Macro- and micro-structure define who sits where and who leads what. Decision rights, governance, and management processes define how power actually flows. If you get these wrong, even a beautifully designed structure will deliver slow, political, and inconsistent decisions. If you get them right, you can often get strong performance even with an imperfect structure.
This chapter focuses on a practical sequence: identify the decisions that really matter, clarify ownership using simple tools, design governance bodies that add value, embed decisions in core management processes, and then clean up the governance clutter that almost every organization accumulates over time.
10.1 Identifying Critical Enterprise Decisions
Most organizations treat decisions as by-products of meetings and processes rather than as design objects. In an org design effort, you must invert this logic: start from decisions, then design roles, forums, and processes to make those decisions well and at speed.
The first discipline is to focus on critical decisions. If you try to map every decision in the company, you will drown in detail and create decision-rights matrices that no one reads. Instead, identify a manageable set of decisions that shape performance and risk.
A simple way to do this is to look across four categories:
- Strategic and portfolio decisions
Examples: entering or exiting markets, large capital allocations, M&A, major product/platform bets, portfolio restructuring. These decisions set direction and reallocate resources at scale. - Commercial and customer decisions
Examples: pricing and discounting policies, target segments, key-account strategies, channel mix, product line-up by market. These decisions drive growth, margin, and customer experience. - Operational and investment decisions
Examples: capacity investments, network and footprint changes, sourcing strategies, major technology projects, service levels, inventory policies. These determine cost, reliability, and scalability. - Risk and control decisions
Examples: risk appetite, policy exceptions, credit sanctions, safety standards, data privacy rules, compliance interpretations. These shape the organization’s risk profile and regulatory standing.
Within each category, ask leaders and key managers a few direct questions:
- “Which decisions, if consistently better and faster, would have the largest impact on our results?”
- “Which decisions are currently slow, repeatedly escalated, or frequently revisited?”
- “Where is it genuinely unclear who has the final say?”
You will quickly generate a list of recurring problem decisions (“we always fight about global discounting rules”, “no one is sure who approves new SKUs”, “capital requests bounce for months”). Prioritize a shortlist, typically 15–30 decisions at enterprise level, and a second layer of unit-specific decisions where needed.
Next, clarify the desired characteristics of each decision. For example:
- How fast does it need to be (hours, days, weeks)?
- How much cross-functional input is truly needed?
- What level of risk is involved?
- How reversible is it (easy to adjust vs hard to unwind)?
This matters because not all decisions deserve the same governance. A highly reversible pricing tweak in a small segment does not need a multi-layer approval chain. A major plant closure or platform choice probably does.
Finally, consider where each decision should sit in the future organization, independent of current personalities. If you were designing from scratch, who would be the natural owner: a BU head, a functional leader, a regional MD, a corporate committee, a risk officer? This gives you a first sketch of the decision architecture that you will refine using more formal tools.
10.2 RACI and RAPID: Tools for Clarifying Decision Ownership
Once you know which decisions matter, you need to make ownership clear. Two simple tools are used widely: RACI and RAPID. Both are useful, but only if applied with discipline and restraint.
RACI: Responsible, Accountable, Consulted, Informed
RACI is a way to clarify roles in tasks or processes. It works best when you are mapping recurring activities or value streams.
- Responsible: who actually does the work. There can be many “R”s.
- Accountable: who is ultimately answerable and has the “yes/no” authority. There should be exactly one “A” per activity.
- Consulted: who must be consulted before action is taken (two-way communication).
- Informed: who needs to be told after decisions or actions (one-way communication).
RACI’s strengths are its simplicity and its familiarity. Its main failure modes are overuse and vagueness. Teams produce giant matrices with dozens of activities and stakeholders, and then argue endlessly about whether someone is C or I. The result is a static document no one uses.
To avoid that, use RACI:
- For a limited set of end-to-end processes that are strategically important or problematic.
- At a meaningful level of aggregation (process steps or decision clusters, not every micro-task).
- With a clear rule: one A, a few R’s, and very few C’s per line.
If your RACI has more Cs than Rs, you are designing for consultation and slow decisions rather than clarity.
RAPID: Recommend, Agree, Perform, Input, Decide
RAPID, popularized by Bain, focuses explicitly on decisions rather than tasks. It is particularly useful for complex, cross-functional decisions that suffer from unclear ownership.
- Recommend: the role that pulls the analysis together and proposes a course of action.
- Agree: roles that must agree because they hold veto rights (often Legal, Risk, Compliance).
- Perform: roles that execute the decision once made.
- Input: roles that provide data or perspectives but do not own the recommendation or decision.
- Decide: the single point of decision—the person or body with the final call.
RAPID is powerful because it distinguishes between input, recommendation, and decision. Many organizations confuse these: the people who know the most about a topic assume they should decide; the people with formal authority assume they must drive the analysis. RAPID allows you to say, for example: product management recommends, finance and risk agree, sales and operations provide input, and the BU head decides.
Use RAPID for your shortlist of critical enterprise decisions identified in 10.1. For each decision, explicitly assign R, A, P, I, D roles. Then test:
- Is the Decide role at the right level (not everything escalated, not everything pushed too low)?
- Is Recommend located where expertise and proximity to data sit?
- Are Agree roles truly few and justified by legal or risk requirements?
- Is Perform properly aligned to the structure so that execution is feasible?
You do not need RAPID for every decision. Use it to solve real ambiguity and recurring conflict. If a decision is straightforward and local, a simple “this role decides” is enough.
Practical tips for using RACI and RAPID
- Keep matrices short and live. They should fit on one or two pages and be updated as roles evolve.
- Use them in real meetings. When a decision is stuck, pull out the RAPID or RACI and ask, “Are we honoring what we agreed?”
- Use them to train leaders. New leaders often struggle most with “what am I allowed to decide?” Clear decision tools accelerate their ramp-up.
The test of success is not that you have a complete matrix; it is that people stop arguing about who decides and instead debate the substance.
10.3 Governance Bodies: Committees, Forums, and Management Meetings
Decision tools help at the level of individual decisions. Governance bodies are how you institutionalize decision-making and oversight over time. Nearly every organization develops too many of them; very few regularly clean them up.
A useful way to think about governance bodies is by primary purpose:
- Decision-making bodies
Examples: investment committees, risk committees, product councils, pricing boards, capital allocation committees. Their job is to decide on defined topics within set limits. - Performance and management forums
Examples: business reviews, operating committees, sales and operations planning (S&OP) meetings, pipeline reviews. Their job is to review performance, manage variances, and steer actions. - Advisory and design forums
Examples: architecture boards, innovation councils, transformation steering groups. Their job is to provide expert guidance, align stakeholders, and shape future designs. - Information-sharing forums
Examples: cross-functional updates, townhalls, communities of practice. Their job is to share information, connect people, and build community.
Problems arise when a forum tries to be all these things at once. A product council that is half information-sharing, half debate, and half decision-making will make few decisions and leave many confused.
For each key governance body in your design (especially those at enterprise and BU level), define:
- Mandate: What decisions does this body own? What topics does it review or advise on? What is explicitly out of scope?
- Authority: What can it approve? What must it recommend to a higher body? What can it only advise on?
- Membership: Which roles must be present? Are they empowered to decide, or are they representing others?
- Cadence and duration: How often does it meet? For how long? Are there fixed agendas or rotating topics?
- Inputs and outputs: What materials are needed in advance? What decisions or actions should come out of each meeting?
A simple one-page “charter” per key committee or forum is enough. The discipline is to align the charter with decision rights: if a committee is the “Decide” for a given RAPID, that should show up in the mandate and authority.
Also consider the vertical architecture of governance. A common pattern is:
- Board or supervisory body
- Enterprise executive committee (ExCo)
- BU leadership teams
- Functional leadership teams
- Cross-functional forums (e.g., S&OP, risk, investment)
Check for duplication: are multiple bodies reviewing the same topics at different levels without clear escalation logic? For instance, if both a BU committee and an enterprise committee review the same investments, clarify thresholds and scope: the BU decides up to a certain size; above that, the enterprise committee decides. If a topic is being reviewed three times up the chain with no added value, simplify.
Finally, ensure that governance forums are staffed and run to actually decide. That means:
- Pre-reads circulated early, focused on options and recommendations rather than raw data.
- Clear decision agendas: “We need decisions on items 1 and 2, input on 3, FYI on 4.”
- Chairs who enforce discipline—closing debates, clarifying decisions, and documenting outcomes.
If your steering committees and councils are mostly “update theaters,” that is a design problem, not just a meeting-facilitation issue.
10.4 Management Processes: Planning, Budgeting, Performance Reviews
Governance bodies meet periodically; management processes run continuously. In org design, you should treat processes like planning, budgeting, and performance reviews as core vehicles for exercising decision rights and reinforcing structure.
Four processes usually matter most:
- Strategic and operating planning
This is where you translate strategy into targets, initiatives, and resource allocations. Design questions include:- Who sets top-down targets, and who proposes bottom-up plans?
- Where are trade-offs between BUs or segments made?
- How are cross-unit initiatives prioritized and funded?
- If you have moved to a more segment-based or product-based structure, planning processes should mirror that: plans and targets by segment or product, with clear owners presenting integrated views. If you retain a strong functional logic, planning should emphasize capability and capacity plans owned by functional heads.
- Budgeting and capital allocation
Budgets are often where decision rights become most contested. Clarify:- Which roles or bodies own budget envelopes at each level?
- Who can reallocate within a BU or function without additional approvals?
- Which types of spend (e.g., headcount, capex, marketing) require central approval?
- Where possible, push budgeting authority closer to accountable P&L owners while maintaining guardrails for enterprise priorities and risk. If everything requires central approval, you will undermine the autonomy implied by your structure.
- Performance reviews and business reviews
Regular performance dialogues embed accountability in the organization. Design:- The cadence and structure of business reviews (monthly, quarterly).
- The role of BUs vs functions in presenting performance.
- How cross-unit issues are surfaced and resolved (performance of a shared service, for example).
- Align these reviews with your macro-structure: if BUs own P&Ls, BU heads should be front and center in reviews, with functional leaders providing insight and challenge rather than owning the numbers themselves.
- Talent and leadership processes
Organizational performance depends heavily on who is in key roles. Integrate org design with:- Succession planning for critical roles in the new structure.
- Talent reviews across BUs and functions to identify where to place top talent.
- Leadership development and moves that reinforce the new design (for example, rotating leaders across units to break silos).
In each process, revisit decision rights:
- Who decides final targets?
- Who approves hiring for critical roles?
- Who signs off on the execution of major cross-functional initiatives?
If you have clarified ownership of decisions in 10.2 but then run planning and budgeting processes that ignore those decisions, the design will not hold. Conversely, well-aligned processes will reinforce decision clarity every quarter without needing new charts or frameworks.
10.5 Simplifying Governance: Reducing Committees and Complexity
Most organizations reach an inflection point where governance complexity becomes a performance drag. Committees multiply to solve specific issues, then persist after the issue passes. New initiatives bring new forums; old ones rarely die. As part of your org design, you should deliberately simplify governance.
Start with a governance inventory. For the scope of your design, list:
- All standing committees, councils, forums, and recurring management meetings above a certain level (for example, director level and up).
- For each: purpose, chair, membership, cadence, typical agenda items, and any formal decision authority.
You do not need perfection; a 70–80% view is enough to see patterns. Then, categorize each forum:
- Decision-making
- Performance management
- Advisory/design
- Information-sharing
Check for duplication: multiple bodies handling similar topics with overlapping memberships. Also note attendance: meetings with large, passive attendee lists are prime candidates for redesign.
Next, apply three questions to each forum:
- Is this forum still needed?
- Does it address a continuing need that cannot be met more efficiently through other means?
- Is it solving a problem that the new structure and clarified decision rights will already address?
If not, mark it for elimination.
- Is this forum at the right level and scope?
- Could decisions or reviews be pushed down to BU or functional level?
- Are topics better handled in a more focused forum with fewer people?
If yes, consider redesign: splitting, delegating, or narrowing scope.
- Is this forum designed to decide or merely to inform?
- If its real job is information-sharing, can you replace it with dashboards, written updates, or occasional town halls?
- If it is supposed to decide, does it have clear authority and pre-read discipline?
If not, redesign the format and mandate.
From this analysis, build a governance simplification plan:
- Forums to eliminate (with timing and communication).
- Forums to merge (for example, combining overlapping advisory councils).
- Forums to redesign (new mandates, memberships, cadences).
- New forums required by the new design (for example, a cross-BU investment committee if you shift to a more decentralized structure but want enterprise coordination on capex).
Communicate the new governance model as part of your overall design communication, not as an afterthought. People should understand:
- Where decisions will now be made.
- Which meetings they no longer need to attend.
- What is expected of them in the meetings that remain.
A short governance “map” can help: a one-page view of top-level committees and forums, their purposes, and their interconnections.
Finally, treat governance simplification as an ongoing discipline, not a one-off event. Build into your management routines a simple rule: every year, each committee must re-confirm its mandate and demonstrate its value, or be sunsetted. Encourage leaders to question whether new issues really require new bodies, or whether they can be addressed within existing forums and decision-rights structures.
If you are deliberate about decision rights, governance, and management processes, you will find that many of the everyday frustrations managers report—unclear accountability, slow decisions, excessive meetings—begin to recede. Structure sets the stage, but decisions and governance determine how the play actually unfolds.
