CEO vs COO: how to split the top job

CEO vs COO at a glance

The shortest version comes from venture firm CRV: "CEO owns the plan; COO owns making it happen." The US Bureau of Labor Statistics says chief executives "provide overall direction", and that in large organizations they focus on policy and strategy while operations managers direct the daily work. The table puts those definitions into startup terms.

The pay figures come from Payscale's CEO and COO salary pages, both updated in October 2026 and listed in the sources below, and they cover every size of organization. Total pay runs from $86k to $430k for CEOs and $85k to $299k for COOs. The BLS puts the May 2025 median wage for chief executives at $213,990, and for general and operations managers at $105,770. A startup CEO who is also a founder often pays themselves well below any of these numbers, so treat them as a ceiling on what the market expects, not a target.

What does a COO do that a CEO doesn't?

A COO turns the CEO's strategy into a working operation. CRV calls the COO the second in command who "owns day-to-day operations". The defining feature, according to Origin Ventures partner Jason Heltzer in a Built In article, is "the breadth of responsibilities". A VP of operations or growth runs one lane. A COO runs several, and usually carries some P&L weight.

That breadth means the job looks different in every company. CRV sorts candidates into four types, and naming which one you need is the most useful thing you can do before writing a job description.

• Operator. Runs a model that already works and makes it run faster and cheaper. Fits a company with repeatable revenue and growing headcount.
• Builder. Turns improvised work into repeatable systems. Fits a company where everything still runs on the founders' memory.
• Complement. Covers what the CEO is weak at. A sales-minded CEO who hates internal process often wants this one.
• Deputy. Co-runs the company with real authority. This is the closest to a true number two, and the hardest to make work.

Built In also describes a strategic-change COO, brought in for a pivot or a move into new markets. Matt Bieber, a COO quoted in the same article, calls that hire "a proactive maneuver". Whichever type you pick, write it down. A COO hired as a Builder and judged as an Operator will look like a failure within a year.

Who to hire first at each stage

The CEO always comes first, so the real question is when, if ever, to add the COO seat. No rule fits every company. These stages are our reading of the sources, and the headcounts are rough.

1. Before product-market fit. No COO. CRV warns that a COO hired before the model works can "install structure too early" and slow down the trial and error you still need. Heltzer puts it plainly: "In a seed deal, you normally do not see a COO." The founders split operations between them.
2. Around 10 to 12 people. CRV says this is when a dedicated operations hire usually makes sense. That's an operations manager or a generalist, not an executive. If the founders are drowning in follow-up rather than operations, a chief of staff may fit better.
3. Series A and early scaling. Heltzer estimates a Series A company has a COO "maybe 50 percent of the time". CRV suggests a fractional COO is usually right until revenue and team complexity are substantial.
4. Complex operations at any size. Companies that mix hardware, logistics or production with software sometimes need one early. Built In cites the meal-kit startup Tovala, which hired a COO "really early" because the CEO couldn't track every detail. Keith Rabois joined Square as its first COO at about 20 employees.
5. Growth stage with an outward CEO. When the CEO spends most of the week on fundraising, press and key customers, a COO who runs the inside of the company is often the right call.

Mark Suster, quoted by Built In, takes the opposite view for early companies: the CEO should run the business and delegate to VPs, which makes a COO unnecessary. That's a reasonable default. Hire against it only when you can name the specific operating problem the COO will own.

What a COO costs a startup

CRV's August 2026 guide gives directional figures for seed through Series B: a base salary of roughly $150K to $250K, total compensation of $200K to $400K, and equity of about 2% to 5% at seed, narrowing to 1% to 2% by Series A, usually over four years with a one-year cliff. It quotes fractional retainers of about $8,400 to $11,300 a month for around 14 hours a week. CRV is an investor rather than a salary survey, so use these as a starting point for conversations with recruiters.

Here's the arithmetic for a 25-person company. A fractional COO at $8,400 a month costs $100,800 a year in cash and no equity. A full-time COO at the bottom of CRV's base range costs $150,000 plus benefits and a meaningful slice of the company. The fractional option buys you roughly a third of a working week. If the problem needs someone in every leadership meeting and every hiring decision, a third of a week won't solve it, and the cheaper option ends up costing more.

Write the decision rights before the offer

CRV makes a point that matches what goes wrong most often: CEO and COO partnerships strain more over unclear authority than over skill. Its advice is to agree how disagreements get resolved and to write a mandate with explicit decision rights before the offer goes out. A one-page memo covering these questions is enough.

• Who can hire and fire department heads, and who has to be consulted.
• Who approves the annual budget, and who can move money between teams during the year.
• Who runs the weekly leadership meeting and sets its agenda.
• Who talks to the board about operations, and whether the COO attends board meetings.
• Which decisions the CEO keeps no matter what, such as pricing or the next fundraise.
• How a disagreement between the two gets settled, and how quickly.

If you can't fill in the memo, you aren't ready to hire. The blank lines show which parts of the business you still want to hold on to.

Signs you chose wrong

• Every decision still comes back to the CEO. Department heads go around the COO because they know the real answer comes from the founder. The title exists, the authority doesn't.
• The COO is pitching their own strategy. CRV lists a competing ego as a warning sign in candidates. After the hire, it shows up as a second vision competing with yours in leadership meetings.
• Process arrived before the product worked. Weekly reports, approval chains and planning cycles appear while the company is still searching for a repeatable sale. That's the early-structure risk CRV describes.
• The job turned out to be one lane. If the COO spends most of the week on one function, Heltzer's point applies: a VP of that function would have been cheaper and a better fit.
• The hire was meant to fix traction. A COO makes a working engine run better. If growth has stalled because the product isn't landing, a second executive won't change that.

Software can make either seat lighter, but it doesn't change who decides. In Nautis, the Chief of Staff builds a briefing at 7:00 am from tasks, pipeline, runway and approvals, so a CEO without a COO sees the operating picture in one place. The Decision Loop is a simple log for big calls with your reasoning and confidence, which you revisit after 30, 60 or 90 days. That's a useful record when two executives disagree about what was agreed. Neither replaces a COO's authority over teams. For what the top role looks like stage by stage, read the startup CEO guide and the startup COO guide.

FAQ

Sources

• CRV: how to hire a COO, timing, fit and what to look for
• Built In: what kind of startup needs a chief operating officer?
• US Bureau of Labor Statistics: top executives
• Payscale: chief executive officer salary, United States
• Payscale: chief operating officer salary, United States