Organizational strategy and management theory for business decision-making. Covers objectives-based management, the knowledge-worker firm, decentralization, the five management tasks, strategy-as-practice, managerial roles, and the distinction between efficiency and effectiveness. Use when structuring an organization, setting objectives, allocating decision rights, or critiquing a strategy document.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
Organizational strategy and management theory for business decision-making. Covers objectives-based management, the knowledge-worker firm, decentralization, the five management tasks, strategy-as-practice, managerial roles, and the distinction between efficiency and effectiveness. Use when structuring an organization, setting objectives, allocating decision rights, or critiquing a strategy document.
A firm is a coordination mechanism that turns inputs into outputs under uncertainty. Strategy is the set of decisions a firm makes about what to do, what not to do, and how to organize itself to do it. This skill catalogs the core techniques from Drucker's management-by-objectives tradition, Follett's integrative coordination, Mintzberg's strategy-as-practice, and related bodies of work, with decision heuristics, worked examples, and failure modes.
Agent affinity: drucker (objectives and knowledge work), mintzberg (strategy-as-practice and managerial roles)
The future is unknowable but decisions must be made
Technique 1 — Management by Objectives (MBO)
Pattern: Define a small set of measurable objectives at the top, cascade them down so every unit and individual can name how their work contributes, and review against the objectives on a regular cadence.
Historical basis. Drucker introduced the term in The Practice of Management (1954) as an alternative to "management by drives" — the practice of enthusiasm-driven campaigns that accumulate without integration. MBO replaces enthusiasm with alignment.
Worked example.A mid-sized manufacturer sets the objective: "Reduce on-time-delivery defects from 12 percent to 4 percent over two quarters." Each unit decomposes this into actions it controls: procurement reduces supplier variance, production smooths line changeovers, shipping tightens the cutoff window. Every Monday meeting begins with the delivery number. When the number moves, the unit responsible is identifiable. When it does not move after two weeks, the unit explains why and proposes a new action.
When to use. Whenever the firm has more activity than results — a signal that effort and impact have decoupled.
When it stalls. MBO fails when objectives are set by the center without bottom-up input, or when they cascade into vanity metrics that the unit can game. A unit that cannot change the number it is measured on will either resent the metric or cheat it.
Technique 2 — Effectiveness vs Efficiency
Pattern: Before optimizing how well something is done (efficiency), verify that it is the right thing to do (effectiveness). Drucker: "There is nothing quite so useless as doing with great efficiency something that should not be done at all."
Heuristic. For every activity on the team's calendar, ask: (1) What would happen if we stopped doing this? (2) If the answer is "nothing important," stop. (3) If the answer is "something bad," ask whether that bad thing is worth the time cost.
Worked example. A sales team tracks 47 weekly metrics. Review reveals that only 6 of them influence any decision the team or its managers would actually make. The other 41 are measured because someone once asked for them. Dropping the 41 reclaims 8 hours per week of analyst time and changes no outcome. That is effectiveness — not "doing the reports faster," but "not doing most of the reports."
When to use. At the start of any performance improvement initiative, before any investment in tools or process optimization.
Technique 3 — The Knowledge-Worker Firm
Pattern: When the workers know more about the job than their managers, the managerial role shifts from directing to enabling. Knowledge workers set their own tasks within a frame, and the manager's job is to set the frame, remove obstacles, and ensure accountability for results rather than activity.
Historical basis. Drucker coined the term "knowledge worker" in 1959 and spent the next 40 years elaborating the consequences. By the early 2000s he had concluded that managing knowledge work was the single most important management challenge of the century.
Worked example. An engineering team is told what to build and when, but not how. The manager's weekly one-on-one is not a status check (the engineer's commits speak for themselves) but a conversation about obstacles, priorities, and growth. Output is measured in shipped features and in code other engineers can build on. Managers are evaluated on whether their reports grow in capability, not on how busy the reports look.
When NOT to use. Manual or routine work where specification is possible and variance is a defect. Treating a manufacturing line like a knowledge-worker firm produces chaos. Treating an engineering team like a manufacturing line produces resignations.
Technique 4 — Decentralization
Pattern: Push decisions to the unit that has the information and the consequences. The center retains decisions that genuinely require the center — capital allocation, executive appointments, and policies that must be consistent across units.
Historical basis. Drucker's 1946 study of General Motors, Concept of the Corporation, was the first academic account of the decentralized multi-divisional firm and its trade-offs. Sloan's GM was the template; Drucker named what Sloan had done.
Decision guide.
Decision
Who owns it
Why
Capital over a unit-specific threshold
Center
Balance sheet is one object
Executive hiring above a level
Center
Talent is a firm asset
Policies that must be consistent (compliance, safety)
Center
Arbitrage would erode the policy
Operational priorities within the unit's budget
Unit
Unit has the information
Hiring at the unit level
Unit
Fit is local knowledge
Product roadmap for the unit's customers
Unit
Customers are local
When it stalls. Units gradually re-centralize whenever a unit fails visibly; the center responds with process, which erodes unit authority, which creates a new failure. The discipline is to tolerate the first few unit failures and intervene only when failure is systemic.
Technique 5 — The Five Management Tasks
Pattern: Drucker's decomposition of the manager's job into five functions: (1) set objectives, (2) organize, (3) motivate and communicate, (4) measure, (5) develop people. Every manager must do all five. Neglect of any one produces a predictable failure mode.
Diagnostic use. When a unit is struggling, walk the five tasks and identify which is missing.
Missing task
Symptom
Objectives
People are busy but nobody agrees on what success looks like
Organize
Work falls through cracks, same work happens twice
Motivate and communicate
People know what to do but not why
Measure
Decisions are made on anecdote and feeling
Develop people
Good performers leave, new hires take too long to productivity
Worked example. A design team is missing its deadlines. The manager's instinct is "we need better objectives." But the review reveals the objectives are fine — what is missing is measurement. Nobody is tracking the lead time between request and delivery, so nobody can see where the queue is building up. Adding a single visible lead-time chart, without changing anyone's goals, clears the backlog in six weeks.
Technique 6 — Managerial Roles (Mintzberg)
Pattern: Rather than asking what managers should do, Mintzberg's 1973 study observed what managers actually do and classified their work into ten roles in three clusters: interpersonal (figurehead, leader, liaison), informational (monitor, disseminator, spokesperson), and decisional (entrepreneur, disturbance handler, resource allocator, negotiator).
Key finding. Managers do not spend most of their time planning and analyzing. They spend most of it in brief, fragmented, verbal communication. The average manager has about 9 minutes between interruptions. Planning happens in the cracks.
Diagnostic use. When a manager says "I want to be more strategic," the question is not "how do we make strategy time" but "which of the ten roles can we reduce load on to free the manager from fragmentation." Usually the answer is disturbance-handler (stop being the escalation path for routine problems) and resource-allocator (delegate budget decisions within limits).
Technique 7 — Strategy-as-Practice
Pattern: Strategy is not only what is written in the strategic plan. It is also what the organization actually does, which is the product of daily practice by many people. The "strategy-as-practice" tradition (rooted in Mintzberg's distinction between intended, deliberate, emergent, and realized strategy) asks: what is the organization's realized strategy, and how does it differ from the intended one?
Diagnostic use. Ask any five line workers "what is our strategy?" If the answers disagree substantially, the realized strategy is not the stated one. The question then is whether the gap reflects workers ignoring the plan (a communication problem) or the plan ignoring reality (a planning problem).
Worked example. A company's stated strategy is "premium brand, high service." Field observation reveals that salespeople compete on price and routinely promise delivery dates engineering cannot meet. The realized strategy is "low price, over-promised service." The gap tells the executives that their strategy document is a fiction. They must either enforce the intended strategy (change sales incentives) or acknowledge the realized one (change the brand).
Technique 8 — Integrative Coordination (Follett)
Pattern: Mary Parker Follett (1868-1933) argued that conflict between departments or stakeholders should not be resolved by domination (one side wins) or compromise (both sides lose a little). The goal is integration — reframing the situation so that both parties' underlying needs are met without subtracting from either. Integration requires identifying what each party actually wants (often different from what they are demanding) and looking for a third option that satisfies both.
Worked example. Sales wants more salespeople. Engineering wants fewer distractions. A dominate-or-compromise resolution would either hire the salespeople (engineering suffers) or not hire (sales suffers) or split the difference (both lose). The integrative question is: what does sales need the salespeople for? If the answer is "customer discovery," the integrative solution may be a dedicated customer-research function that serves sales without interrupting engineering. Both underlying needs are met.
When it stalls. Integration is not always possible. Zero-sum situations (fixed budget, incompatible product directions) must be resolved by decision, not integration. Pretending integration is possible when it is not drags out conflict and erodes trust.
Technique 9 — Purpose Statement
Pattern: A one-sentence statement of what the firm or unit exists to do, precise enough to rule out plausible alternatives. The test of a purpose statement is: does it help decide what not to do?
Bad purpose statement. "We make our customers successful." (Every company says this. It rules out nothing.)
Good purpose statement. "We help mid-market manufacturers reduce their scrap rate below 2 percent." (Rules out: startups, enterprises, non-manufacturing, manufacturers whose scrap rate is already low, problems other than scrap rate.)
Worked example. A startup pivots three times in a year, each time after board pressure. On the fourth pivot the founder writes a purpose statement so specific that the next board meeting becomes a conversation about whether this is still the right purpose — not about which new market to chase. The pivots stop.
Technique 10 — Scenario Planning
Pattern: For decisions that must commit resources over a long horizon under irreducible uncertainty, construct 3-4 distinct plausible futures (not forecasts) and test whether the decision survives across all of them. Scenarios are stories about the future, not predictions of it.
Worked example. A utility is deciding whether to build a 40-year gas plant. Three scenarios: (1) carbon regulation stays weak, gas remains competitive; (2) carbon regulation tightens, gas becomes a bridge fuel that retires early; (3) renewables + storage drop faster than expected, gas is stranded. A decision that is profitable in all three is robust. A decision that is profitable only in scenario 1 is a bet, not a strategy.
Strategy Selection Heuristics
When approaching an unfamiliar organizational situation, use this decision tree:
Is the problem "we are busy but not delivering"? Apply MBO (Technique 1) or Effectiveness-vs-Efficiency (Technique 2).
Is the problem "our experts do not know what to do"? Knowledge-worker frame (Technique 3).
Is the problem "decisions are slow"? Decentralization (Technique 4).
Is the problem "something is falling through cracks"? Five management tasks diagnostic (Technique 5).
Is the problem "the manager wants to be more strategic"? Managerial roles analysis (Technique 6).
Is the problem "our plan does not match what we actually do"? Strategy-as-practice (Technique 7).
Is the problem "two departments are fighting"? Integrative coordination (Technique 8).
Is the problem "we cannot decide what to focus on"? Purpose statement (Technique 9).
Is the problem "we cannot predict the future"? Scenario planning (Technique 10).
Common Mistakes
Mistake
Why it fails
Fix
Copying a framework from a famous firm without understanding why it worked there
Context is not portable
Understand the frame, not the form
Setting objectives without cascading them
Lower units optimize locally
Every objective must decompose to unit actions
Measuring what is easy to measure
Drives the wrong behavior
Measure outcomes, not activity
Treating strategy as an annual document
Realized strategy drifts between cycles
Review the gap between intended and realized quarterly
Resolving conflict by compromise as default
Both parties lose
Try integration first, fall back to compromise only if integration fails
References
Drucker, P. (1954). The Practice of Management. Harper & Brothers.
Drucker, P. (1973). Management: Tasks, Responsibilities, Practices. Harper & Row.
Drucker, P. (1946). Concept of the Corporation. John Day.
Mintzberg, H. (1973). The Nature of Managerial Work. Harper & Row.
Mintzberg, H. (1994). The Rise and Fall of Strategic Planning. Free Press.
Follett, M. P. (1942). Dynamic Administration: The Collected Papers of Mary Parker Follett. Harper.
Whittington, R. (2006). "Completing the practice turn in strategy research." Organization Studies, 27(5), 613-634.
