Glossary.

A tenfold increase in an organization's capability to learn, evolve, create value and stay strategically aligned. It is about more adaptability, impact and relevance, not more activity.

The book's people-centric, AI-ready system of nine principles for building organizations that can learn, adapt and perform. Org Topologies is the map and method that powers it.

An approach to raising performance through people-centered, strategy-driven, AI-ready organizational redesign: a visual map plus the MADE change method.

The 2x2 visual language that places organizational units by skills mandate (horizontal) and work mandate (vertical), so structure, dependencies and misfit become visible and shareable.

The process of developing and selecting an organization's structure (roles, reporting lines, units, interfaces, coordination) to meet its strategy and mandates; also the field that studies how structure improves effectiveness.

Deliberately aligning business strategy, the org goal and the change process, instead of letting the design accrete by accident.

How well the organization delivers what its strategy requires; achieved when org design elements are kept in sync with the strategy.

Jay Galbraith's model of five interlinked design elements (org design criteria drawn from the org goal, structure, processes, rewards, people) that must be adjusted together. MADE adds a learning loop around it.

The one primary thing an organization deliberately optimizes for, chosen to support its strategy and used to settle trade-offs.

The external result a strategy aims at, such as profit or market share. It is uncertain; org design, by contrast, is internal and largely in your control.

The primary concern is the org goal being optimized. A secondary concern is something valued and kept in view but pursued only as far as it does not undermine the primary one.

Presenting two valued options as opposites when a range of choices exists; reframing with a primary and a secondary concern reveals new options.

A hidden commitment that works against a stated improvement goal, which explains why people resist even changes they support.

An org design whose structure and mandates support what the organization is optimizing for. No topology is better in itself; the question is which one fits.

A gap between what the structure is optimized for and what the strategy needs, made visible by mapping and assessing.

The ability to change direction easily and cheaply, guided by the learning needed to deliver better customer value with less friction. It is the org goal of the Adaptive Topology.

The ability to follow value into the unknown, learn there and still deliver, so the organization recombines as conditions change without reshuffling its structure.

The abilities, such as adaptability, impact and relevance, that ultimately raise performance; each topology trades some capabilities off against others.

What an organization can collectively see, learn and solve. It emerges from the mandates assigned to units, not from individuals, and mandates set its ceiling.

A cohesive set of org design elements that reinforce one another toward a goal. Scrum is one: adopt only its events and leave structure, people and rewards untouched, and it fails for lack of systemic alignment, not for lack of merit.

Choosing structure, mandates, processes, rewards and people on purpose, around one primary org goal. The alternative is design that accretes through ad hoc reactions to internal political pressures, which sends mixed signals to everyone.

Leaving change to outsiders or to a leadership blessing instead of owning it. People then resist being changed, comply on the surface and mimic progress, and the improvement fades as a temporary overlay.

Leaders set the direction and people shape the change from within. Together they move decisions closer to the work, shorten feedback loops and reduce dependence on central coordination, as at Poster.

Real org design starts from inherited structures, people, habits, commitments and budgets while the business keeps running. It is refactoring on the move, not drawing an ideal diagram on a blank whiteboard.

Cost is a constraint to design within, not an org goal to optimize for. A top priority of cutting cost, taken to its end, means shutting the company down.

A bundle of authority and responsibility to decide, explore, act and deliver results within a defined scope. Mandates set the maximum complexity a unit can handle.

The horizontal axis: how many skills a unit has and is authorized to apply in its work. Units on the left are incomplete, on the right complete.

The vertical axis: the kind of work a unit is trusted to handle, from narrow tasks at the bottom to shared ownership of whole outcomes at the top.

A unit that lacks the skills, or the mandate to use them, to deliver customer value end-to-end, so it depends on others and creates handoffs. Sits left of the vertical divider.

A unit with the skills, data and authority to deliver value independently, without relying on other units. Sits right of the vertical divider.

The partial products of work (tickets closed, features shipped, documents done). More output does not create more customer value.

What customers and stakeholders actually value, such as the whole meal, a healthy patient or a customer who stays. Units above the horizontal divider own them.

The line on the map separating incomplete units (left) from complete units (right).

The line on the map separating units optimized for partial outputs (below) from units designed to deliver whole outcomes (above).

The economic reading of the horizontal axis: the cost of getting work through handoffs, which falls as a unit moves right.

The economic reading of the vertical axis: how hard it is to redirect effort to new opportunities, which falls as a unit moves up.

Lower-left quadrant: incomplete units working on outputs in narrow domains (individual contributors, groups, teams, AI agents). Strong at precision, dependent on others for end-to-end value.

Lower-right quadrant: complete units working on outputs within a bounded scope, delivering with good flow but not owning the whole.

Upper-left quadrant: incomplete units that oversee the whole solution space and coordinate, prioritize and control, but do not create customer value directly.

Upper-right quadrant: complete units owning outcomes and shaping what is built next, blending dreaming, thinking and doing; the Adaptive Topology.

People or teams growing beyond their primary craft, building skills across domains and applying them in real work. It is the mechanism that expands the skills mandate and, with outcome ownership, moves units diagonally.

Broadening a person's skills across several domains, so units can take on more of the work themselves.

An organizational model in which small autonomous teams share purpose and transparency and are linked into a network that co-owns a broad outcome.

A team that combines several functions. By itself it only moves a unit horizontally; Adaptive Topology needs units that are both cross-functional and cross-component.

A reliance between units. Painful only when people are forced apart and cannot work synchronously; when they work on shared work together it becomes ordinary collaboration.

Adam Smith's example of division of labor, used as a picture of the Resource Topology: workers pinned to narrow subtasks, efficient in stable, low-variability work.

Moving units out of narrow lanes (workers in Resource, teams in Delivery) by expanding both mandates up to the level of the whole product.

Organizing teams around architectural components so each delivers only its part of a feature; this structure makes development sequential, a structural outcome rather than a mindset problem.

In Lean and the Toyota Production System, the flow of value end-to-end from the customer's view, activated by customer pull, from concept to cash.

How broadly an organization defines its product, from one component to the whole; a broader definition widens the work mandate.

Scrum's statement of when work is truly finished and usable: integrated, tested, documented where needed and ready for real use. The horizontal axis reads as the growing completeness of this definition.

Early signals chosen to show that an organization is on track to deliver outcomes. Outcomes such as a customer who stays are lagging and hard to measure, so smart proxies give a reliable sense of progress sooner.

A recognizable pattern, drawn from observation, of which quadrants an organization occupies and so what it is optimized for. There are three common ones and they can be combined.

Directing units over Doing units: work divided by function, deep specialists, planning by resource managers. Org goal: efficiency of specialized skills under stable demand.

Directing units over Delivering units: stable cross-functional teams in bounded areas. Org goal: end-to-end flow of predictable work from demand to delivery.

Driving units: broad skills and work mandates, units that swarm around outcomes. Org goal: minimize the cost of learning and reconfiguration in volatile environments.

Metaphor for the Resource Topology: precision through specialization, effective where work can be planned in advance.

Metaphor for the Delivery Topology: stable cross-functional teams optimized for flow, coordinating through synchronous collaboration inside teams.

Metaphor for the Adaptive Topology: an organization that senses change, integrates learning and adjusts direction continuously.

Letting multiple topologies coexist in one organization where parts do different kinds of work; each should be self-sufficient, accountable for outcomes and kept evolving.

Organizations typically drift from Adaptive toward Resource as they hire and specialize; the book calls this aging. The reverse path, Resource to Delivery to Adaptive, is descaling.

The short spike in performance after a transformation followed by drift back to the old baseline, because the underlying paradigm did not change.

Putting faster engines (specialists, AI) into a jammed system makes motion louder but flow no better: local speed does not become global performance.

Each part improving what its own system rewards, which can lower performance of the whole.

The cycle of recurring, wasteful reorganizations that each treat symptoms while increasing instability.

Teams clustering fluidly around high-value outcomes in an Adaptive Topology, like swarm intelligence in nature.

Change without real structural backing: fresh slide decks, labels and jargon while the deeper performance problems stay as they were. Each new fad, now often co-written with generative AI, repeats the cycle.

Treating change as a project with start and end dates, deadlines and milestones. It teaches the organization to change the change so it looks good on paper, and diverts attention from real behavioral change.

The exhaustion that sets in when repeated shake-ups never address the real problems. Morale drops and performance stalls.

Applying a single organizing logic, metric set or governance model across the whole company in the belief that consistency equals efficiency. Across unequal contexts it destroys value and becomes a misfit at the Gemba.

The common practice of slicing an organization into bounded parts, such as squads, tribes or stream-aligned teams, each with a narrow mandate. It fits locally until work shifts, then falls between the cracks and another reorganization follows.

Gene Gendel's picture of how client companies are taxed three times: by consultancies selling big-bang transformations, by frameworks that demand retraining and licensing, and by the tools that lock those frameworks in place.

Quick, smooth movement of known work inside a team's lane. In the Delivery Topology the flow is local: when an outcome needs many teams, local speed does not add up to global speed.

A temporary team spun up to finish cross-cutting work that does not fit the stable team boundaries. It signals a fall back into resource allocation and crisis management when unplanned work arrives.

Sequential development, where each team optimizes its part, hands over and waits. Feedback arrives late and iteration collapses into stage gates. The book treats it as a structural outcome of component teams, not a mindset problem.

Work where demand, scope or technology change often, as in product development, where no feature is ever built twice. It calls for fast feedback, end-to-end responsibility and continuous learning; narrow stable lanes drift out of fit.

A downward spiral in which teams with weak trust retreat into isolated tasks to avoid the discomfort of collaboration. External coordinators then assign work, and the units slide back toward the Resource Topology.

Efficiency has no single meaning. In the Resource Topology it is maximum use of specialized skills; in Delivery, fast reliable flow of known work; in Adaptive, lowering the cost of learning, reconfiguration and redirection.

Exploit optimizes the delivery of value already understood (Delivery Topology); explore optimizes learning to discover what is valuable (Adaptive Topology). Do not turn them into camps: master both. Driving units integrate the two.

The detailed version of the map, with four levels on each axis and sixteen archetypes, used for precise assessment.

A common pattern of how an organizational unit is structured and operates. The 4x4 map names sixteen of them.

Short names for the sixteen 4x4 cells: the row of the work mandate (TASKS, CAPS, PART, WHOLE) plus the skills column from 1 to 4. CAPS-3 reads end-to-end skills at the capabilities level.

The four levels across the 4x4 map: Functional, Multi-Skill, End-to-End and Expanding. The 2x2 divider sits between Multi-Skill and End-to-End.

The four levels up the 4x4 map: Tasks, Capabilities, Partial Solution and Whole Solution. The 2x2 divider sits between Capabilities and Partial Solution.

Older letter-plus-digit archetype codes (e.g. A1, B2, C3, Y1) used on pre-2025 posts.

The five unit symbols on every map: human, AI agent, pool, team, team of teams.

The Org Topologies method for continuous, systemic change: Map, Assess, Design, Elevate, and repeat, built around Galbraith's Star Model.

First MADE step: bring the invisible system of collaboration into view by placing units on the map, at the Gemba, asking which parts create value and how they interact.

Second step: judge, not people but the system, whether the current org design fits the business ambitions and strategy.

Third step: choose the one primary org goal and shape structure, processes, people and rewards to support it.

Fourth step: build the capability to make the new design work, by moving right and up with Elevating Katas as small experiments.

The two halves of MADE: Study is Map plus Assess; Improve is Design plus Elevate.

Mapping again after elevation to check what actually changed and close the learning loop.

A quick, low-risk pass through MADE or a half-day mixed working session that lets people experience a different way of working before committing to structural change.

Curiosity (Map), honesty (Assess), commitment (Design) and openness (Elevate).

A deliberative workshop in which affected people learn why an org goal is chosen, what it enables and which trade-offs it implies, to build commitment.

Gemba is where value is created; Go See means leaders observing the real work there instead of relying on reports. Mapping and Go See reinforce each other.

A timeboxed immersion where leaders work with teams for days or weeks as active participants.

A shared decision space that pulls leaders out of local views toward strategic goals and systemic constraints.

A core skill of MADE: performance comes not from individual elements but from how they interact. Improving visible roles, processes and tools while leaving interaction patterns untouched brings no lasting gain.

Making sense together of how work really happens, by observing the system rather than judging people. In MADE it reduces the psychological threat of change and prepares learning and commitment before any structural change.

The gap between leaders and the real system of work, created by physical or structural separation. Reports and dashboards then filter reality. Go See and the Gemba counter it.

Lasting change is owned by those who create it; leaders must own it and cannot delegate it, and people support what they help create.

Organization design turns intent into a system and drives performance more reliably than standalone culture programs.

Reshape the skills and work mandates to change what the organization can see, learn and act on.

Choose deliberately what to optimize for, accept what you will not optimize, and pick the topology that fits.

Change as a continuous evolution of fitness using MADE.

Broaden mandates so people and teams follow value, learn and deliver, instead of re-slicing the organization every time work shifts.

Choose one primary org goal, accept its trade-offs and embed it into structure, mandates, incentives and decisions.

Embed learning across domains so people can take broader responsibility and recombine skills as work evolves; amplify with AI.

Use AI to accelerate the chosen org design direction, after design, not instead of it.

A structured, repeatable improvement routine that turns org design intent into lived behavior, applied mainly in the Elevate step.

The Elevating Katas Catalog in Appendix A, grouped by Star Model element: Strategy, Structure, Processes, Rewards, People.

Each kata name starts with "Elevate by" followed by a gerund phrase, e.g. Elevate by Expanding the Product Definition.

Purposeful adoption of AI paired with deliberate organizational design: first find the missing capabilities and structural constraints, then use AI to build and accelerate them.

Naive, sporadic, ad hoc spreading of AI across existing silos, which amplifies the current pattern and the Ferrari effect.

The line that sums Principle 9: set AI aside first, understand the structure, then use AI as a lever.

An AI system acting as an organizational unit. It can sit on the map like any unit: single-purpose agents are Doing units; agents can join Driving collaboration.

Designing the mandates of organizations where AI agents work across digital workflows, so they do not recreate a Taylorist split between thinking and doing.

AI-assisted spec-driven development: humans specify intent and AI agents turn it into plans, tasks and code.

AI treated as an available service that lowers the cost of learning and doing, so specialization is no longer scarce.

The book's custom GPT and 10X ORG expert, powered by Org Topologies; he/him.

The book's closing stance on AI: elevate people to higher mandates instead of eliminating them.

Routine expertise applies known procedures well in stable conditions; adaptive expertise extends, recombines or reinvents skills when conditions change.

The gradual conversion of novel work into routine work that machines absorb, in waves; the boundary of what stays human keeps moving.

Challenge the limits that organizational design puts on people, and build environments that elevate natural human strengths, such as learning, combining and recombining skills, instead of competing with AI on narrow tasks.

An employee's ability to remain effective as work conditions shift. It correlates with adaptive expertise, and the book stresses that experience alone does not produce it: it is developed through exposure to the unexpected.

The result when everyone races to adopt the same AI tools, models and frameworks without changing structure: no advantage, only higher costs for consultants, tools and usage.

A leader who is accountable for understanding the system, choosing a direction and energizing others to co-own and improve it; MADE helps leaders become true owners.

A leader, consultant or change agent who sees the people behind the boxes and lines on the map.

A diverse leadership group that guides the change together.

Certified Org Topologies Practitioner: the Academy course for improving your own organization.

Certified Org Topologies Consultant: the Academy class for consultants and transformation leaders.

A broader role name that replaces rigid titles such as Junior Java Developer in an Adaptive Topology, with clear progression levels.

In the book's view, managers grow capability and improve the system rather than direct work.

People opting in to shape and lead a change instead of being pushed into it. A core group of volunteers carries the change with energy and credibility, and peers choosing to join gives it legitimacy and momentum.

Cognitive load is the mental effort needed to do a task and learn while doing it. People overload when the cost of learning is high, not when they know much; narrow ownership often raises load through handoffs and context switching.

Mastery in one narrow skill. It makes sense when expertise is scarce and demand for it effectively unlimited, a rare case; AI democratizes expertise, so most people need to become multi-learners.

Keeping people and their accumulated learning together, instead of treating them as interchangeable resources shuffled between temporary groups. The Adaptive Topology compounds this continuity into long-term performance.

Skills people hold beyond their job title, from an old course or earlier work, which can rise to the occasion when needed. Capability extends beyond role names.

A team with broad skills that owns a whole product instead of one component. Poster moved from twelve narrow specialist component teams to six versatile whole-product teams; Y Soft runs twenty as a team-of-teams.