Mintzberg's 5 org configurations: which one are you, and which one should you be?

60-Second Summary

Five configurations: Simple, Machine bureaucracy, Professional bureaucracy, Divisionalised, Adhocracy.
Each is coordinated by a different mechanism — and that mechanism determines what HR practices actually fit.
Most startups begin as Simple structures and evolve into Adhocracies or Machine bureaucracies depending on the business model.
Playbooks copied across configurations usually fail. Google's promo process doesn't fit a 30-person startup. A factory's standard work doesn't fit a creative agency.
Configurations shift as you scale; the painful transitions cluster around ~50, ~200, and ~1000 people.

Why does a fast-growing startup feel chaotic but alive, while a 50-year-old manufacturer feels orderly but slow? Why do hospitals and law firms not respond to 'standard operating procedures' the way airlines do? Henry Mintzberg answered this in 1979 with one of the most durable classifications in management theory. Understanding which configuration you are in saves you from importing playbooks that quietly do not fit.

Why a 1979 model still wins

Most management frameworks decay. Mintzberg's didn't, because he wasn't describing a fad — he was naming the structural mechanism by which different kinds of work get coordinated. Surgery is not coordinated the way a factory line is. Consulting is not coordinated the way a bank is. The mechanism is what changes, and the configuration is the predictable shape that mechanism produces.

The 5 configurations, plain-English

Configuration	Coordinated by	Typical examples	Feels like
Simple structure	Direct supervision by founder/owner	Startups under ~30 people, family businesses, small shops	Fast, personal, every decision routes through one or two people
Machine bureaucracy	Standardised work processes	Manufacturers, airlines, banks, fast food, call centres	Predictable, rule-heavy, change is slow but operations are tight
Professional bureaucracy	Standardised skills (training & credentials)	Hospitals, universities, law firms, audit firms, consultancies	Highly skilled individuals, weak central control, peer-to-peer norms
Divisionalised form	Standardised outputs (each unit accountable for its P&L)	GE, Berkshire Hathaway, conglomerates, holding companies	Headquarters sets targets, divisions choose how to hit them
Adhocracy	Mutual adjustment in project teams	Consulting firms, R&D shops, design studios, ad agencies, modern tech startups at scale	Project-based, fluid teams, lots of meetings, deep specialisation

Coordination — the hidden mechanism

The most useful idea Mintzberg gave us is that organisations have a primary coordination mechanism, and structure follows from it. There are five mechanisms — direct supervision, standardisation of work, standardisation of skills, standardisation of outputs, and mutual adjustment — and each one suits different kinds of work.

A factory standardises the work itself: every operator does the same step the same way. A hospital can't — surgery cannot be scripted at the step level — so it standardises the skill instead, through medical training, credentials and protocols. A consulting firm can't standardise either, because every client engagement is different, so it relies on mutual adjustment: smart people talking to each other constantly to figure out what to do next.

Once you see the coordination mechanism, the structure stops being mysterious. The mechanism explains why hospitals have so much administrative paperwork (the only way standardised-skill organisations achieve consistency at scale), why consultancies have so many meetings (mutual adjustment is meeting-heavy by definition), and why startups eventually stop scaling on founder energy (direct supervision caps at about Dunbar's number).

What playbooks fit which configuration

Simple → light process, founder-led hiring, no formal ladders, decisions made in hallways. Imposing Google's perf system here will kill you.
Machine bureaucracy → detailed job descriptions, structured comp bands, formal performance ratings, multi-step approvals. Imposing startup informality here breaks compliance.
Professional bureaucracy → peer-led calibration, partner/principal tracks, credentialing investment, light managerial intervention. A heavy command-and-control style will be silently ignored.
Divisionalised → strong financial targets per unit, internal market dynamics, leadership rotation, succession depth. HR runs as a corporate centre + embedded BUs.
Adhocracy → project staffing, fluid teams, deep specialist careers, dual ladders, OKRs, narrative reviews. Static org charts will be obsolete the day after publication.

The painful scaling transitions

Configurations shift as the company grows, and the shifts are painful. People who thrived in one configuration often struggle in the next. The patterns repeat across thousands of companies:

Around 30–50 people: Simple structure breaks. Either you evolve toward Adhocracy (product-led tech) or toward Professional bureaucracy (services). This is when 'we need to hire HR' usually happens.
Around 150–250 people: a layer of middle management appears. Processes that lived in someone's head get written down. Decision rights become formal. This transition kills founder-mode habits.
Around 1,000 people: Divisionalisation becomes tempting. P&L decentralises. Politics enter the building, often for the first time, because there's now budget to fight over.
Around 5,000 people: the organisation is large enough that subcultures harden into distinct micro-configurations. A successful HR strategy starts to look like running an internal federation, not a single company.

The Spotify-model warning

Most 'org model' fads (Spotify's squads/tribes, holacracy, the Netflix culture deck) describe specific configurations that worked in specific contexts. Copying the artefact without copying the underlying coordination mechanism produces theatre. Always ask 'which Mintzberg configuration are they actually operating in?' before importing the artefacts.

Diagnose your configuration in 10 minutes

What is the primary coordination mechanism in our company today? (Pick one — the strongest.)
What configuration does that mechanism imply?
Where do we behave as if we were a different configuration? (e.g. Simple structure imposing Machine bureaucracy processes)
Where are we headed next as we grow, and is the leadership team ready for that mechanism?
Which 2 HR practices currently in flight fit our configuration, and which 2 are imported from a configuration we are not?

FAQs

Frequently asked questions

Aren't most modern companies a mix?

Yes. Mintzberg himself said most real organisations are hybrids — a Machine bureaucracy at the centre with Adhocracy R&D pockets, say. The point is to know your dominant configuration and where you're hybridising.

Where do remote-first companies fit?

Remote-first amplifies whichever configuration you already are. Simple structures lose hallway coordination first; Adhocracies lose mutual adjustment hardest; Machine bureaucracies often run remote relatively smoothly because the work is standardised.

What configuration is best?

There is no 'best' — only fit. The right configuration depends on the work, the strategy, the regulatory environment, and the stage. The wrong question is 'which is best'; the right question is 'which does our work require?'

How does this relate to Galbraith's Star?

Galbraith tells you which 5 levers to align. Mintzberg tells you the underlying coordination mechanism. Together they cover the 'what to design' and the 'why these designs cluster' sides of the same problem.

Written by Pawan Joshi.Sources cited inline.

First published 16 Jun 2026See site changelog →