Oracle, Spotify and Comcast Put 6 Executives in 3 Co-CEO Jobs
One CEO is increasingly a polite fiction. Oracle, Spotify and Comcast have put six executives into three top jobs because the modern chief executive role has become too big for one person to bluff through.
One CEO is increasingly a polite fiction. Oracle, Spotify and Comcast have put six executives into three top jobs because the modern chief executive role has become too big for one person to bluff through.
That does not mean co-CEOs are clever by default. Most businesses can barely get one boss to make a decision on time. Give the wrong two people the same title and you have not doubled leadership capacity — you have installed a very expensive traffic jam.
The six-executive signal hiding in plain sight
Axios flagged the pattern this week: Spotify, Comcast and Oracle have each named two co-CEOs, while the seven Fortune 100 companies that completed planned CEO transitions this year chose insiders. That is the bit worth paying attention to. Big companies are not handing the keys to heroic outsiders and praying for magic. They are splitting the workload among people who already understand the engine.
At Oracle, the change was unusually explicit. On September 22, 2025, Safra Catz moved from CEO to executive vice chair after running the company since 2014. Oracle elevated Clay Magouyrk, previously president of Oracle Cloud Infrastructure, and Mike Sicilia, previously president of Oracle Industries, to co-CEOs. Larry Ellison stayed chairman and chief technology officer.
That is not a random pairing. Magouyrk came up through cloud infrastructure and had led Oracle’s cloud engineering work. Sicilia came from the applications and industry side, including Oracle Health. One man is built around the machinery underneath the AI boom; the other is built around getting industry-specific software into customers’ hands. Infrastructure and applications. Build it and sell it. Clean enough.
Spotify made a similarly deliberate change. Founder Daniel Ek became executive chairman on January 1, 2026, while Gustav Söderström and Alex Norström became co-CEOs. Söderström had been co-president and chief product and technology officer. Norström had been co-president and chief business officer. Spotify said the pair had already been leading much of the company’s strategy and operations since 2023.
Again, the logic is obvious. Product and technology on one side; commercial execution on the other. Ek remains involved in long-term direction and capital allocation rather than pretending he can — or should — remain the bottleneck for every operating decision.
Comcast followed with Michael Cavanagh becoming co-CEO alongside chairman Brian Roberts in January 2026. Cavanagh was not imported with a glossy turnaround deck. He was already inside the company. That matters more than people admit. When the business is large, complicated and under pressure from multiple fronts, a new leader needs years to learn where the bodies are buried. An insider starts with the map.
The old CEO job has quietly become four jobs
The classic CEO role was already absurd. You were meant to set strategy, allocate capital, hire and fire senior people, keep the board calm, sell investors on the future, talk to customers, manage regulators, make acquisitions and stop the culture turning into a bin fire.
Now bolt on AI. Not merely buying a few software licences and calling it transformation. Real AI decisions reach into product, pricing, data, cybersecurity, workforce design, customer trust, legal risk and capital expenditure. A CEO who does not understand the strategic implications is exposed. A CEO who tries to personally master every technical and commercial detail is kidding themselves.
That is why the Oracle model makes sense on paper. Oracle’s cloud infrastructure opportunity and its industry-applications opportunity are interconnected, but they are not the same operating problem. One demands relentless capacity, reliability, engineering and giant customer commitments. The other demands domain expertise, product adoption and credible outcomes in regulated industries. Putting a single generalist above both can work — until it does not.
Spotify’s arrangement is even more revealing because it formalised something already working. The company did not wake up one morning and decide two CEOs sounded trendy. It had operated with Söderström and Norström as co-presidents for nearly three years before granting the title. That is the sensible sequence: test the operating relationship first, then change the org chart. Not the other way round.
Too many boards do the reverse. They announce a structure, put out a press release about complementary skills, then discover six months later that “complementary” was a nicer word for “they cannot agree on anything.”
Continuity is winning because disruption is expensive
The fashionable boardroom fantasy is the outsider saviour: bring in a new face, slash a few layers, say “transformation” 40 times, unveil a new font and hope the share price gets excited.
Sometimes that is necessary. If the business is genuinely broken, continuity can be cowardice dressed as stability.
But that is not what these companies are doing. Oracle, Spotify and Comcast have all elevated leaders already steeped in their businesses. Axios noted that all seven Fortune 100 companies with planned CEO transitions so far in 2026 selected internal candidates. That is a very clear boardroom verdict: when the operating environment is moving at a stupid pace, institutional knowledge has become more valuable.
There is a practical reason. The new CEO no longer gets a leisurely 100-day honeymoon. The market wants answers immediately. Employees want clarity immediately. Customers want reassurance immediately. Activists, analysts and competitors are already sharpening knives before the appointment hits the wire.
An insider knows the people, the constraints, the customer promises and the ugly trade-offs already made. They can start making decisions on day one rather than spending six months being briefed by people trying to protect their turf.
That is not glamorous. It is just good operating hygiene.
The contrarian bit: two CEOs are often a warning, not a solution
Here is the part most leadership articles skip because everyone wants to sound supportive: co-CEO structures are dangerous.
They work only when the company has already answered the question everyone else will avoid: who decides when the two bosses disagree?
If the answer is “we will collaborate,” you do not have a governance model. You have a future argument.
The co-CEO title can mask three very different realities. It can be a genuine division of labour, like product versus commercial execution. It can be succession training, where an incumbent shares power before stepping away. Or it can be a board ducking a hard choice between two candidates.
Only the first two are promising. The third is poison.
A business does not need two people owning every decision. It needs decisions split cleanly, shared decisions kept rare, and a tie-breaker that everyone understands before pressure arrives. Oracle’s pairing looks designed around two major operating domains. Spotify’s pairing was rehearsed for years. Those are better signs than a board merely announcing that two talented executives will “work closely together.” Of course they will. That is not the question.
The real question is whether the executive team knows whose call it is on product, pricing, hiring, capital, acquisitions, customer escalation and public messaging. If they do not, staff will shop decisions between bosses. Politics will flourish. Speed will die.
And no amount of executive coaching fixes an org chart that gives two people overlapping authority with no rules.
Coca-Cola shows the alternative: keep one CEO, split the load below
Co-CEOs are not the only answer. Coca-Cola offers a cleaner alternative for plenty of companies.
Henrique Braun became CEO on March 31, 2026, succeeding James Quincey, who stayed on as executive chairman. Coca-Cola also created a chief digital officer role, appointing Sedef Salingan Sahin to unify digital, data and operational excellence. Sahin reports to Braun; the company did not create a second CEO to prove digital matters.
That is an underrated move. It acknowledges that digital execution is now too important to be buried inside a vague transformation committee, while preserving a single point of accountability at the top.
For many founders and mid-market operators, this is the better lesson. You probably do not need a co-CEO. You may need a real operating leader with authority over the part of the business you keep treating as a side project — technology, product, sales execution, supply chain or customer retention.
The title matters less than the mandate. Give someone responsibility without decision rights, budget or direct access to the CEO and you have just hired another person to prepare slides.
What this means for you
Do not copy the co-CEO label because Oracle or Spotify did it. Copy the discipline underneath it.
First, write down the five decisions that currently clog your business. Not the vague ones. The real ones: who owns pricing, who can hire senior people, who approves product roadmaps, who speaks to major customers, who gets to kill a failing project.
Second, assign one final decision-maker to each. One. You can have input from ten people. You cannot have ten owners. Shared accountability is usually unaccountability wearing a nice shirt.
Third, if you are considering a second-in-command or a co-leadership structure, run a 90-day test before changing titles. Give each person a defined domain, a budget, measurable outcomes and a weekly decision log. Watch where decisions stall, where people bypass the structure and where the two leaders genuinely make each other better.
Fourth, make the tie-breaker explicit. It might be the chair, the founder, the board, or one CEO having final say in a specified category. But write it down before the first serious disagreement. If you wait until the fight, the fight becomes about power rather than the decision.
The lesson from Oracle, Spotify, Comcast and Coca-Cola is not that one CEO is obsolete. It is that the lone genius CEO has always been a bit of a fairy tale.
The best leaders are not the ones who insist on carrying every problem themselves. They are the ones who build a structure where the right decisions get made quickly, by people who actually own the consequences. That is how you scale without turning the whole place into a meeting about meetings.