The Look I Couldn’t Let Go

When I started building MyBrandPilot, I thought the hard part would be the “smart” stuff.

The AI agents. The strategy generation. The website ingestion through Firecrawl. The multi-step workflow. The whole ambitious, probably-too-big-for-one-person thing.

And yes, all of that was hard.

But the design decision I obsessed over most was much simpler:

What should this thing feel like when someone opens it?

At first, the project had different names. MyAdPilot. StrategyPilot. Eventually it became MyBrandPilot. The names shifted as the idea got clearer, but the feeling I wanted stayed the same.

I wanted it to feel like elite consulting.

Not flashy. Not cute. Not “look at this cool AI toy.” I wanted it to feel like the kind of place where serious brand and marketing decisions could happen.

That may sound dramatic for a person with no coding background building software with an AI app builder. Believe me, I had plenty of moments where the screen was broken, the AI misunderstood me, and I thought, “Who exactly do I think I am?”

But the design mattered to me because the product itself was ambitious.

MyBrandPilot was not just a single-page tool. I was trying to shape it into a phased, enterprise-style SaaS product. There were organizations, roles and permissions, website analysis, specialist AI agents, and even a required Critic pass so the output wouldn’t just be the first shiny answer the AI produced.

That kind of thing can easily become a mess.

And if the interface feels messy, people assume the thinking underneath is messy too.

I know I do.

When I open software and it looks cluttered or gimmicky, I brace myself. I start looking for the catch. I don’t trust it as much. Maybe that’s unfair, but it’s true.

So I kept coming back to the same question: would this feel credible to someone making real business decisions?

Not “would it impress another AI builder?”

Not “would it look trendy on social media?”

Would it feel calm, structured, and serious enough that a person could sit down and think?

That led me toward a consulting-style design system. More restrained. More polished. More like a sharp presentation than a loud dashboard. I wanted the product to suggest confidence without shouting.

The funny thing is, I probably spent too much time thinking about this.

There were moments when I was nudging the builder about layouts and sections while some bigger piece of the app still needed work. Part of me knew I was avoiding the harder problems. Design can become a hiding place. It feels safer to adjust the room than to deal with the machinery behind the wall.

But I don’t think it was only avoidance.

For this project, the design was part of the promise.

If MyBrandPilot was going to produce marketing deliverables and strategy work, the experience needed to respect the seriousness of that work. The look had to say, “Slow down. We’re going to think this through.”

That mattered to me because I’m building as someone who remembers work before everything became an app. I remember printed reports, meeting rooms, binders, decks, and the weight of a recommendation when real money or reputation was on the line.

AI can make things fast. Sometimes too fast.

So I wanted the design to put a little gravity back into the process.

Not to make it old-fashioned. Not to pretend AI isn’t changing everything. But to make sure the tool didn’t feel like a slot machine for business advice.

That was the design decision I couldn’t let go of.

I wanted MyBrandPilot to feel trustworthy before it tried to feel clever.

Small takeaway: sometimes the “look” is really a decision about trust.