Published August 11th, 2025 • 19 minutes
Also available on: Apple Podcasts • Spotify • YouTube
Key Takeaways
- Co-founders create coordination overhead that can kill startups
- AI tools reduced some of my reasons for needing co-founders
- Speed of execution is your biggest advantage in 2025
- Equal equity rarely reflects equal contribution over time
- Decision-making by consensus is startup suicide
- Consider advisors instead of co-founders for domain expertise
- If you have co-founders, evaluate honestly and act decisively
The Co-founder Obsession is Wrong
Co-founder advice is everywhere in startup culture. But Dropbox is not an example of a solo-founded company: Drew Houston built it with Arash Ferdowsi. My argument here comes from my own co-founder experience.
Editorial update, October 11, 2026: Dropbox's own letter from its founders names Drew Houston and Arash Ferdowsi. The original solo-founder example was incorrect. My historical team-count account is preserved; the pair-count formula does not establish that additional founders cause failure or that AI can supply their responsibilities.
I've been on both sides. Started my first startup at 18 with five co-founders. Now I'm running two companies solo with a lean team of five people total.
The reality: In 2025, you probably don't need a co-founder. And AI is changing everything.
My Co-founder Journey: From 5 to 0
When I started my first company as an 18-year-old student, I barely knew anything. I read 'Zero to One', watched Sam Altman's startup series, and felt like I needed co-founders to fill my knowledge gaps. That made sense then.
At our peak, we had five co-founders and 14 total team members.
We were applying to Y Combinator, following all the conventional wisdom about needing a team.
Then reality hit.
By 2023, we almost went bankrupt. We had maybe 2-3 months of runway left. Four of my co-founders left. All of them. I was left with three team members.
From 2023 to 2025, we've only grown our headcount by one person. Five people total, running two companies. Think about that.
Why Co-founders Actually Hurt You
1. Coordination Nightmare
Possible pairwise communication paths grow quadratically. With five people, there are 5 × 4 / 2 = 10 possible pairs. That count is not a measurement of communication workload; in our case, keeping everyone aligned was difficult.
The only time all five of us could meet was 8-10 PM. We'd work until midnight, then rinse and repeat. If one person wasn't there, it took massive effort to get them up to speed.
2. Equity Drama
Breaking down equity is one of the most time-wasting exercises I've ever done. How much should each person get? 20/20/20/20/20? It seems simple, but it's not.
As your startup grows, you realize it doesn't make sense for a CEO who's personally invested everything to have the same equity as someone contributing part-time.
3. Unequal Risk and Investment
Not everyone takes the same risk. Imagine you're 28, quit your job, paying your own salary out of pocket, putting everything on the line.
Your co-founder Alex is still working at Google making $200K, contributing 2-3 hours per day. This compounds over time and tears teams apart.
4. Decision-Making Paralysis
I used to ask all five co-founders to agree before making any decision. I would wait until everyone agreed before moving forward. Looking back, I want to punch myself in the face.
Nothing kills speed like having to get approvals from multiple people.
Why AI Changes Everything
Why I Rely Less on a Technical Co-founder
Editorial update, October 11, 2026: This section was originally headed "You Don't Need a Technical Co-founder Anymore". Its narrower heading preserves the original section link here.
Back in the day, if you weren't technical, you needed a brilliant engineer co-founder to build your product.
Today, that's no longer the case. You have:
- Cursor and Claude for coding
- GitHub Copilot
- Windsurf and other AI tools
These tools can lower the barrier to prototyping. They do not remove responsibility for security, reliability or the domain expertise a product needs.
You Don't Need a Marketing Co-founder
"But George, I'm technical and bad at marketing/sales. I need a co-founder for that."
Equity is not the only way to add help. For my business, hiring for those roles became an alternative to adding a co-founder. That depends on the work, available cash and the contribution a potential partner would make.
How can you tell if someone is actually good at marketing or just screwing with you? You don't know. Never give away 30-50% equity for unproven marketing skills.
AI as Your Co-founder
I lean on Claude, ChatGPT, and other AI tools for advice and references. Then I make the final decision.
You can't make decisions just because AI tells you to, but AI can help me generate perspectives and analysis. It does not establish domain expertise or replace the accountability of a human partner.
When You Might Actually Need a Co-founder
Deep domain expertise in specialized fields. If you're building an aerospace company, rockets, satellites, or hard chemistry/deep tech, you might need someone with that specific background.
But consider making them an advisor first, not a co-founder.
What to Do If You Already Have Co-founders
1. Evaluate Honestly
Are they as passionate about your product as you are? Do you see them scaling with the company? Most co-founders don't scale as fast as the company grows.
I cannot establish the division of work inside other founding teams from their public profiles. My own disagreements are the evidence for this account.
2. Consider Buying Them Out
If you don't see your co-founder as your best wingman to the next level, it might be time for an exit. This is bitter, especially if they're friends, but sometimes necessary.
Evaluate if they're still adding value. If the relationship is strained, buying them out might be the best option.
3. Clear Decision-Making Authority
Don't seek consensus on everything. Make it clear that you'll make fast decisions when things evolve quickly.
You need to be the "bad person" who makes decisions fast and takes responsibility for them.
4. Performance Reviews for Co-founders
Yes, even co-founders need performance reviews. This is counterintuitive, but necessary at scale.
If team members see a co-founder not contributing much while still getting paid and holding equity, it destroys morale. Sit down regularly and assess what they've accomplished.
The Reality: Speed is Everything
In 2025, your biggest advantage as a founder is speed of execution. AI helps with parts of my operational work. Your competitive edge is how fast you can move.
Nothing kills speed like having to convince multiple people before taking action. Sometimes the best co-founder is no co-founder.
What I Learned
Going from 5 co-founders to 0 wasn't easy, but it was liberating. I don't have to worry about what someone else thinks about every decision. I know my space, understand my customers, and can focus on serving them instead of managing co-founder relationships.
I still have a team. I work closely with people every day and enjoy it. I just don't need co-founders who take significant equity and slow down decision-making.
If you're starting out, think carefully about whether you actually need a co-founder or if you're just following conventional wisdom that no longer applies.
Have strong agreements/disagreements? Email me at [email protected] or find me on Twitter @TheGeorgePu. I'll read your comments in future episodes.
New episodes every Monday, Wednesday & Friday at founderreality.com
Editorial update, October 11, 2026: This was the episode's announced schedule in August 2025, not a current publication commitment. Read the current free publication at Founder Reality.