. Keep shipping and keep telling people. You’ll have to craft and iterate on your message; you’ll make mistakes; the only way that happens is by doing.

If you’re wondering, yes that product is/was PitStop. Maybe one day it’ll ship.

AI is a huge enabler

I’ve been able to tackle much larger projects thanks to AI — really, all of them. I use Claude Code the most, but Codex/ChatGPT and some Gemini are also super helpful. I love using MCPs and sub-agents for things like managing Linear. Breaking down ideas, projects, tasks, and bugs has gotten so much easier that a lot of the time I can hand the issue off to an agent like Claude Code or Codex and it does a pretty solid job.

It’s not perfect, but it’s like having another half of a brain to complete my half of a brain.

A year ago I wasn’t tackling something like Vizzly on my own this fast. It’s a fully-featured visual testing product with background build processing, full SDKs, and product details like position-based comments. That was not happening without AI and not without a team.

Small, sustainable > VC rocket ship

Because AI levels me up, I think more small teams, small companies, and non-VC founders can make this work. That’s the route I’m trying to take: a SaaS that generates revenue and isn’t beholden to huge-growth-for-ROI.

I want modest projects that work well and people enjoy using. In return, hopefully they pay my bills. Naive? Maybe. Possible? I think so.

Cold emailing is hard (and necessary)

Building interest is hard. Cold emailing is hard. I hate getting those emails too. But you have to take your shots. No one knows who you are or what you do. Maybe someone will find it useful, but it’s up to you to let them know.

No one gives a shit about your product. It’s your job to change that.

Lay the marketing groundwork early

I’ve spent time making sure SEO and basic marketing pieces are in place from the start. I also felt like the site needed to look “legit”: About page, blog, the usual. These are long-term things, but laying the groundwork early should pay off.

Do I love pitching my peers? Not really. “Yo, here’s my mixtape (product), check it out.” It feels sales-y, but I believe Vizzly is genuinely valuable, so I’m doing it anyway.

I don’t think users want “AI” (in the product) right now

Irony alert after that AI section: I’m avoiding AI in the product for now.

Most implementations I see are a bolted-on OpenAI API call with a prompt. It doesn’t feel great or useful. Maybe as we figure out how to build with LLMs, this will improve. For now, it’s not helpful for most users, including me.

I’m bullish on AI. I just don’t want to ship “AI” for the sake of saying “AI.”

Where I’m at

Two months in:

  • Ship, then ship again.
  • AI helps me build faster; I don’t need it inside the product (yet).
  • Cold outreach is uncomfortable and still required.
  • Marketing groundwork matters more than I wanted to admit.
  • Indie SaaS actually feels possible.

Back to shipping.