How To Build a Company You Actually Want to Run | Julian Reacts

Key Takeaways

• Make things before you make a business. Dhravya Shah open-sourced everything and never "put anything under a paywall" — the audience that followed him for fun became the launchpad for Supermemory.
• The pull to build is a signal. Francis Davidson incorporated his next company the same month he left Sonder: "building a company is probably the closest thing I'll ever come to doing artwork."
• Run the ten-year test. Thais Castello Branco's bar for an idea: "Even if I sold Taste tomorrow… I would go and start it again."
• "The cure to burnout is not rest. It's momentum." Flo Crivello planned six months off at his lowest point — then pivoted instead, and the energy "came back immediately."
• Protect the work only you can do. Claire Vo front-loads the work she avoids; Paul Bakaus warns that hiring builders first turns the founder into "the operations person. I think that's a trap."

Listen: Apple Podcasts | Spotify

Introduction

Every startup question eventually collapses into a simpler one that gets asked far less often: is this a company you actually want to run? Not for a launch week. For years. Through the season when nothing works, through the hire that changes your job description, through the Tuesday afternoons that make up most of a founder's life.

In the third Solo Talk episode, Julian Weisser goes back through six conversations from the show — Dhravya Shah (Supermemory), Francis Davidson (Odessia), Thais Castello Branco (Taste Labs), Flo Crivello (Lindy), Claire Vo (ChatPRD), and Paul Bakaus (Impeccable) — and pulls one idea from each that deserved more spotlight than it got the first time. Played in sequence, they turn out to be a single argument: choose work that pulls you, test the pull before you commit a decade to it, and then build the company so you keep doing the part of the work you're best at.

None of the six answered the question the same way. That's the point.

Make Things Before You Make a Business

Before Supermemory was a company, it was one of many things Dhravya Shah made because making them excited him. "It was kind of my way of doing art," he says. "You know how people make paintings." Everything was open source. Nothing went behind a paywall. The first tweet about what became Supermemory was a self-deprecating joke about giving away the code.

What looked like play was quietly building the most durable asset an early founder can have: an audience that trusts the builder, not the pitch.

"People were excited about me as a builder… because people really think of me as someone who's just in for the love of the game." — Dhravya Shah

Julian's reaction cuts against the default startup playbook: "All too often, people are trying to shoehorn themselves into a business idea as quickly as possible to launch a startup and raise funding. I think that's generally the wrong way to build a great business." Explore your interests in public, he argues, and the people who love what you make will be there when you make the big swing.

The Pull to Build Again

Francis Davidson has already lived the full arc — he built Sonder into a roughly $2B company and took it public, then watched it collapse. The revealing detail is what he did next: "Since the moment in June that I left Sonder, I incorporated the new business immediately. I knew I would start another company. This is who I am."

He is careful to say he's no artist — "there's no creative genius really behind what I do. I'm much more of an operator." And then he lands the episode's most quotable paradox:

"Building a company is probably the closest thing I'll ever come to doing artwork." — Francis Davidson

Julian calls this the artistic urge: an intrinsic pull to put something into the world, independent of what anyone will think of it. His advice is unhedged — if you have it, listen to it.

A Problem Worth Ten Years

Thais Castello Branco took the opposite path from Francis. She didn't start the moment she could; she waited for a problem that felt like hers. Her bar was explicit: something she could spend ten more years thinking about, something that didn't feel like work.

"Even if I sold Taste tomorrow — which is not going to happen — I would go and start it again." — Thais Castello Branco

The test had teeth. She left Exa — which, as Julian notes in the episode, recently announced $100 million in annualized revenue — with no rational incentive to go. "It really was something that became almost stronger than me," she says. "I cannot stop thinking about it."

Julian turns the three openings into a framework: are you a Francis, who knows he'll build regardless of what it is? A Thais, who won't commit until the problem owns her? Or a Dhravya, who makes things until one of them becomes the company? There's no right answer, he says — only a right answer for the individual.

Momentum, Not Rest

The episode's sharpest idea comes from its darkest stretch. Flo Crivello describes the two years before Lindy as "chewing glass and staring at the abyss" — Sisyphean grinding with nothing to show for it. He was sure he needed six months off. Instead, he tried the pivot.

"The cure to burnout is not rest. It's momentum. It comes back immediately — the moment you're excited again about what you're working on." — Flo Crivello

Julian is careful about what this does and doesn't mean. "Keep going" is not an instruction to keep doing the thing that isn't working — that, he says, is extremely draining, and betting on it requires conviction that some external condition will change. The move is to pull the thread: stop, or dramatically change what you're doing, and let new momentum supply the energy. Like going for the run you don't feel like taking, the thing you're avoiding is usually the thing that gives the energy back.

Build the Company Around Your Shape

Claire Vo runs ChatPRD alone, and she is unsentimental about her own tendencies: "I will literally build any feature — any feature on the face of the earth — to avoid marketing." Her response isn't willpower. It's architecture: a constellation of nine AI agents covering marketing, sales, support, and an EA, plus a day designed around her weaknesses — sales calls, marketing, and podcast work in the morning, building in the afternoon, because she'd build anyway.

Julian's gloss: inspiration has a short half-life, and willpower is weakest exactly when the avoided work is hardest. Do the things you don't want to do first.

Paul Bakaus completes the thought from the other direction. He's building a bigger team — and refuses the standard first hire. Most founders, he argues, hire an engineer or a designer to get the building off their plate, and wake up as their company's operations person.

"You are the operations person. That's now your role. I think that's a trap." — Paul Bakaus

His alternative: hire a chief-of-staff-type operator and keep the building. Julian extends it into the episode's closing argument — early investors are investing in you and their belief that you'll figure it out. The things that make you special are the things worth preserving. And the people you hire to protect them are, in a real sense, replacing your co-founders: compensate them like it.

The Case for Choosing Your Own Way

The quiet thesis of this Solo Talk is that the six founders disagree — productively. Dhravya explored his way in; Thais waited for conviction; Francis never needed a reason. Claire stays truly solo with systems; Paul hires so he can stay a builder. What they share is that each built the company around who they actually are, not around someone else's template.

That's the version of "build a company you want to run" worth taking seriously: not a vibe, but a set of testable choices about what you make, when you commit, and what you refuse to hand off.

About This Episode

Solo Talk is the Solo Founders Podcast's clips-and-commentary format: Julian Weisser revisits past conversations and pulls one under-discussed idea from each. This edition features Dhravya Shah (founder of Supermemory, who raised $3M at 20), Francis Davidson (founder of Odessia; previously built Sonder to ~$2B and took it public), Thais Castello Branco (founder of Taste Labs, $18.5M seed), Flo Crivello (founder of Lindy, $50M raised solo), Claire Vo (founder of ChatPRD, the bootstrapped AI product used by 100,000+ product managers), and Paul Bakaus (founder of Impeccable).

Subscribe to the Solo Founders Podcast on YouTube, Apple Podcasts, or Spotify.
Join our newsletter for weekly insights from solo founders: solofounders.beehiiv.com
Apply to the Solo Founders Program: solofounders.com/program