We realized AI wasn't our product. Removing unnecessary work was.

When we started building Pixnom🟠 we thought our biggest challenge would be teaching AI to build great websites.

We were wrong!

The hardest problem turned out to be something much more ordinary.

Forms..

Lots of forms.

If you've ever built a website for a business, you've probably gone through the same routine.

Business name, Phone number, Address, Business hours, Services, Description, Photos, Social links.

It doesn't seem like much.

But after the tenth field, you're no longer creating anything.

You're just copying information from one tab to another.

One afternoon, while reviewing our onboarding, someone asked a question that made the whole room go quiet.

"Why are we asking people to tell us things they already told Google years ago?"

It sounds obvious now.

At the time, it wasn't.

A local business has already spent years building its online identity.

They've uploaded photos, Collected reviews.

Changed opening hours when seasons changed.

Updated phone numbers.

Added services.

Responded to customers.

Kept everything accurate because that's where people discover them.

Then they decide it's finally time to build a website...

..and the very first thing we ask them to do is repeat all of that work.

The more we thought about it, the stranger it felt.

We weren't helping users move forward.

We were asking them to rewind..

So we scrapped our original onboarding.

Instead of starting with a blank page, we started with something every business already has:

A Google Maps listing.

One link.

That's it.

The goal wasn't to impress anyone with AI.

The goal was much simpler.

Respect the time people had already invested.

Everything else came from that decision.

The AI could organize the information.

Write a first draft, Structure the pages, Suggest content.

But none of that would have mattered if the first experience still felt like homework.

That project changed the way we think about building software.

Now, whenever we're designing a new feature, someone inevitably asks:

"Are we solving a problem... or are we asking users to solve it again?"

It's become one of those annoying questions that keeps showing up in meetings.

The kind that slows decisions down.

The kind that kills features.

The kind that's usually worth listening to.

Ironically, people who discover Pixnom often tell us,

"The AI is impressive."

Internally, though, that's almost never what we talk about.

We talk about everything users didn't have to do.

Because we've learned that people rarely remember software for the things it lets them create.

They remember how it made them feel.

Confident.

Fast.

Unstuck.

Like they made real progress in the first five minutes instead of filling out another form.

That has become our definition of good product design.

Not adding more intelligence.

Removing more unnecessary work.