Don’t Wait for Perfection: How We Took Aha Studio From 6 AI Workers to 31 in 9 Days

Build in Public · #25

Aha Studio went from six AI workers to 18 in its first week, and to 31 by the time the early bird closed, 9 days after launch. That only happened because we didn’t wait for perfection. Every worker was brought to a good standard and tested before it went in, always with the plan to come back and improve it. Here’s how that works in practice, including the rough edges I’m happy to show you.

Aha Studio was one week old when I demoed it live for the second time.

By then it had gone from six AI workers to 18.

Two days later, when the early bird closed, it had 31.

Was every one of them perfect? No. Were they good, tested, and doing real work for people? Yes.

That’s the standard I build to.

Good and tested, then improve

The rule is simple:

1. Get the worker doing its workflow to a good standard.
2. Test it. Properly, on real jobs.
3. Put it in.
4. Come back to it and make it better. That was always the plan, from the first day.

With AI you can more or less speak things into existence now. Then you spend your time tweaking, polishing and getting it right. The building is fast. The polishing is where the time goes, and it never really stops.

The foundation came first

Building fast doesn’t mean building carelessly.

We spent weeks and weeks building the base: the foundation every worker sits on. That’s what made it possible to let loose afterwards and build tons of workers quickly.

I deliberately started with six workflows I knew inside out, before building more. Then we planned a couple of weeks of consolidating: pulling all the little bits and pieces together, so every worker shares the same features, like being able to steer and edit what it builds in plain conversation.

Once that foundation is right, new workers come thick and fast.

The rough edges, in public

Here are a few things that weren’t finished when I showed them live, and I said so:

• Some URLs were longer than I wanted. I’d already spotted it. With everything else we were building, it was one of those jobs I still had to get round to.
• The funnel builder is at stage 0.5. It has the essentials: webinar registration pages with SMS registration, wait lists, the basics. It’s nowhere near a full funnel platform, and I told people that.
• We were testing like hell that day, mostly the calculators, improving the output and the design.

None of that stopped the workers doing useful jobs for people that week.

Let the members steer it

There’s a “suggest a feature” button inside Aha Studio. Members were already using it in the first week, and some of their suggestions were already live.

If I like a suggestion, I get it built.

That only works because the product is out there being used. Feedback from real members on real jobs is worth far more than another month of me guessing in private.

Improving also means removing

AI moves fast. A worker that’s useful today might be irrelevant in six months, or might not deliver the outcome we expected.

So I’ll take things out as well as put them in. Adjust a worker, build a new version, or drop it. That’s part of not waiting for perfection too: nothing is precious, and nothing is finished.

What I’d tell anyone sitting on a product

• Define “good enough” before you start, and test against it.
• Build the foundation properly, then move fast on top of it.
• Be honest about the rough edges. Point them out before your buyers do.
• Put a feedback button in, and act on what comes through it.
• Plan to come back. Shipping is the start of the work.

My Take: The entire Aha Studio was built super fast, but we didn’t wait for perfection. We got the AI workers doing their workflows to a good standard and tested them, always with the knowledge and the intention to come back to each one and improve it. If I’d waited until every worker was perfect, there would be six of them sitting on my computer, still waiting for their first user. Instead people were using 18 in the first week, telling me what they wanted, and helping me make each one better.