Image credit: Xbox
New here? Every two weeks, I look at a leadership decision in the news the way a diligence team would: what did the decision-makers actually know, and how did they know it? Issues 1 to 4, on Disney, Paramount, Mattel and Lululemon, are already out.
Key takeaways:
Xbox created its first chief operating officer role in July as part of a turnaround. The executive in the role left about three months later. Xbox gave no reason and named no successor.
The public record says a lot about the role. On paper, it overlapped with another executive’s job, left two of the biggest studios outside it, and sat under a CEO who had been a COO herself.
A newly created role is a bet with its own terms. Boards should check those terms before they check the person who takes them.
For the executive being offered the new role, the same questions apply, from the other side of the table.
The idea in this issue: the New Role Test. Before backing a brand-new job, write down three things: what it decides on its own, what’s been left out of it, and what the CEO is holding on to. If that’s hard to write down, the job isn’t ready, and whoever takes it is carrying that risk.
The announcement:
On October 23, Xbox CEO Asha Sharma sent a short memo. “Helen Chiang is leaving Microsoft today. I’m grateful for her years of service and wish her well.” The COO’s duties are being split among other leaders for now. “We’ll share more on COO leadership as plans are finalized.” Asked why, Xbox told The Verge it doesn’t “discuss individual personnel matters.”
Chiang spent 21 years at Microsoft. She worked on Xbox Live and later led Minecraft and its studio, Mojang. In July, Sharma named her Xbox’s first-ever chief operating officer. The job lasted about three months.
Most coverage focused on the departure. This issue doesn’t. I don’t know why she left, and the reasons may never be fully public. What we can read is the public record of how the role was built. That’s where the bet was, and it’s the part a board can actually check.
How the role came together:
Sharma took over in February, after Phil Spencer retired and president Sarah Bond left. She came from Microsoft’s AI product group and had not worked in the game industry before. Matt Booty was promoted to Chief Content Officer in the same announcement.
By July, Sharma was resetting the business. Her memo said Xbox’s operating margins were “three to 10 times lower than comparable businesses.” About 3,200 jobs would go over the next 12 months, half of them right away. That’s roughly a fifth of Xbox’s workforce. At least four studios would leave, and there would be fewer layers of managers.
And a new COO would own the results for the whole business. In Xbox’s words, “end-to-end P&L responsibility across content, hardware, platform, and services.” The aim was to get every part of Xbox working the same way. That’s the bet in one sentence: one leader, picked from inside, could get a big, struggling business pulling in one direction, while the CEO reset the strategy and the company cut a fifth of its people.
Where the job was unclear:
Run the New Role Test on the July announcement, and two questions show up before anyone started the job.
On paper, two people owned the same thing. The COO was responsible for results “across content.” Xbox already had a Chief Content Officer, promoted five months earlier. When two senior leaders appear to own the same thing, someone often has to sort it out later, sometimes in a hurry.
“The whole business” had exceptions. In the same announcement, Mojang and King, the studios behind Minecraft and Candy Crush, moved to report directly to the CEO. Mojang was the studio the new COO had led. The job was described as end to end, minus two of the company’s biggest franchises.
The leader’s old team was left open. A leader promoted from inside brings credibility from somewhere. For this COO, that was Minecraft, and that part of the company was reorganized in October. Whatever the reasons behind the exit, a new job should spell out what happens to a leader’s old team when the leader moves up.
The CEO who did the job:
Sharma was chief operating officer at Instacart from 2021 to 2024, through its IPO. CEOs who have run operations themselves often have strong views on how operations should run. That’s not a flaw. That experience is often why a CEO is brought in to fix a business.
It does mean a new COO under that kind of CEO needs to know, in writing, which decisions are actually theirs. When the CEO has done the job, the COO title can mean less until someone writes down what it covers.
The fair reading:
There is a strong case for what Sharma did. A business with margins that far behind its peers needs one person accountable for results, and a 21-year insider who knew the business was a sensible pick. Resets move fast, and some roles are built to be revised.
Since the departure, one outlet has reported a different account of the exit, based on a single source it could not confirm. Microsoft has not commented. This issue takes no view on why she left. Its subject is how the role was built. Xbox may name a new COO soon, or fold the job back into the CEO’s office, and either could be the right call.
The point is narrower. The terms of the role were visible in July. They were the part of the bet a board could check.
Five tests before creating a new number two:
The who-decides test. List the decisions the role makes alone. If the list overlaps another executive’s, settle who wins before the announcement.
The what’s-left-out test. A job described as “the whole business” that has exceptions is a different job. Name the exceptions and why each one exists.
The what-the-CEO-keeps test. Ask the CEO which decisions they will hand over and which they’ll keep. This matters most when the CEO has done this job before.
The how-we’ll-know test. Set goals for the job that still make sense after the shake-up around it, and a date to check them.
The plan-B test. Ask what happens if the job isn’t working in six months. If the honest answer is “it goes back to the CEO,” the role may not need to exist yet.
From the coaching side. A newly created role is an honor. It can also be one of the riskiest seats in the building, because no one has done that exact job there before. If you’re the one being offered it, ask these before you accept:
“Show me the decisions that are mine.” Not the title. The calls you can make without asking.
“Who gives something up when I arrive?” Someone usually does. Know who, and whether they agreed.
“What are you keeping, and why?” The answer tells you where the real job ends.
“What does success look like in six months, in numbers?” If it’s vague now, it will be judged later by feel.
“What happens to my old team?” Your old team is where your credibility comes from. If it’s about to be reorganized, you want to know before the press release, not after.
A good leader can still struggle in a badly built role. These questions don’t guarantee the job works. They make sure you’re betting on a job, not a hope.
The question to ask first:
Before you approve a new role, ask yourself: if we had to write down on one page what this job covers, could we do it today?
If you are weighing a CEO, a founder or a new senior role right now, the free Source Map at gannett.partners lets you check your own evidence in about ten minutes. Enter who told you what, and see how many of your sources are really the same source. It’s designed so what you enter stays in your browser. And if you’d like to talk a situation through, I’m glad to.
And if you’re the leader carrying the seat, Gannett.Partners also coaches founders and CEOs. A thinking partner to help you see yourself clearly. [gannett.partners/coaching]
Get Leadership, Underwritten by email.
New case studies on leadership decisions in the news, sent straight to your inbox. Sign up at leadershipunderwritten.com.
P.S. New: the Team Test, a free one-page worksheet based on Issue 4. [Get it here for free]
Based on public sources as of November 1, 2026. For general information only. Not legal, investment or employment advice. Views are the author’s. Companies named are discussed from public sources; no affiliation is implied.
Chris Gannett is the founder and CEO of Gannett.Partners, an independent leadership advisory firm based in Dallas and working with leaders globally. The firm gives investors and boards an independent answer on whether they have the right leader, and helps that leader grow. Investors, family offices and boards hire it for CEO Diligence, its flagship practice. Founders and CEOs hire it for Executive Coaching, built around goals they set. The two are kept apart: the firm does not evaluate a leader it is coaching or has coached in the past year.