LinkedIn Post Ideas for Chiefs of Staff

"So what do you actually do?" I've been asked this at every dinner party for two years, and honestly, my CEO has asked a version of it too, mid-quarter, when she wondered aloud why a task had landed on my desk instead of someone else's. Here's the honest answer. My week falls into three buckets. Bucket one, roughly 40%: proxy work. Representing the CEO's intent in rooms she can't be in — a vendor negotiation, a cross-functional dispute between two VPs, a board member follow-up. I'm not deciding for her. I'm carrying her actual position accurately into places she isn't. Bucket two, roughly 35%: cadence design. Board prep timelines, the weekly leadership sync agenda, the quarterly planning process. Nobody else owns making sure the company's decision rhythm actually works, so it defaults to me. Bucket three, roughly 25%: the fire du jour. A leadership departure that needs careful internal communication. A customer escalation that's become political. This bucket is unpredictable by design — it's whatever the company most needs someone senior and unattached-to-a-department to absorb that week. The ambiguity isn't a bug in the role. It's the entire value proposition — a chief of staff who has a fixed job description has usually been narrowed into something closer to an EA or a project manager. If your CEO still isn't sure what you do, ask her which bucket she thinks is missing. That conversation is usually more useful than a job description.

Ten days out from a board meeting, here's the actual timeline I run. No step has ever been skipped without a visible cost in the room. Day 10: First deck draft assembled from function leads' inputs, reviewed by me for narrative coherence before the CEO sees it. I cut anything that doesn't serve the three decisions we need from the board this cycle. Day 8: CEO review one. She reads it cold, flags what's missing or wrong. This is deliberately her first real look — I don't want her opinion baked into the structure before she's reacted to it fresh. Day 6: Pre-wire calls begin. I personally call two board members known to ask the hardest questions, walk them through the key numbers informally, and surface objections before they're voiced live. The CEO should never be surprised in the room by a question I could have seen coming. Day 4: Full rehearsal, out loud, timed. The CEO presents to me and two function leads as if we're the board. We interrupt with the sharpest questions we can think of. Day 2: Final deck lock. Nothing changes after this point except typos — last-minute edits are where errors sneak in. Day 0: I sit in the room, silent, tracking action items live so the follow-up doc goes out within 24 hours, not a week later. Board prep is the highest-stakes recurring task in this role. Almost nobody documents the actual mechanics publicly. This is mine.

I ran a full calendar audit on my CEO three months into the role. One recurring meeting was costing her six hours a week, and it wasn't even the biggest line item — it was the invisible one. The audit method: I logged every meeting for four weeks, tagged each by type — decision-required, informational, relationship-maintenance — and tracked whether she spoke more than she listened, as a rough proxy for whether her presence was actually load-bearing. The meeting that died: a daily 30-minute cross-team standup she'd been attending since the company was 15 people, now attended by 22. She spoke for maybe two minutes most days. Everyone else was performing their update for her benefit, not coordinating with each other. I proposed cutting her attendance entirely and replacing her presence with a written summary I'd compile from the standup notes, delivered to her in five minutes over coffee instead of thirty minutes live. What filled the six hours: two of them went to the pre-wire calls I'd been doing badly, squeezed into gaps. Two went to a genuinely new initiative — direct 1:1 time with the four function leads she'd been meaning to spend more time with for a quarter. Two went back to her, unscheduled, on her own request. Nobody in the standup noticed her absence within two weeks. The team had been performing for an audience that was barely watching. Calendar surgery is the most invisible leverage in this job, and the easiest to prove with a stopwatch.

Chief of staff is not a career. It's a launchpad, and I think pretending otherwise does the role a disservice. I say this as someone two years into the seat with an explicit plan: I'm doing this for roughly two more years, then moving into an operating role — likely VP of Ops or a general management seat — where I own a P&L or a function directly instead of amplifying someone else's. My logic: the CoS role teaches an extraordinary range — you see the whole business, sit in every important room, learn to translate strategy into execution across functions you don't own. What it doesn't teach, past a certain point, is direct accountability for an outcome that's unambiguously yours. Every win in this seat is technically the CEO's or the team's. That's correct and appropriate, and also exactly why staying past the learning curve starts to cost you. I know career CoS people who've done this for eight, ten years at different companies, and some are genuinely happy with a role built around proximity to power rather than ownership of an outcome. That's a legitimate choice. It's just not mine, and I think being explicit about the two-year plan, with my CEO, from day one, made me better at the job — not worse — because I wasn't trying to make the role permanent by making myself indispensable in ways that actually slow the organization down. Career CoS folks, I expect disagreement. What's your version of the plan?

I had to tell the executive team we'd missed a board-committed revenue target by 14%, on behalf of a CEO who was traveling and asked me to run the meeting. It remains the hardest room I've been in. What I got wrong: I opened with context and caveats, trying to soften the number before I said it. By the time I said 14%, three people in the room had already guessed something worse from my tone, and I'd spent two minutes building anxiety instead of clarity. What I'd redo: lead with the number, cold, in the first sentence. Then context. People can handle a hard fact faster than they can handle watching someone circle one. What landed badly: I used language that sounded like I was defending the number — "given the market conditions" — when the room wanted ownership language, even though the miss wasn't mine to own personally. I should have said "here's what happened and here's what we're doing," full stop, without any framing that read as an excuse on the CEO's behalf. What worked: I'd prepared three concrete next steps before walking in, not just the bad news. The room's mood shifted the moment they saw there was already a plan, even an imperfect one. Being the messenger on a missed quarter taught me that the news itself rarely breaks trust. How it's delivered — clearly, without defensiveness, with a next step already in hand — is what people actually remember.

My weekly sync with the CEO has run roughly 80 times now. The template has changed six times. Here's where it landed. Section one: Decisions needed. Every item requires a specific ask and a deadline — never "thoughts on this?" Vague items get sent back to the owner to sharpen before they make the doc. Section two: Radar. Things that aren't urgent yet but could become so — a customer showing early churn signals, a competitor move, a leadership team dynamic worth watching. This section has caught more real problems early than any other part of the doc, precisely because it has no bar for urgency to get listed. Section three: Blocked threads. Anything stuck waiting on a decision, a person, or information, more than five days. Naming it here has fixed more bottlenecks than any status meeting, because visibility alone often unsticks things. Section four: One question I ask her directly, every week, about her own bandwidth or state: what's taking more energy than it should right now? This section is newest, added after I realized I was tracking the company's health closely and her own far too loosely. What I cut along the way: a general updates section that became a dumping ground nobody read closely, and a metrics recap that duplicated the ops review. Every CoS is quietly improvising a version of this. Compare notes — what's in yours that isn't in mine?

Five moves that actually work for influence without authority, the core skill of this seat, learned mostly by getting them wrong first. 1. Pre-wire every contested decision. Never let two VPs discover they disagree for the first time in a room with the CEO present. I talk to both beforehand, separately, so the room surfaces a resolved disagreement or a clearly framed choice, not a live argument. 2. Borrow sponsorship deliberately. When I need a cross-functional initiative to move, I ask the CEO for one specific sentence I can quote directly in the room — not a vague "the CEO supports this," but her actual words. Borrowed authority only works when it's precise enough to be checked. 3. Control the agenda, not the outcome. Being the person who sets what's discussed and in what order shapes decisions more than being the person with a vote, most of the time. 4. Write the first draft. Of the memo, the plan, the announcement. The first draft anchors every conversation that follows it, even when it gets heavily edited. 5. Never spend borrowed capital on something small. I save the CEO's name and the pre-wired conversations for decisions that actually matter, so they carry weight when I use them. None of these are manipulation. They're mechanics. Practical politics for people who live it daily, not people who theorize about it.

Behind the scenes of the annual planning week I run, the machinery nobody outside leadership ever sees. Six weeks before: I collect a one-page pre-read from every function lead — their top three priorities, their honest resourcing gaps, one disagreement they expect to have with another function. That last question surfaces more useful friction than any planning template I've tried. Three weeks before: I stage two disagreements deliberately. Not manufactured conflict — real tensions I already know exist between functions, scheduled as explicit debate sessions before the offsite, so they don't derail the actual planning week when time is scarcest. Offsite week: the schedule is built so the hardest cross-functional tradeoff happens on day two, mid-morning, when energy is highest and nobody's checked out yet. Day one is alignment on context. Day three is commitments and owners, never new debate. What gets decided in the room versus deferred: resourcing tradeoffs and the top five priorities get decided live. Anything requiring data we don't have yet gets explicitly deferred with an owner and a date — never left ambiguous, which is how planning weeks quietly produce decisions nobody actually made. The part I'm proudest of: we now track which staged disagreements from the pre-work actually surfaced in the room versus resolved themselves beforehand. Last cycle, four of six resolved in side conversations before the offsite even started, which meant offsite time went to the ones that genuinely needed it. Planning season makes this the most searched content in the CoS community for good reason.

I used to argue, confidently, that chiefs of staff should never have direct reports. I've changed my mind, and I want to walk through why. My old position: the CoS role derives its power from having no organizational turf to defend. The moment you manage people, you start optimizing for your own team's success instead of the company's, and you lose the neutrality that makes cross-functional influence possible. What changed it: I took on two direct reports last year — a junior CoS and an executive assistant to the leadership team — not by choice initially, but because the workload demanded it. I expected it to compromise my neutrality. It mostly didn't. What it did do was force me to build actual management muscle I'd been avoiding, and it gave me leverage to delegate the tactical 30% of my job so I could spend more time on the strategic 70%. Where my old position still holds: I'd never take a direct report from a function I'm meant to be neutral referee over — no engineers, no salespeople reporting to me directly. The reports I have are explicitly CoS-function support, not operating-team ownership. So: nuanced yes, with a narrow definition of what kind of direct report. I expect disagreement from both camps — the purists who think any direct report dilutes the role, and the career CoS folks who've run larger teams successfully. Convince me further.

My CEO once handed me a first draft of a difficult investor email, said "you know my voice better than I do right now," and walked into another meeting. That one still surprises me when I think about it. Chiefs of staff, I want your version. What did your CEO delegate to you that you didn't expect — the moment that told you something real about how much trust, or how much sheer capacity constraint, was actually driving the decision? Mine, in more detail: it was the week of a difficult board update after a rough quarter, and she was double-booked with an investor call. She didn't just ask me to draft it — she asked me to send the first version directly to two board members for early reaction, under my own name, representing her thinking. It worked. But it also meant if I'd gotten her position wrong, the damage would have happened before she ever saw it. The range of answers I expect from other CoS folks: fundraising deck ownership, a firing conversation run solo, product launch decisions made in the CEO's absence. The variance is the whole point of this role — no other job at this level swings this wide week to week. Answer below. I think this comment section ends up being one of the better recruiting pitches for the function that exists on this platform — the range on display here is what makes people want the seat.