In 2009, seven of us co-founded a startup called Ekkli. In the year and a half we existed, we could not close a single investor meeting. The reason was sitting in the room with us, and we built the company that way on purpose. It is the most useful lesson about founding I have, and almost nobody talks about it.
The product was a collaboration tool for democratic decision-making. It heard everyone’s opinion in a group, ran semantic comparison across the messages, visualized where the discussion was converging, and helped large groups reach a conclusion without any single person running the meeting. It was a clever product. We were proud of it.
The Product Was Democratic
The premise was good. Most group decisions stall because three people dominate the room and the rest go silent. The tool let everyone contribute at the same time, weighted the contributions semantically, and made the convergence visible. On one screen, you could see what a hundred people thought about a question and where the consensus was forming.
The seven of us were good friends. We met for overnight coding sessions in homes, cafes, and shared offices. The culture was warm. It was a really good time.
So We Made the Company Democratic Too
This is where the mistake started. We decided that if the product was about hearing every voice equally, the company should run the same way. All seven founders would be equal. No CEO. No CTO. No founder hierarchy. We would make every decision the way the tool made decisions, by surfacing every opinion and converging.
It is the kind of decision that sounds principled when you are sitting in a kitchen with friends, agreeing on it. It is the kind of decision that becomes structurally impossible the moment the company has to interact with the outside world.
The Meeting We Could Not Close
The first sign was the first investor meeting we tried to take.
We had a meeting on the calendar. The investor wanted to know who was on the board. We did not have a board. So we sat down, the seven of us, to decide.
We tried a rotating board. We tried picking three people for that meeting only. We tried picking three people for the company. We could not agree on who. Not because anyone was being difficult. Because nobody wanted to overrule anyone else. We were a group of friends running a tool that said “everybody’s voice counts equally”, and we could not produce three people willing to be the official voice of the company for an hour.
We never closed the meeting. We never closed any of the meetings we tried to take after it. The structural problem we had built into the company was that nobody had the right to speak for it. So nobody did.
Then We Spent a Year on Symptoms
We did not see this as a structural problem at the time. We saw it as a series of unrelated issues we could solve one by one.
Lead generation was not working, so we tried various approaches. We could not agree which one to commit to, so we tried them all in small versions, and none of them got fully executed.
Customer conversations were not landing. We rotated who took the calls. Different founders gave different signals to the same customer because nobody owned the relationship.
The product roadmap kept growing because every founder had features they wanted to add, and no one was empowered to say no. Every release was a compromise that excited none of the people who built it.
Every one of these problems looked like its own thing to us. Every one of them was a downstream symptom of the same root cause. We were seven equal founders, and equal meant nobody decided.
The Lesson, More Than Fifteen Years On
Eventually we closed the company. We had a really good time together. The product was clever. The friendships are intact. The company never made it past the first investor meeting it tried to close.
The lesson I take from Ekkli, fifteen years on, is the one I now give to every founder who walks into a call with me before they have signed a co-founder agreement.
Co-founders should be complementary. Not equal, not similar. Complementary.
Everyone already knows the surface version of this. You want one technical co-founder and one commercial one. You want different domain backgrounds. You want non-overlapping skill sets. Fine. That is the easy half.
Why Personality Complementarity Is the Quiet Half
The half almost nobody talks about, and the half that killed Ekkli, is personality and soft skills.
If one of you is the pusher, the other one needs to be the executor who finishes what got pushed. If one of you is the introvert who thinks before speaking, the other one needs to be the extrovert who carries the room when carrying matters. If one of you defaults to “yes, let us try it”, the other needs to default to “no, not this quarter”. If both of you are agreeable, you will agree your way past every difficult conversation the company needs to have. If both of you are pushers, you will burn each other out by month six.
This is what we got wrong. We picked each other because we liked each other. We were all easy people to spend time with. None of us was the unpleasant one who would tell an investor “I am the CEO, talk to me.” None of us was the one who would tell another founder “we are not building that feature, drop it.” We were seven nice people who agreed too much.
The diagnostic question I give founders considering a co-founder: in the meeting where you and your co-founder fundamentally disagree, who breaks the tie, and why does the other one accept it? If there is no clean answer, you have not yet found a complementary co-founder. You have found a friend.
You can absolutely be friends with your co-founder. The Ekkli friendships outlasted the company. But the things that make somebody a great friend, like being agreeable, accommodating, and nice, are sometimes the same things that make them a terrible co-founder for the conversations a startup actually needs.
Let’s Talk
If you are evaluating a co-founder situation, or you are already in one and trying to figure out which of your problems are downstream of a founder dynamic you have not named, that is the kind of question I work through with founders all the time. I take on senior async architecture and product-engineering work for early-stage teams making exactly these calls. If that sounds like your situation, reach out.
