Org TopologiesThe home of modern org design
Visit the Academy

FAQ

Questions,
answered.

Short answers to the questions practitioners keep asking, written from the book 10X ORG. For more, ask Aiden in the Org Topologies Slack.

FROM THE BOOK 10X ORG

Questions

Each answer is short and says where to read more in the book.

  1. How Do I Place a Team or Unit on the Map?
  2. What Do Doing, Delivering, Directing and Driving Mean?
  3. What Is the Difference Between Delivering and Driving?
  4. Where Do We Start with Org Design?
  5. How Do We Move Toward an Adaptive Topology Without Creating New Silos?
  6. With AI, Do We Still Need Teams, and How Big Should They Be?
  7. I Have Not Read the Book. What Are the Essential Ideas?
  8. What Does a Broader Mandate Look Like in a Real Business?
  9. What Is the Difference Between Resource and Delivery Topology, and When Does Each Fit?
  10. How Do I Help Leadership See Org Design as a Management Tool?
  11. What Are the Risks of Widening a Team's Scope of Work, and How Do We Handle Them?
  12. How Do I Read TASKS, CAPS, PART and WHOLE, and Find My Team's Cell?

How Do I Place a Team or Unit on the Map?

The map has two axes. The horizontal one is the Scope of Skills Mandate: how many of the skills needed to deliver value a unit has, and is allowed to use.

The left is incomplete, the right is complete. The vertical one is the Scope of Work Mandate: whether the unit is trusted with narrow tasks and outputs (lower half) or owns outcomes (upper half).

Questions that help you place a unit:

  • Skills: Can this unit finish its own Definition of Done, or does it wait for other units? Is it allowed to learn new skills when the work needs them?
  • Work: Does it get tickets, specs and checklists, or does it decide trade-offs and shape the work itself? Would a customer judge its results on their own, or only as part of a bigger whole?

Map the unit as it really works today, not as its charter says. Do the skills axis first, then the work axis, and expect a debate on the second one.

Being aligned with company goals (for example OKRs) is not the same as owning an outcome. Do the exercise with a small group that sees the system from different angles, and keep it free of judgment: the aim is clarity.

In the book 10X ORG: Chapter 5, "Map to Understand"; "Scope of Skills Mandate"; "Scope of Work Mandate".

What Do Doing, Delivering, Directing and Driving Mean?

They are the four kinds of intelligence a unit can have, one for each corner of the map:

  • Doing (lower left): narrow specialists focused on closing tasks. Strong at their craft, not expected to deliver end-to-end value on their own.
  • Delivering (lower right): units with a complete set of skills that deliver value well within a bounded area. They are not asked to own the whole outcome.
  • Directing (upper left): units that oversee the whole solution space and coordinate, plan and control other units' work, such as a PMO. They usually cannot create value directly.
  • Driving (upper right): units that own outcomes and shape what gets built next, blending strategy and execution without waiting for handoffs.

In the book 10X ORG: Principle 3, "Shape Intelligence with Mandates"; Primer, archetype grid.

What Is the Difference Between Delivering and Driving?

Both are complete in skills. The difference is the scope of work.

A Delivering unit ships efficiently inside a boundary, and tends to optimize local output. Someone else steers the outcome and aligns it with other units. A Driving unit owns the outcome and decides where to focus, so it can change direction without a reorganization.

A practical way to read it: if a unit's direction comes from outside as roadmaps, tenders or tickets, it is Delivering-shaped. If it senses demand, picks what to work on and answers for the result, it is Driving-shaped. Labeling a bounded work area a "value stream" does not make a unit Driving, because a real value stream runs from the customer's need to cash.

In the book 10X ORG: "Delivering Intelligence"; "Driving Intelligence".

Where Do We Start with Org Design?

Start by seeing the current system. Map the units and how they interact as they really work, before discussing improvements.

The book's phrase for it is "fix the road first": make the roads visible, then decide where the traffic needs help. Bring a small guiding group with different views (engineering, product, HR). Cluster units by skills first, then place them by the work they own.

Next, ask what the organization needs to be good at, and choose a topology on fit: Resource, Delivery or Adaptive. Not every part of a company needs the most adaptive design, and that is fine. Then improve one thing at a time, for example with an Elevating Kata for the misfit you found.

In the book 10X ORG: Chapter 5, "Mapping"; Principle 4, "Fit for Purpose"; Elevating Katas (Appendix A).

How Do We Move Toward an Adaptive Topology Without Creating New Silos?

Silos come from slicing the organization and its work into narrow, permanent lanes. An Adaptive topology does the opposite: it lets people and teams work together on larger, shared outcomes instead of cutting them into predefined lanes.

Two things make that possible. People need multi-learning (growing beyond their primary craft, which AI tutors and copilots now make practical) and real agency.

A warning sign that you are rebuilding silos: a team keeps a fixed narrow area and its own backlog, and its mandate has not grown. Also test the fit.

The book describes a company where pushing adaptability everywhere created friction. One product area changed fast and fit Adaptive, a late-stage product with low variability fit Delivery better, and separating the two worked.

In the book 10X ORG: "Adaptive Topology"; "Misfits of Adaptive Topology"; "Scope of Skills Mandate" (Multi-Learning).

With AI, Do We Still Need Teams, and How Big Should They Be?

The book does not give a team size, and it does not say teams go away. Its argument is about mandates.

AI can speed up task-level work, and people shift from executing to supervising, handling exceptions and exercising judgment. But AI cannot make up for missing end-to-end responsibility. Without changing what a unit owns, faster execution only speeds up local output, not the performance of the whole system.

The more useful question is what each unit, human or AI, is allowed to own and learn. Used as a teacher, AI makes multi-learning cheaper, so a Delivering unit can understand customer context and trade-offs, and take on more end-to-end work if its design grants the decision rights. Whether one team can cover a larger piece of software is therefore a mandate and fit question for that organization, not a number.

In the book 10X ORG: Principle 3, "AI for Doing", "AI for Delivering", "AI for Directing", "AI for Driving"; Principle 8 (Embed Multi-Learning).

I Have Not Read the Book. What Are the Essential Ideas?

Org Topologies starts from one idea: how well an organization performs depends on how it is designed, so look at the design before the people. Then it gives you four tools:

  • The map. Two mandates, what skills a unit has and may use, and what work it is trusted with. Every unit sits somewhere on it.
  • Four intelligences. Doing, Delivering, Directing and Driving describe the four corners.
  • Three topologies. Resource (work divided by function), Delivery (cross-functional teams with bounded work) and Adaptive (shared ownership of outcomes). None is best in general. The rule is fit for purpose.
  • Elevating Katas. Small repeatable routines that move a unit toward a better fit.

Add multi-learning (people growing beyond their craft, which AI tutors make cheaper) and you have the core. A good first step is to map your own organization and see where the design and the purpose disagree.

In the book 10X ORG: Principles 3 and 4; Chapter 5, "Map to Understand"; Appendix A (Elevating Katas).

What Does a Broader Mandate Look Like in a Real Business?

The book uses everyday examples. In a restaurant, a unit that chops onions or runs the grill is incomplete in skills and works on outputs, while a unit that delivers the whole meal and the dining experience is complete and owns the outcome.

In an insurer, clerks who verify claims or chase documents are narrow. A broker who analyzes risk, advises, talks to the client and executes the trades is delivering an outcome with a complete skills mandate.

Two moves widen a mandate: more skills (horizontal), and more ownership of the result (vertical). For a team that only builds a slice of a product, ask what it would need to own the customer's whole result, and who would then stop being a hand-off. Widening needs real decision rights, not just a larger scope on paper.

In the book 10X ORG: "Scope of Skills Mandate"; "Scope of Work Mandate"; "Customer Impact".

What Is the Difference Between Resource and Delivery Topology, and When Does Each Fit?

In a Resource topology, work is divided by function. People specialize deeply, managers and coordinators allocate the skills, and the goal is performance through efficiency. It fits stable conditions with predictable work, standard quality and little change, such as a central policy function whose rules others carry out.

In a Delivery topology, the strongest dependencies are removed by building cross-functional teams that finish work within clear boundaries, with fast flow. It fits work that is interdependent but still predictable.

A team that owns a whole checkout flow is an example. Its limit shows when goals shift often: teams can run fast in a direction that is no longer right.

The choice is about fit, not progress. Moving from Resource to Delivery means broader skills and a narrower, well-bounded scope of work.

In the book 10X ORG: "Resource Topology"; "Delivery Topology"; "The Mandates"; "Fit Is All You Need".

How Do I Help Leadership See Org Design as a Management Tool?

The book frames org design as the way structure serves strategy: how roles, accountabilities, units and coordination are chosen so that the organization can do what it intends to do. Leaders engage when it is about their goals, not about a framework.

Three moves from the book. First, map the current system together, with a small guiding group that sees the organization from different angles, and keep it free of judgment.

Second, remind everyone that a unit's place on the map is not a verdict on its people. It shows the authority and accountability that were designed for it.

Third, take one real outcome and follow how it moves: where it waits, who decides, which hand-offs it crosses. A concrete misfit persuades more than a model does. Also be ready for the valid concerns that follow (see the next question).

In the book 10X ORG: Chapter 4, "Growing Misfit"; Chapter 5, "Mapping"; "Defining Mandates".

What Are the Risks of Widening a Team's Scope of Work, and How Do We Handle Them?

The book takes the concerns seriously. When everyone is responsible for everything, no one is accountable and things slip through.

Specialists can feel their craft is losing value. Line managers whose job was to grow skills and ensure quality may wonder what their role becomes. There can also be liability when one team works in another team's domain.

Its answer is not to blur ownership. Clear accountability still matters: every unit needs ownership of its outcomes, and the closer the costs, investments and trade-offs sit to the unit that owns them, the better it can optimize.

Widen the mandate on purpose, with decision rights and clear ownership, and treat specialist depth as something to keep and build on, since multi-learning extends it. The book does not discuss code quality or maintainability in particular.

In the book 10X ORG: Chapter 6, "Valid Concerns"; "Scope of Skills Mandate" (Multi-Learning); "Fit Is All You Need".

How Do I Read TASKS, CAPS, PART and WHOLE, and Find My Team's Cell?

A cell name is row-number. The row is the scope of work (bottom to top), the number is the scope of skills (left to right).

Rows: Tasks (narrow tasks for one expertise, like a sous-chef preparing ingredients), Capabilities (a full feature or service element, like a claims group), Partial solution (one business vertical, like Uber Ride versus Uber Eats), Whole solution (the whole customer problem, like a two-person startup).

Numbers: 1 Functional (one specialty), 2 Multi-skill (cross-functional but still incomplete), 3 End-to-end (everything needed to deliver the value), 4 Expanding (end-to-end, plus the mandate to acquire new skills).

So CAPS-3 is an end-to-end unit that owns a full capability. Both scales are the two axes of the Org Topologies map, and mapping is the first step of the MADE method.

How to find your cell:

  1. Pick one outcome. A unit is complete or incomplete for that outcome, so the same team can land in different cells for different outcomes.
  2. Skills first (the number). Can the unit deliver that outcome without depending on others for skills (3), and may it learn new skills when needed (4)?
  3. Then work (the row). What part of the outcome does the unit own, and may it change that part when it learns something? Climb four customer questions and stop at the first "no":
    • Does the customer understand or care about what the unit produces? No: ingredient, Tasks.
    • Does the customer transact for it (pay, order, sign off)? No: dish, Capabilities.
    • Does the unit deliver the whole service, start to finish? No: dish, Capabilities.
    • Do customers judge the unit by its own work, and can it act on that feedback? No: meal, Partial solution. Yes: experience, Whole solution.
  4. Read the cell and combine row and number, for example PART-3.

Answer for what is true today, not what could be. A unit that is rated but may not change its own work stops at Partial solution. The row is a design decision about the mandate, not a measure of ability, so a strong team can sit low on purpose.

In the book 10X ORG: Chapter 5, "Mapping"; "Customer Impact"; Primer, the archetype grid, the five vertical levels and the examples table.

Short answers only. Learn from Aiden: ask Aiden, our AI built on the book 10X ORG, in Slack for more.