From Pivot Hell to $1M ARR: The Magic Patterns Origin Story

We had a big customer paying but barely touching the product. When we pressed on why, they said editing was painful. We leaned into natural language editing, and that spark became Magic Patterns.

Welcome to Before It Clicked. Today, we dive into how you and Teddy went from iMessage analytics to Magic Patterns, from a maze of pivots to profitability and a Series A led by Dalton Caldwell. Set the stage—what is Magic Patterns and where are you now?

Magic Patterns is an AI design tool for product teams to go from idea to working UI fast, sometimes without Figma. We launched in October twenty twenty-three, crossed one million in annual recurring revenue this summer with just the two of us, and grew steadily by staying focused on real value.

Rewind to the YC application. You had the text analytics project. How did that become a fundable pitch?

It started as a college side project to analyze iMessage, went viral on Reddit in twenty eighteen, then sat on a shelf while we worked as front-end engineers. For YC in twenty twenty-two, we made it open source, offline, and charged two dollars and ninety-nine cents for premium graphs. We had about seventy dollars in monthly revenue, but the point was showing we thought like business builders.

What did Dalton say on the acceptance call, and how did that steer you?

He respected the hustle but urged us to work on what we knew deeply. For us that was front-end and design tooling, not personal text analytics. That nudge set the theme for every pivot.

Day one after that—what did you try first and how did you validate?

We built a Chrome extension to capture interactive slices of a site so engineers could share local work for feedback. Interviews were plentiful, usage was not. We over-indexed on interviews, under-indexed on shipping, and learned the hard way that fast launches beat endless conversations for self-serve tools.

Walk us through other swings and what you learned.

We tried a landing page generator when models were still weak, a design token manager with open source interest but little real usage, and a component library editor—like Storybook on steroids—that resonated only with a narrow slice of Series A and B companies. We even flirted briefly with healthcare and a gaming experiment, then snapped back to design tools.

You mentioned core tenets to escape endless pivots. What anchored you?

Be honest and passionate enough to share it publicly. Do not assume the business model until you map the market. Think bigger when the ICP is too narrow. Learn fast by talking to enough customers, but ship faster.

How did the component editor die and point you to Magic Patterns?

A paying customer barely used it because editing was clunky—we were re-building Figma. Instead of adding more toggles, we let people edit with plain English. That shift unlocked the idea behind Magic Patterns.

You launched Magic Patterns after two months of building. Why so confident with no pre-launch interviews?

Months of failed attempts clarified what teams wanted, and the models had improved. We started coding on August twentieth and shipped in early October. A few hundred users tried it at launch, and seeing servers light up globally told us this had real pull.

When did you start charging, and how did growth unfold?

We launched free, then quickly moved to credit-based pricing after Dalton warned us about AI costs. Month one was about two hundred and sixty-three dollars in recurring revenue. It took roughly a year to reach ten thousand monthly and seven more months to hit one hundred thousand.

What slowed the early ramp, and what flipped the curve?

We chased side quests like a Figma plugin and an email-template variant and messaged too narrowly around components before the tech was ready. The curve bent when models improved and buyers searched for tools that worked with their own design systems—right where we were strong.

You were profitable at serious revenue with just two people. Why raise?

Ambition and capacity. I was buried in sales and enterprise processes while Teddy carried code. Customers wanted to know our team and procurement contacts. We realized we were capping impact, and the market was bigger than two people.

For founders stuck in pivots, what are your crisp takeaways?

Ship sooner. Love the problem and the users. Be ruthless about whether the market is big enough. Do not overfit to one noisy customer. Aim to find at least a handful who are genuinely thrilled.

What is next for Magic Patterns, and how can people reach you?

We are hiring across sales, engineering, and go-to-market to build the best design tool for product teams. It still feels like day one. Connect with me on LinkedIn, and if you hit support at odd hours, there is a good chance it is me replying.