Twelve essays on what leaders actually do — when the textbooks fail, the calendars fill, and the team is watching.
A primer · Volume VII, No. 9 · The Deck Catalog · 14 pages
In this issue
The Drucker Question — What does a leader actually do?
Style — Goleman's six modes
Situational Leadership — Hersey & Blanchard
Servant Leadership — the Greenleaf reversal
Level 5 — Collins on the quiet ones
Decision-making — RAPID, DACI, OODA
Communication — the 7C framework
Feedback — radical candor and its limits
Hiring — the most leveraged decision
Case · Satya Nadella's Microsoft turnaround
Case · Boeing's slow drift
Reading & viewing
01 · The Drucker Question
What does a leader actually do?
Peter Drucker spent six decades posing the same disarming question to executives: "What is it you are trying to accomplish?" Most could not answer in a sentence.
Drucker held that the leader's job reduces to three things, in this order. First, define the mission — what business are we in, what success looks like, what we will not do. Second, recruit the right people and put them in roles that match strengths. Third, hold people accountable to outcomes, not activity.
Everything else — strategy, culture, execution, communication — sits downstream of those three. A leader who hasn't done them is, in Drucker's word, an "executive": someone who responds to the inbox.
"Management is doing things right; leadership is doing the right things." — Peter F. Drucker, The Effective Executive, 1967
02 · Style
Goleman's six modes
Daniel Goleman, building on the Hay/McBer database of 3,871 executives, identified six distinct leadership styles in the 2000 HBR essay "Leadership That Gets Results." The best leaders fluidly switch among them.
Style
Modus operandi
Best when
Risk
Coercive
"Do what I tell you."
Crisis, turnaround, with problem employees
Erodes flexibility, motivation, ownership
Authoritative
"Come with me."
Vision needed; new direction
Fails when team has more expertise than leader
Affiliative
"People come first."
Healing rifts, motivating in stress
Poor performance can be tolerated
Democratic
"What do you think?"
Building buy-in, gaining input from capable team
Endless meetings, no decisions
Pacesetting
"Do as I do, now."
Quick results from highly motivated experts
Burns out team; leader becomes bottleneck
Coaching
"Try this."
Long-term capability building
Slow; doesn't work if employee resists growth
Goleman's data: leaders who used four or more styles created the best climate and business performance. Pacesetting, used habitually, was negatively correlated with team performance — counter-intuitive but consistent in the data.
03 · Situational Leadership
Match the style to the moment
Paul Hersey and Ken Blanchard's 1969 framework argues there is no single best leadership style. The right approach depends on the readiness of the follower for a specific task. Plot followers on two axes: willing/unwilling and able/unable.
The four resulting quadrants demand four leadership behaviors: Telling (high directive, low support), Selling (high directive, high support), Participating (low directive, high support), and Delegating (low directive, low support).
The same employee may need Telling on a brand-new domain and Delegating on her core expertise. A leader who applies the same style across all situations gets sub-optimal results in most of them.
The model has been criticized as oversimplified and under-validated empirically. But its diagnostic discipline — assess the person and the task, then choose — remains the field's best on-the-ground advice.
Figure 1. Hersey & Blanchard, Situational Leadership Model. Match style to follower readiness.
04 · Servant
The Greenleaf reversal
Robert Greenleaf, a 38-year AT&T executive, retired in 1964 and wrote an essay that inverted the leader-follower hierarchy. The leader's first job, he argued, is to serve the team — to remove obstacles, develop people, and enable the work, not the other way around.
Servant leadership is now embedded in agile methodology, in much of the U.S. military officer training, and in some of the most durable corporate cultures (Southwest, Costco, Marriott).
Ten characteristics in Greenleaf's framework: listening, empathy, healing, awareness, persuasion, conceptualization, foresight, stewardship, commitment to growth, building community.
"The first test, and the most difficult to administer, is: do those served grow as persons?"— Robert Greenleaf
05 · Level 5
Jim Collins's quiet ones
For Good to Great (2001), Jim Collins's team examined every Fortune 500 company over 35 years and isolated 11 that achieved sustained outperformance after a key leadership transition. Each was led by what Collins termed a Level 5 leader.
The defining trait: professional will + personal humility. Level 5 leaders were almost universally introverted. They credited the team for success and looked in the mirror for failures. They built institutions that thrived after they left.
"Level 5 leaders look out the window to attribute success — to factors other than themselves. When things go poorly, they look in the mirror." — Jim Collins, Good to Great
Counter-finding · the celebrity CEO
Collins also noted that 10 of 11 named CEOs of the comparison group (companies that didn't make the leap) were prominent industry celebrities. Big names did not correlate with sustained outperformance. The opposite, often.
06 · Decision-Making
Three frameworks for making the call
RAPID (Bain)
Recommend · Agree · Perform · Input · Decide. Each role assigned to specific people for each decision. Removes ambiguity about who decides what.
DACI (Atlassian)
Driver · Approver · Contributor · Informed. The simpler cousin of RAPID, used in product orgs.
OODA (Boyd)
Observe · Orient · Decide · Act. John Boyd's fighter-pilot framework. The leader who cycles faster than the opponent wins. Originally tactical; now applied broadly to crisis management.
Figure 2. OODA loop. The decisive variable is cycle speed, not loop quality.
07 · Communication
The 7 Cs of clear messaging
The communication checklist used in journalism, in plain-language law, and in operational briefings. Drop one and the message degrades.
Clear — single interpretable meaning. Avoid jargon, hedges.
Concise — no unnecessary words. The reader's time is the budget.
Concrete — specifics, numbers, names. "Some progress" vs. "shipped 3 of 5."
Correct — factually accurate. Trust is a thousand small confirmations.
Coherent — logical sequence. Conclusion first, then support.
Complete — answers the question being asked. No critical omissions.
Courteous — respects the recipient. Not synonymous with bland.
"If you can't explain it simply, you don't understand it well enough." — attributed to Albert Einstein
08 · Feedback
Radical candor and its limits
Kim Scott's framework, born from her time at Google and Apple, plots feedback on two axes: care personally and challenge directly. Both high → radical candor. High care, low challenge → ruinous empathy. Low care, high challenge → obnoxious aggression. Low both → manipulative insincerity.
The strength of the framework is that it treats truth-telling and warmth as additive, not opposed. The most-cited corporate failure mode is "ruinous empathy" — managers withholding hard feedback to preserve the relationship and so denying their reports the information needed to grow.
Critique: candor without context can be cruelty. Cultural variance matters; a level of directness that energizes a Dutch team can wound an American or a Japanese one. The skill is calibration.
Two practical rules: feedback should be specific (what action, when, what effect) and timely (within 24 hours of the observed behavior). The annual review is, mathematically, the worst possible delivery vehicle.
SBI: a feedback grammar
Situation — when and where it happened. Behavior — what specifically the person did or said. Impact — the effect on you or the team. Skip the labeling ("you're disorganized"); describe the act.
09 · Hiring
The most leveraged leadership decision
Andy Grove called hiring the manager's most consequential and most under-prepared act. A bad hire takes 9 months to identify, 6 more to exit, and costs ~1.5× annual salary in direct and indirect terms.
Best practices that mostly hold up across studies:
Structured interviews. Every candidate, same questions, same rubric. Unstructured interviews predict performance only marginally above chance (Schmidt & Hunter meta-analysis, 1998).
Work samples. A 90-minute task representative of the actual job. Highest single predictor of on-the-job performance.
Reference checks done well. Not "would you hire her again?" but "describe a time she failed; how did she respond?"
Slow no, fast yes. Bezos: when in doubt, don't. Rejecting a strong candidate is recoverable; mis-hiring is not.
10 · Case
Satya Nadella · the Microsoft turnaround
Feb 4, 2014. Satya Nadella becomes Microsoft's third CEO. The company's stock has been flat for over a decade. The Windows-Office franchise is profitable but increasingly irrelevant to mobile and cloud. Internal culture has been described as "stack-ranked, blame-cultured, and intramurally hostile."
Nadella's first email to employees on day one: "Our industry does not respect tradition. It only respects innovation." His first two strategic moves: open-source .NET, and ship Office for iOS.
The pivots
Cloud-first. Azure passed AWS in growth rate by 2018. Pivoted Microsoft from packaged software to recurring cloud revenue.
Cultural rewrite. "Growth mindset" (borrowed from Carol Dweck) replaced "know-it-all" with "learn-it-all." Stack ranking abolished.
OpenAI bet. $13B invested 2019–2023. Made Microsoft the AI infrastructure layer for the late 2020s.
Result
Market cap: $300B (2014) → $3T+ (2024). The most successful CEO transition of the 21st century, by total value created. Lesson: culture eats strategy for breakfast — but the leader's first job is to choose the strategy.
11 · Case
Boeing · the slow drift
Boeing was, for half a century, an engineering culture. Pilots, designers, and machinists ran the company. Then in 1997 Boeing acquired McDonnell Douglas in a stock-funded deal that was, in effect, McD-D taking over Boeing. Headquarters moved from Seattle to Chicago in 2001. Engineering reported to finance.
By the 737 MAX program (2011 onward), the priority was margin and speed-to-market against Airbus's A320neo. The MCAS software, designed to handle a known aerodynamic quirk, was added late, certified on a sole-sensor input, and not adequately disclosed to pilots.
Two crashes: Lion Air 610 (Oct 2018, 189 dead), Ethiopian 302 (Mar 2019, 157 dead). 346 lives, the global grounding of the type, $20B in direct costs, criminal investigation, and a generational reputational fall.
The Alaska Airlines 1282 door-plug blowout (Jan 2024) broadened the diagnosis. The 2024 Senate report cited "a culture that prioritized financial performance over safety."
Lesson: leadership is downstream of value choices. When the values shift, the outcomes follow — sometimes 20 years later.
"The CEO who says 'culture' as if it were unrelated to what he measures, what he rewards, and whom he promotes — that CEO has not yet led." — editorial commentary, multiple sources
12 · Reading
What to read & watch
Books
Drucker — The Effective Executive (1967)
Collins — Good to Great (2001)
Goleman — Emotional Intelligence (1995)
Grove — High Output Management (1983)
Lencioni — The Five Dysfunctions of a Team
Scott — Radical Candor
Sinek — Leaders Eat Last
Heifetz — Leadership Without Easy Answers
Nadella — Hit Refresh
Horowitz — The Hard Thing About Hard Things
