How founders should split roles as a company scales
Most co-founder blowups don't start with a fight. They start with a fuzzy sentence like "we'll figure out titles later." Then the company grows, the work changes and two people who used to do everything together are suddenly stepping on each other in every meeting, wondering when their partner became an obstacle. Co-founder conflict sits in the top three reasons startups fail, right beside running out of cash and building something nobody wants. That should get your attention. Cash and market problems are hard to control. Role problems are self-inflicted.
I built Ringba over years of slow, deliberate decisions, and the role decisions were among the hardest. Not because the answers were complicated, but because they involved egos. Including mine. This is the playbook I wish someone had handed me early.
The default split, and why it works until it doesn't
Most venture-backed startups launch with two or three co-founders, and investors have historically preferred it that way. Y Combinator leaned toward teams for years, though solo-founder acceptances have grown since the mid-2010s. The classic split is CEO plus CTO: the CEO owns fundraising, sales, and hiring, while the CTO owns product and engineering.
Here's the thing. That split works well up to about 10 to 20 employees, because at that size the org chart and the actual work are the same thing. Everyone reports to a founder. Every decision has an obvious owner. You don't need process because you can settle anything over lunch.
Then you cross roughly 30 to 50 people and the model breaks. Not gradually. Suddenly. You hire a real executive layer, some mix of engineering VPs, sales leaders, and marketing heads, and the founders' jobs flip from doing the work to managing the people who do it. This is where the CTO who loves writing code discovers the job now means sprint rituals, performance reviews, and budget meetings. Some founders make that leap. Many don't want to, and pretending otherwise stalls both the person and the company.
Study the splits that worked. At Airbnb, Brian Chesky took the external CEO role while Nathan Blecharczyk owned technical infrastructure and Joe Gebbia owned design, which meant three founders held three lanes with almost no overlap between them. At Stripe, Patrick Collison runs the company as CEO while John Collison handles business operations. Different companies, same principle: each founder holds territory the others don't touch.
The lesson isn't the titles. It's the clean edges.
Who should be CEO?
The founder best at the CEO's actual job: raising money, recruiting executives, telling the company's story, and making final calls when the team deadlocks. Not the founder with the original idea, the most equity, or the loudest voice. Idea credit and operating ability are different things, and confusing them gets expensive.
Simple. Not easy. The CEO question is where founder egos collide hardest. My honest opinion: if you can't settle it in one direct conversation, you have a bigger problem than titles, and you should fix it before you raise a dollar.
There's a sobering statistic worth sitting with. Studies of venture-backed companies suggest a large share of founding CEOs are gone from the role by IPO or later funding rounds. Boards replace them. Sometimes founders replace themselves, which takes more self-awareness than most people have. Either way, the CEO seat isn't a lifetime appointment, and treating it like one is how you end up in a boardroom fight instead of a planning meeting.
Pick the person who can do the job for the next 18 months. Revisit after that. Nobody picks forever.
Write it down before you scale
Everything above is theory until it's on paper. Two documents matter, and skipping either one costs you.
First, a written role charter. Some teams use RACI matrices, but I prefer the DRI model popularized at Apple, where every functional area gets exactly one directly responsible individual whose name sits next to that function on a single page. Product, sales, hiring, all of it gets one name each, and when a decision stalls you look at the chart, find the name, and that person decides. Role charters exist to kill decision-rights disputes, and decision-rights disputes are what "co-founder conflict" looks like day to day. Nobody fights about the mission. They fight about who gets to say no.
Second, vesting. The standard is four years with a one-year cliff, and skipping it is one of the most expensive early mistakes a founding team can make, because vesting is what protects the company when a founder's role shrinks or they walk in year two. Without it, a departed founder keeps a full stake while everyone remaining does the work. I've watched that scenario poison companies from a distance. There's no clean fix afterward.
Neither document takes more than a weekend. The charter is a one-page table, and the vesting terms are boilerplate your startup lawyer has written a hundred times. Role problems are self-inflicted, and these two pages are the prevention.
Cheap insurance, honestly.
Revisit titles every 12 to 18 months
This is the most commonly missed move, so I'll be blunt. Put a recurring calendar event on the books, every 12 to 18 months, where the founders ask whether the titles still match the jobs. The "CTO" who wrote the first codebase and the CTO a 100-person company needs are often two different jobs. One builds. The other manages managers and hasn't shipped code in a year.
That conversation only feels threatening if you've never had it. Founders who normalize it early treat title changes like reorgs, not demotions. The founding CTO becomes Chief Architect and does the deepest technical work in the company while a hired VP of Engineering runs the org, and everyone wins. Founders who dodge the conversation get the same outcome eventually, just delivered by a board. With more bruises.
If it's too loaded to have alone, bring in an executive coach or facilitator at $200 to $600 an hour. Steep, sure, until you price the alternative, because legal fees for a founder breakup routinely run $25,000 to well over $100,000 before you even count the distraction. A few facilitated sessions are a rounding error.
I write more about these operating decisions at adamyoung.com, and I covered how role clarity shaped Ringba's growth in The Pay Per Call Revolution. The pattern repeats at every stage. Clarity is boring, and boring compounds.
FAQ
What if both founders want to be CEO? Pick based on the job description, not seniority or idea credit. Whoever is better at fundraising, recruiting, and external storytelling takes the seat for 18 months, with a scheduled review. Truly stuck? Hire a facilitator before you hire anyone else.
Should co-founders always have equal equity? Not necessarily. Equal splits are common and defensible, but unequal ones work fine when the reasoning is documented and everyone signs off early. The ratio matters less than vesting: four years, one-year cliff, no exceptions. Including you.
Can a founder step back without leaving? Yes, and it's healthier than most people assume. A founding CTO can become Chief Architect or Chief Scientist while a hired VP runs the department, which keeps the founder's board influence and deep work intact while the org finally gets the management it needs. Write the new charter the same week.
When is it too late to add vesting? Never, but it gets harder every month. Retroactive vesting needs every founder's consent, and consent gets pricey once someone's already thinking about leaving. No vesting in place? Call your lawyer this week.