Heres how it felt in the weeks before I resigned from my last startup: I couldnt sleep. I couldnt eat. Resting pulse at 120. I had reached a point where I couldnt agree with my co-founder over the future of the company. I had to step away from the startup for which I shed blood, sweat and tears. I didnt want to do it, but I reached a point, physically and mentally, where I couldnt handle the stress anymore.

This is the first public post in which Ive ever talked about it, but through advising hundreds of startups, Ive learned that my story is not uncommon.

Every co-founder situation is different, but one common problem that keeps popping up revolves around how the founders engage in conflict: either not enough, or far too much.

Being successful will mask co-founder problems

Founder drama happens even in situations where you wouldnt expect it to crop up. Success will cover up many sins. When things are going up and to the right, things might be going wrong underneath and you wont be aware of it. Its the black ice of startups. Its dangerous because every startup will hit the skids sooner or later. You cant count on good times forever winter is coming.

Posterous, the startup I co-founded in 2008, grew 10X yearly and became a top 200 Quantcast website in that time. But by the end of 2010, growth had flatlined. When things were going well, we were too busy keeping the site online to have anything to disagree about.

I learned the hard way that if you havent prepared for conflict in your co-founder relationship, youll be at each others throats right at the moment when you most need to be working well together.

We learned history is not enough youve got to maintain it like any relationship.

The mistake that my co-founder and I made was in avoiding the dynamics of our co-founder marriage altogether. We rarely spoke directly and honestly with one another. We didnt stop to reflect on what he needed or I needed. We never sought professional support to ensure the health of our partnership. When the honeymoon ended, there was no healthy foundation to support the company.

During my time as a partner at Y Combinator, we always looked closely at how well co-founders knew each other before they started. Most people think of good co-founding pairs in purely functional terms: a business person paired with a technical person. This is deeper than that, because when conflict does arise (and it always does), if you have nothing in common other than the startup, youll struggle to find common ground at the worst of times. Its necessary for founders to have something in common, but not sufficient in and of itself.

In my case, I had known my co-founder for more than eight years, and we had been friends since college. We had history, but we learned history is not enough youve got to maintain it like any relationship. It isnt enough that youve been friends for years. It matters what your relationship is like now.

Avoiding conflict

With hindsight, I now realize my rift with my co-founder was entirely preventable. We stopped spending time together because we were avoiding conflict. I wanted so much for us to succeed, and I wanted so much for us to be great co-founders (and to maintain the narrative that we were close and had a good partnership) that I skipped the hard work that it takes to get that relationship and do our best work: embracing conflict and resolving it. Its a problem that Ive recognized over and over again in founders with whom Ive worked both as an advisor and investor.

If you havent spent time together outside of work, ask yourself why? If you see your co-founder coming down the hall, do you alter your course to avoid them? Do you try to keep your interactions at a minimum? If so, thats a clear sign youre avoiding conflict by just avoiding them, period. Thats just not going to work.

Founders sometimes take the avoidance route to an extreme. One recently told me that he decided to talk to his co-founder only once monthly, claiming it to be the only valid way forward. This was a pretty extreme case of avoidant behavior! I told them they had to either radically spend 10X more time working through issues and resolving them, or prepare to split.

Successful co-founders actually embrace conflict.

Its the same script all over again: Co-founder conflict is bad, so if we minimize how often it happens, thats the best possible case. Its a trap!

My executive coach, Cameron Yarbrough, points out that this is usually the moment the Four Horseman of the Apocalypse show up: defensiveness, criticism, contempt and stonewalling. When psychologist John Gottman (author of the Four Horseman concept) identifies those behaviors in marital relationships, hes able to predict relationship failure with uncanny accuracy. The same thing holds true for co-founders.

Successful co-founders actually embrace conflict, and are constantly in the process of resolving it. If you cant argue and arrive at the best solution, youre not doing the work to actually have a real, healthy working relationship.

You have to actually lean into the conflict and come out with a solution that makes sense, over and over again. If you find yourself avoiding it, then you have to consciously expend effort to fight that default behavior.

Dont agree on something? Dont leave the room until you have a resolution.

An hour is not enough? Cancel your weekend, go on a hike and figure it out.

In these situations, theres nothing more important than for you and your co-founders to do the work and come out of it stronger.

Too much conflict? Establish boundaries

Of course, fighting all the time is no good either. Its a recipe for a frayed relationship sooner or later. When founders are in a situation where they are fighting about everything all the time, it usually means that their individual roles are not well-defined enough. Two hacker founders refuse to give up ground over an architectural decision, product-oriented founders with similar skill sets fight over direction and so on.

Heres the best way to handle it: Make a list of all of the areas needed for your business. Then figure out who is best at each part, and assign one person to it. If someones better at sales, they should own that. Likewise for DevOps or any other specific task that is core to your business. That person is officially the owner of that thing. Everyone agrees to hear each other out when a decision comes up, but once the owner decides, all debate is over. Everyone moves on. You cant debate things forever, and co-founders need to be able to trust each other.

Embracing conflict, fightingfair

If this is your first company, this might be the first time youve had to make decisions at this stage. What does it actually mean to embrace conflict? What is fighting fair?

Embrace conflict instead of abandoning yourself. Some founders know what they want, and know whats right, but give up before the fight even starts. If this sounds like you, dont feel bad about it that was me too. Ive always valued harmony in my interactions with everyone I work with. But with time, and, again, sometimes the hard way, Ive learned you cant sacrifice what you know to be right in order to get to that harmony early. Youve got to fight. Dont swallow your words. If you have a point, make sure you are heard.

Its not aggression either. You shouldnt bulldog your way to a decision. The loudest in the room shouldnt necessarily and automatically be the one who wins. This is actually conflict avoidance of a different stripe one that doesnt give any space to any competing idea at all. You may be sure youre right, but in a fair and balanced conflict, theres no downside to listening first and letting the other side know you hear them.

Co-founders need to be able to trust each other.

Fighting fair is collaborative and data-based.One concrete thing before you start to work through conflict is to always remind yourselves: Youre on the same team.

Everyone in the room wants to win, and all of you want to make this company successful. With that, youre ready to go talk about the problem as a process, where different viewpoints are aired and evaluated directly. You fail at this only when you try to skip to the end, either by giving up before you begin (self-abandonment) or asserting youre right before anyone even gets to get a word in edgewise.

One concrete way to get more direct experience with this is whats called a T-Group, which is a technique developed for the Stanford GSBs Interpersonal Dynamics program to train people in precisely this kind of fighting fair. Nonprofit InnerSpace regularly hosts them, and many founders describe the experience as being extremely valuable.

Get help

Some of you reading this will have been through all of the exercises above, and more. For those of you who are at the end of your rope with your co-founders, I have one final piece of advice: Get help!

Talk to your most trusted friends, investors and mentors. Startups are crazy things, after all. Youre trying to do something nobody else has done, and it can feel very lonely, like youre the only one who has ever had this problem. Trust me, it helps to get outside of your head and talk through what youre seeing with other founders and friends.

Dont be afraid to bring in the pros. Be open to getting professional help, either individually (to help you respond to the ongoing conflict) or as a group (similar to how a marriage counselor can save a marriage). I cant recommend executive coaching enough for founders, especially when a company-killing conflict is on the line. You have employees and customers who depend on you to make the right call, and you owe it to them to make sure you do. Athletes have coaches and trainers who help them get to peak performance. Knowledge work can be just as demanding, and Ive seen many founders find their partnerships saved this way.

Co-founder disputes are the No. 1 early startup killer, but it doesnt have to be thatway

Co-founder disputes have historically been one of the top reasons startups fail at the earliest possible stage. Most that do fail do so because conflict (either too much or too little) is left unresolved for too long; with these tools, youll at least be a little more prepared against that possibility.

Embrace the conflict just the right amount and youll get through this, too.

Thanks to my executive coach, Cameron Yarbrough, for reading drafts of this.

Read more: