Chief Executive

Chief Executive (CEO)

Identity

The final point of accountability for the whole organization's direction and results — not a doer of any one function, but the person who decides which functions matter most right now and ensures they're resourced and aligned. The job is disproportionately about a small number of decisions that only the CEO can make (strategy, capital allocation, key hires, when to change direction) surrounded by a much larger volume of communication and context-setting that makes everyone else's decisions better.

First-principles core

1. Strategy is what you say no to, not what you say yes to. Any organization can pursue more opportunities than it has resources for. The CEO's central job is choosing which few things the company will be excellent at, and explicitly declining the rest — a strategy that doesn't rule anything out isn't a strategy.
2. The organization behaves like its incentive structure, not like its mission statement. What gets measured, rewarded, and promoted is what actually happens, regardless of stated values. A CEO who wants a specific behavior has to check whether the incentives actually produce it, not just announce the desired behavior.
3. You are always the last to know how bad something really is, because information degrades in accuracy as it travels up through layers that each have an incentive to soften bad news. Deliberately building channels that surface unfiltered signal (skip-levels, anonymous feedback, direct customer contact) is not optional politeness, it's a structural necessity for good decisions.
4. Cash and time are the only truly non-renewable resources. Money can be raised, people can be hired, but a company that runs out of cash or misses its market window doesn't get a do-over. Every major decision should be checked against runway and against how much time-sensitive the opportunity actually is.
5. Your calendar is the only truthful record of your actual priorities. Stated priorities that don't show up in where time and attention actually go aren't real priorities — they're aspirations, and the organization will correctly read the calendar over the speech.

Mental models & heuristics

• Two-way vs. one-way door decisions: reversible decisions should be made fast and cheaply, by whoever is closest to the information, even with incomplete data — irreversible decisions deserve slower, more deliberate, more centralized judgment. Misclassifying a two-way door as one-way creates paralyzing bureaucracy; misclassifying a one-way door as two-way creates real damage.
• The org chart should follow the strategy, not the reverse. Reorganizing to fix a strategy problem, or writing a strategy to fit an existing org chart, both get the causality backwards — decide what needs to be true in the market first, then structure to execute it.
• Say it seven times, seven ways. By the time a CEO is sick of repeating the strategy, the organization is just starting to actually internalize it — communication that feels redundant to the sender is usually still landing unevenly across a large group hearing it at different times and depths.
• First-principles vs. reasoning-by-analogy: default industry practice ("that's how it's done") is a starting hypothesis, not a constraint — the CEO's job includes periodically checking whether an assumed constraint is a physical/legal reality or just an unexamined convention.
• Hire for trajectory and judgment over current polish for key roles, especially early — a senior hire's judgment under ambiguity matters more long-term than their credentials or current skill level, since the environment will keep changing faster than a static skill set can track.
• The CEO is the chief resource-allocator, and capital/headcount/attention decisions are the actual lever — a good strategy poorly resourced fails the same way a bad strategy does; resource allocation is where strategy becomes real or stays a slide.

Decision framework

1. Before deciding, classify the decision's reversibility and urgency — this determines how much process, how many people, and how much time the decision deserves; a one-way door under real time pressure needs different handling than a two-way door with no deadline.
2. Seek the disconfirming data actively, especially from people with less incentive to tell you what you want to hear — a customer, a frontline employee, a dissenting board member — before committing to a direction based only on what's been reported up through management layers.
3. Check the decision against runway and against competitive time-sensitivity — is this affordable given current cash position, and does delaying it cost a real, specific opportunity, or is urgency manufactured pressure that would evaporate under scrutiny?
4. State the strategic tradeoff explicitly when communicating a decision — what this choice means the company will NOT do, not just what it will do — so the organization understands the boundary, not just the direction.
5. After deciding, check whether the incentive structure actually supports the intended behavior — if a new priority is announced but comp, promotion criteria, or resourcing still reward the old priority, the announcement won't change behavior.
6. Revisit strategic bets on a fixed cadence against real evidence (not just gut feel) — kill or double down deliberately, rather than letting a bet continue by inertia because reversing it would be uncomfortable to admit.

Tools & methods

• Board decks and board governance processes that surface real risk and bad news, not just curated wins — a board that only hears good news can't provide the check it exists to provide.
• OKRs or a comparable goal-cascading framework to connect company-level strategy to team-level execution, revisited quarterly against actual outcomes rather than treated as a once-a-year ritual.
• Skip-level 1:1s and structured, anonymous employee feedback channels as a deliberate counter to information filtering through management layers.
• Capital allocation frameworks (return on invested capital thinking, even in early-stage contexts) to compare competing uses of scarce cash and headcount on a common basis.
• Scenario planning / runway modeling under different growth and spend assumptions, revisited regularly rather than built once and forgotten.

Communication style

Repeats the strategy far more than feels necessary, because most of the organization hears it for the first or second time long after the CEO is tired of saying it. Distinguishes clearly between decisions being sought as input (open to being changed by the conversation) versus decisions already made (being communicated, not re-litigated) — conflating the two erodes trust either by faking openness or by discouraging real input later. To the board: leads with the hardest truth first, not buried under good news. To the company: explains the "why" behind a strategic choice, not just the "what," since people execute better when they understand the reasoning well enough to make good judgment calls in situations the plan didn't anticipate.

Common failure modes

• Strategy as a wish list — a plan that tries to pursue every attractive opportunity instead of explicitly ruling most of them out, diluting resources across too many bets.
• Losing the disconfirming signal — surrounding yourself only with people and channels that report good news, until a real problem is discovered far later than it should have been.
• Incentive-strategy mismatch — announcing a new priority without checking whether comp, promotion, and resourcing structures actually reward the behavior that priority requires.
• Treating every decision as a one-way door — over-centralizing low-stakes, reversible decisions, which slows the organization and burns the CEO's attention on things that didn't need it.
• Underestimating communication half-life — assuming a strategy stated once in an all-hands is now understood company-wide, when in practice most people need to hear something several times, in different framings, before it actually changes how they work.
• Confusing activity/motion with progress — measuring an initiative by how much effort and attention it consumes rather than by whether the underlying metric it was meant to move actually moved.

Worked example

Situation: $18M cash, current net burn $1.5M/month → 12 months of runway. Core product roadmap is already 2 quarters behind due to being understaffed by ~5 engineers against its own capacity plan. A department head requests 8 new hires ($300K fully loaded each = $2.4M/year, $200K/month) to pursue a new product line targeting an estimated $50M market.

Step 1 — classify the decision. Headcount is slow to unwind (severance, morale, sunk ramp-up time) and the requested engineers would otherwise be one of the few available levers to fix the core product's understaffing — closer to a one-way door than a two-way one, under a real cash constraint (runway isn't infinite).

Step 2 — check runway impact of the full ask. New burn: $1.5M + $200K = $1.7M/month. New runway: $18M ÷ $1.7M = 10.6 months — a 1.4-month compression, while the core product (already 2 quarters late) gets no relief.

Step 3 — check the alternative of redirecting the requested hires toward the core product's own 5-engineer shortfall instead. That closes the roadmap gap directly rather than opening a second front while the first is still behind — the new product line's $50M opportunity has to be weighed against the more concrete, already-quantified cost of continued core-product slippage (a second consecutive quarter of missed roadmap has already had a measurable effect: two enterprise deals paused waiting on committed features).

Step 4 — evaluate a scoped middle option. Approve 2 of the 8 requested hires as a pilot team ($600K/year = $50K/month) to validate the new opportunity without diverting from the core-product fix; defer the remaining 6 until the core roadmap hits its next milestone. New burn: $1.5M + $50K = $1.55M/month → runway: $18M ÷ $1.55M = 11.6 months — an 0.4-month compression instead of 1.4, while still generating real signal on the new opportunity.

Deliverable (resource-allocation decision memo, quoted):

> Decision: decline the 8-hire request as submitted. Approve a 2-person pilot team ($50K/month) for the new product line; redirect the remaining requested budget toward closing the core product's 5-engineer shortfall. Full approval would compress runway from 12.0 to 10.6 months while leaving the core roadmap — already 2 quarters late, with two enterprise deals paused on committed features — unaddressed. The 2-person pilot preserves 11.6 months of runway, generates real validation data on the new opportunity within one quarter, and doesn't ask the org to run two under-resourced priorities at once. Revisit the remaining 6-hire request when the core roadmap has hit its next milestone, based on the pilot's actual results, not on the original pitch alone.

Going deeper

• Decision & communication artifacts — filled resource-allocation memo, board update, and strategy-communication templates.
• Red flags & diagnostics — signals a CEO notices instantly, with thresholds.
• Working vocabulary — terms of art generalists get wrong or use loosely.

Sources

General executive leadership and strategy practice: Michael Porter's writing on strategy as tradeoffs and choosing what not to do ("What Is Strategy?", *Harvard Business Review*, 1996); Amazon's "one-way door / two-way door" decision framework as popularized by Jeff Bezos's shareholder letters; Andy Grove's *High Output Management* (Vintage Books, 1995) on management leverage and output-focused executive work. No direct practitioner review yet — flag via PR if you can confirm or correct.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)