Co-Founding

May 06, 2025

After working on Benchflow.ai (ai eval platform) for ~half a year, I broke up with my co-founder this February. I won't go into detail about the breakup here, but the main takeaway for me is that I committed to the co-founder too quickly. Two weeks ago, I started to explicitly search for a co-founder. I hadn’t done this before because I thought I’d be a productive solo founder, and because I had a personal inadequacy bias (not wanting to ask others for help) and social judgment anxiety (being judged for not already having a strong co-founder). I suspect people are reserved about this topic for similar reasons as to why they are reserved about their goal of finding their dream partner/wife.

This post is partly about what I learned about co-founders and partly about my own search.

My search

I realized some of the best people to ask as part of my search are the subscribers to this blog and their friends. I enjoy working with and spending time with people who are intellectually curious. They should have the ambition to build a massive, fast-growing, and venture-backed company solving an important fast-growing problem. Ideally, they're even more technical than I am.

While I can get excited about many ideas that have the potential to grow into a generational company, I like problems that involve the creation or dissemination of knowledge. Some areas I have thought about recently are ai executive coaching, compliance ai for aerospace and aviation, ai roll-ups (LPOs and SEO agencies), and eval labeling/expert marketplaces.

If someone comes to mind that is also exploring or planning to start something, please forward them this blog post or blurb:

I (Moritz Wallawitsch) previously co-founded BenchFlow.ai (backed by Jeff Dean) and RemNote.com (GC-backed, ~2M ARR). Exploring several ideas including BPO ai roll-ups and ai exec coaching, but open to others. Interested in connecting with other technical founders in similar situations. (moritzw.com)

What I realized and learned

Early-stage VCs are right to prioritize the co-founders and their relationship over everything else because a great co-founder team is able to pivot around until they find the right problem and reach PMF. 70% of successful startups pivoted at least once before finding the right business model. I wrote about this ideation process on my blog before. The Cursor/Anysphere team for example, started in text to CAD, moved to an ai email client idea before they started working on Cursor.

A great co-founder is someone you would have met anyway. I.e., you share existing friends with them. You should also enjoy spending time with their friends and partner. You can test this by going to the same event or hosting a dinner or poker night.

Another interesting criterion I re-discovered is the airplane seat check. I.e., would you enjoy sitting next to them on a long plane ride. If in a meeting, do you naturally go over the allotted time?

In my experience, it is a lot more fun, rewarding, and productive to come up with the idea together (or independently come up with the same idea) that you work on. Analogously, parents feel more invested in children that they helped conceive compared to those from a previous relationship.

The majority of the companies I admire had multiple co-founders. Examples include: Palantir, Anduril, early Apple and PayPal, Flexport, Faire, Scale AI, Cursor, Mercor, Ramp, Stripe, and Contextual AI.

I think a match in skills is important but underappreciated is a match in thinking styles. The minds of the co-founders should complement and augment each other. Brian Armstrong said:

I was searching for someone who would be genuinely excited about my idea, but also complementary to my skills—someone who could push me to think bigger and offer perspectives I hadn't considered. I wanted to leave meetings feeling energized and having learned something new.

Scaling Knowledge

Discussion about this post

Scaling Knowledge

Co-Founding

My search

What I realized and learned

Resources

Discussion about this post