Clear decision roles matter more than broad agreement

Why companies keep approving decisions they’ve already made

For years, marketing managers at a retail bank needed approval from finance and risk management before sending direct-mail campaigns to customers. Finance checked the financial assumptions behind each campaign, while risk management examined projected losses. Even when a mailing was identical to one the bank had previously approved, both departments had to review it again.

The reviews protected the bank against financial risks, but they also treated every campaign as a new decision. Marketing managers could repeat an established activity only after finance and risk management had reconsidered assumptions they had already accepted. The bank followed its own rules, yet those rules created delays without changing the underlying decision.

Eventually, the bank established guidelines defining which marketing decisions managers could make independently. Campaigns within those limits could proceed without further approval. New offers and campaigns targeting new customer populations still required review when they fell outside the guidelines. Finance and risk management retained control over decisions involving different financial assumptions or exposure to losses.

The change addressed a common organizational problem: companies often involve people in decisions because those people have traditionally held approval authority. Their involvement can continue long after the conditions that justified it have been resolved. Adding senior approvers or requiring broader consultation can make matters worse, because each additional participant creates another opportunity to reopen the discussion.

For executives, the question is which responsibilities improve a particular decision. Some people need to develop the recommendation, others must contribute expertise, and a smaller group may need to verify specific requirements. Someone must make the final choice, and someone must carry it out. Decision speed and quality depend on assigning those responsibilities deliberately.

Why more participation can produce worse decisions

A media company encountered this problem when it gave every editorial unit responsibility for developing content recommendations. Each unit could advance its priorities, but bottlenecks emerged as different groups proposed competing approaches. With several teams occupying the same recommendation role, nobody had clear responsibility for combining their proposals into a decision that could move forward.

The company resolved the bottleneck by assigning one person to integrate the editorial units’ contributions, establish priorities, and recommend trade-offs. Editorial teams continued providing expertise and explaining their concerns. Responsibility for developing the recommendation now rested with one person, who could draw on the information needed to make it.

Clear recommendation ownership matters because competing proposals require someone to evaluate the alternatives. When several departments independently determine what should be recommended, disagreements can remain unresolved until they reach the final decision-maker. One accountable recommender can address those disagreements earlier and present a coherent proposal, leaving the final authority to concentrate on the remaining trade-offs.

A consumer-products company experienced a related problem at the executive level. It introduced a process intended to focus senior leadership attention on major product innovations but failed to specify which decisions required executive involvement. Consequently, senior leaders found themselves debating details such as the height of lettering on product packaging. Their participation followed the process, yet it consumed attention intended for more consequential innovation decisions.

Senior involvement can also change how managers behave. When executives routinely participate in operational details, managers may wait for their opinions before proceeding. Decisions slow down, and assigning responsibility to employees closer to the work becomes less useful. Consensus-driven organizations face a similar problem when meetings include dozens of participants whose contributions have little bearing on the eventual choice.

But limiting participation creates a different risk when it excludes essential expertise. At a pharmaceutical company, drug-development decisions incorporated scientific and medical expertise. Those perspectives helped determine whether a drug could be developed and whether it offered clinical value. However, the process overlooked whether insurers would cover the resulting treatment. The company consequently invested in pharmaceuticals with poor prospects for insurance reimbursement.

The missing information concerned commercial viability. People who understood payers’ willingness to provide coverage could have explained how reimbursement prospects affected the investment decision. Additional scientific review would have left that question unanswered because the necessary expertise came from a different function.

The three companies had different problems with decision responsibilities. The media company distributed recommendation ownership across too many groups. The consumer-products company involved executives in decisions that could be handled elsewhere, while the pharmaceutical company excluded information that could materially change the preferred course of action.

Addressing these problems requires separating the work involved in making a decision. Developing a recommendation, contributing expertise, verifying a requirement, choosing an option, and implementing it are distinct responsibilities. Each needs a defined role and an appropriate level of authority.

RAPID assigns five responsibilities to every decision

RAPID is a framework for assigning five decision responsibilities: Recommend, Agree, Perform, Input, and Decide.

The letters identify responsibilities, although their order does not prescribe the sequence of activities. Developing recommendations, consulting specialists, and conducting required reviews often happen together. Implementation considerations should shape the proposal before final approval, while the final decision establishes what the organization will do.

The framework separates preparing a decision from approving it. A senior executive may make the final choice, while another person coordinates the analysis across departments. The executive needs a recommendation that explains the alternatives and their consequences, so responsibility for preparing it must be explicit.

These roles can overlap. Someone developing a recommendation may also provide specialist input, participate in implementation, or hold final decision authority. What matters is that everyone knows which responsibility each person carries and where that authority ends.

Recommend: Give the proposal one accountable owner

The Recommend responsibility, or R, belongs to the person accountable for developing a proposed course of action. The recommender gathers information, evaluates alternatives, consults relevant participants, and combines competing perspectives. The proposal should explain its advantages, disadvantages, and trade-offs so the final decision-maker can make an informed choice.

Preparing that proposal can take considerably more work than approving it. In the RAPID approach, the recommender is described as typically performing 80% of the work associated with a decision. The figure shows how much responsibility rests with the person coordinating the analysis and preparing the proposal.

A mining company illustrates what happens when nobody clearly owns that work. After expanding into a new geographic region, it needed an external-relations organization to support its mine. Corporate headquarters wanted the organization established and assumed the regional vice president would recommend its structure. The regional vice president expected headquarters to develop the proposal because headquarters had initiated the idea.

Neither side prepared the recommendation, so the decision stalled. Progress became possible only after the executives clarified who held the R responsibility. The company had people who could approve an organizational structure, but nobody had accepted responsibility for investigating the requirements, evaluating possible arrangements, and proposing one.

RAPID therefore recommends assigning one R to each decision. That person can delegate analytical tasks, consult many departments, and gather competing proposals. But one individual remains responsible for combining the work into a recommendation, as the media company achieved when it brought its editorial proposals under one owner.

Ownership also requires an agreed process. Early in the decision, the R should meet with the final decision-maker to define the problem, establish the criteria, and clarify what the recommendation must address. They should agree on deadlines, required information, and the depth of analysis the decision warrants.

Those expectations guide the next stage. The recommender identifies people with relevant expertise, functions responsible for verifying requirements, and teams affected by implementation. The R then organizes meetings and reviews, evaluates alternatives, resolves differences where possible, and prepares the proposal for final consideration.

A large technology company gives recommenders a checklist to guide this preparation. Two questions capture the central challenge:

• “Who should be consulted to develop a complete recommendation?”
• “What is the right trade-off between rigor and speed for the recommendation?”

The first question helps prevent gaps in expertise. The second connects the amount of analysis to the consequences of the decision. A high-risk commitment may justify extensive investigation, while an easily reversible choice may need a shorter process. The recommender must establish that difference before organizing the work.

The R can hold any suitable position in the organization and need not report to the final decision-maker. Formal hierarchy determines some organizational authority, but recommendation ownership depends on who can coordinate the necessary work and produce a useful proposal.

Input: Bring the necessary expertise into the recommendation

Once the recommendation has an owner, the next question is whose knowledge that person needs. RAPID calls this responsibility Input, or I. Input providers contribute information, experience, analysis, and operational perspectives that help the recommender evaluate available choices.

An engineering leader might explain the technical requirements of a proposed system change. A finance specialist could assess its financial implications, while employees responsible for operations could identify dependencies affecting implementation. Each contribution answers a different question the recommender needs to resolve.

The pharmaceutical company’s missing payer perspective shows why Input selection must follow the decision’s requirements. Scientists and medical professionals could assess development and clinical value, but insurance reimbursement required commercial knowledge. The recommender needed that perspective early enough to influence the investment proposal.

Timing and quality matter because late contributions can force teams to repeat completed analysis. Input providers should offer credible information, clear reasoning, and useful analysis within the agreed schedule. The recommender must consider their contributions and reconcile competing views while retaining responsibility for the final recommendation.

This arrangement allows employees affected by a decision to contribute without acquiring approval authority. Someone responsible for implementation may also provide Input because practical experience can reveal constraints during planning. Holding both I and Perform responsibilities can improve execution, provided the advisory role remains clearly defined.

Agree: Check the requirements that a decision must satisfy

Some proposals must satisfy legal obligations, regulatory requirements, safety standards, financial constraints, or other established conditions. RAPID assigns responsibility for checking these requirements to Agree, or A. An Agreement holder examines a specific condition and works with the recommender to resolve problems before the proposal proceeds.

For example, a safety executive might confirm that a proposed change in working practices meets safety requirements. A brand manager could assess whether an action protects the company’s brand. Finance and risk-management functions may verify compliance with budget limits or the organization’s accepted level of risk.

Agreement differs from Input because an Agreement holder is responsible for a defined requirement. An Input provider offers advice that informs the recommendation. An Agreement holder identifies conditions that must be addressed, which means an unresolved objection requires a formal response.

The R and A should therefore work together while developing the recommendation. The recommender may revise the proposal, provide additional analysis, or explore alternatives to satisfy the requirement. Early cooperation reduces the chance that a late objection will force the organization to reconsider completed work.

An essential Agreement holder must also be involved before the recommendation reaches the final decision-maker. Otherwise, a proposal may advance without meeting a necessary condition. The same risk arises when an Agreement holder waits until after approval to raise concerns that could have been addressed earlier.

Agreement authority has defined limits. Under RAPID, an A does not possess unconditional power to block, reverse, or reject a decision. If the recommender and Agreement holder cannot resolve a disagreement, the recommender escalates the concern to the final decision-maker.

The D then considers the objection alongside the recommendation and determines whether to proceed, revise the proposed action, or accept an identified risk within the organization’s authority. Legal and regulatory obligations still apply. Escalation clarifies who resolves the disagreement while preserving the requirements governing the decision.

Many decisions require no Agreement holder. Others may need one or two people responsible for clearly defined conditions. The number of A roles should reflect the requirements involved, since additional approval responsibilities increase the work needed to advance a proposal.

Decide: Establish one final authority

The Decide responsibility, or D, belongs to the person who makes the final choice and commits the organization to acting on it. RAPID generally assigns one D to each decision because competing final authorities can leave proposals unresolved. When nobody accepts final responsibility, a decision may stall even after the recommendation is complete.

The D evaluates the proposed action, considers its alternatives and consequences, resolves outstanding disagreements, and chooses the course of action. This person also needs enough authority to commit the funding, personnel, and operational resources required for implementation.

The responsibility begins before final approval. At the start, the D should work with the recommender to define the problem, establish decision criteria, clarify timing, identify necessary inputs, and agree on the expected analytical rigor. Early alignment gives the R a clear basis for preparing the proposal and judging when further investigation is useful.

When the recommendation reaches the D, the analysis should address the questions established at the beginning. The decision-maker can then focus on unresolved trade-offs and any disagreement involving an Agreement holder. Once the choice is made, the D communicates the decision, its reasoning, and the resources available to implement it.

Some decisions belong to collective bodies. In those cases, the group should establish its decision rule in advance. Members might agree to decide by majority vote or authorize the chair to make the final call. Without a clear rule, participants may disagree about whether a decision has been approved.

The D’s position within the organization also matters. Final authority needs enough perspective to understand consequences across the business and enough power to allocate resources. Yet decisions should remain as close to implementation as practical, because moving routine choices upward can overwhelm senior executives and separate authority from operational knowledge.

Perform: Assign responsibility for carrying out the decision

The fifth responsibility, Perform or P, identifies who is accountable for implementing the approved choice. The performer turns the decision into an operational plan, coordinates the necessary teams, manages dependencies, and ensures the work moves forward.

Implementation responsibility should be established when the decision is made, with prospective performers involved earlier whenever their knowledge can improve the recommendation. They may identify staffing requirements, facilities needs, systems integration work, or other constraints affecting feasibility.

A proposed organizational change, for example, may depend on resources and operational arrangements that executives overlook during discussions. Involving the performer before approval helps the organization understand those requirements and judge whether the proposed schedule is realistic.

Major initiatives also produce further decisions during execution. Teams may need to choose technical approaches, establish operating procedures, or resolve resource conflicts. Consequential follow-on choices can require their own RAPID assignments, with clear recommendation ownership and final authority.

The D and P may be the same person, but their responsibilities remain distinct. The D establishes the approved course of action and commits the necessary resources. The P organizes the work required to deliver the result, turning formal approval into execution.

Decision authority needs clear boundaries

Assigning RAPID roles can create bureaucracy if their scope remains undefined. Every additional formal participant introduces another potential review, and a narrow approval responsibility can gradually expand into influence over the entire proposal. A process may appear accountable on paper while continuing to produce delays and competing authorities.

A medical-device company encountered this problem with its marketing brochures. Its regulatory compliance team had to approve materials before publication to ensure the brochures met applicable regulations. Over time, however, the team’s influence expanded into aesthetic decisions, including the colors used in the brochures.

Those interventions slowed production because regulatory reviewers were exercising authority over design choices. The company responded by narrowing the team’s Agreement responsibility. Reviewers remained accountable for ensuring that brochure text complied with federal regulations, while their approval rights no longer extended to unrelated aesthetic decisions.

The adjustment preserved the compliance team’s essential role while making its authority more precise. Regulatory expertise remained necessary, but reviews now focused on the conditions that justified the team’s involvement. The same principle applies to finance, safety, and brand reviews: each function needs authority over the requirements it is responsible for checking.

Approval rights should follow the risks involved

The retail bank applied this principle by defining which marketing decisions could proceed under established guidelines. Finance and risk management had already determined the conditions they considered acceptable, so managers could act independently within those limits. A new offer or customer population that fell outside them triggered additional review.

The guidelines changed when specialist oversight occurred. Previously, finance and risk management reviewed every individual mailing. Under the revised arrangement, they reviewed decisions that introduced different assumptions or risks. Managers gained authority over routine activity while specialist functions retained responsibility for exceptions.

This works because requirements can sometimes be expressed as operating rules. Once an organization establishes acceptable limits, employees can make repeated decisions within them without reopening the original assessment. The quality of those rules matters because managers need to recognize when a proposed action exceeds their authority.

Clear limits also prevent seniority from becoming a reason for unnecessary participation. The consumer-products company’s packaging discussions showed how executive attention could be absorbed by operational details. The organization needed criteria for identifying decisions that genuinely required senior judgment.

An executive may offer valuable experience or a perspective worth considering. That contribution can justify an Input role, while formal Agreement authority should follow a specific requirement. Assigning an A because someone “should have a chance to weigh in” gives an advisory contribution the procedural weight of mandatory review.

The consequences go beyond scheduling. Agreement holders can raise concerns that the recommender must formally address, and unresolved disagreements can reach the D. Input providers influence the recommendation through the quality of their analysis, while the recommender remains responsible for combining their views.

Confusing these responsibilities can allow Agreement holders to intervene after approval or treat their authority as an unconditional veto. Organizations therefore need to define what each A reviews, when that review occurs, and how unresolved concerns reach the D.

Narrowing authority can also shift influence within the organization. Employees who previously held Agreement responsibilities may move to Input roles, and someone accustomed to signing off on proposals may experience that change as a demotion. The concern is understandable because approval rights can carry organizational status and practical influence.

Managers need to explain how the revised responsibility works. An Input provider can materially shape a recommendation through timely analysis, specialized knowledge, and a clear account of risks. The organization must then enforce the boundaries consistently, including when senior employees request sign-offs outside their assigned responsibilities.

Consultation should reflect risk and reversibility

The next boundary concerns how much scrutiny a decision needs. Some choices have consequences that are expensive, difficult, or impossible to reverse. Others can be changed readily if the initial decision produces an unsatisfactory result. These more reversible choices are sometimes called “two-way door” decisions.

A two-way door decision generally allows lighter consultation because the organization can revise its approach as new information becomes available. The recommender still needs enough evidence to make a responsible proposal, but the cost of a mistake may justify moving faster.

Decisions involving major financial commitments, safety risks, regulatory obligations, or significant operational consequences require greater care. Consultation and review should reflect the potential harm of an error, the difficulty of correcting it, and the information needed to make an informed choice.

The R and D should establish the appropriate level of scrutiny early. Applying the same process to every decision can consume time better spent on consequential choices. Too little analysis of serious commitments can expose the organization to risks that better consultation would have revealed.

RAPID depends on the wider decision process

Role assignments work best as part of a broader decision process. Organizations need to identify which decisions deserve formal attention, establish who participates, define procedures, agree on timing, and set expectations for behavior.

Five elements capture these requirements:

The What element determines where formal coordination is worth the effort. Leaders need to identify consequential decisions and distinguish them from routine choices employees can make within established rules. Without that distinction, elaborate procedures can spread into work requiring little consultation.

Who assigns accountability, while How and When establish the practical conditions for exercising it. Participants need to know which analysis is required, when information must arrive, and how reviews fit into the schedule. Those expectations help prevent late interventions and repeated discussion.

Behaviors determine whether the process works in practice. Recommenders must seek relevant expertise, Input providers must contribute promptly, Agreement holders must respect the limits of their reviews, and decision-makers must bring choices to a conclusion. Clear assignments lose their value when participants keep reopening settled issues or demanding authority beyond their responsibilities.

AI can improve decision input while people retain accountability

Artificial intelligence makes the separation between analysis and decision authority increasingly important. AI systems can process large amounts of information, identify patterns, combine data from different sources, and generate forecasts. These capabilities help organizations evaluate alternatives and estimate possible outcomes.

Within RAPID, this contribution fits primarily into the Input responsibility. An AI system may generate information that improves a recommendation, while the organization assigns people responsibility for evaluating that information and determining the resulting action.

Consider an illustrative inventory-planning decision at a retailer. Buyers need to determine how much stock to order for an upcoming period. An AI system analyzes demand information and produces a forecast, helping buyers estimate required inventory and compare possible purchasing quantities.

The forecast provides an analytical basis for the decision, but buyers may know something that changes its interpretation. Suppose a competitor is preparing a promotion that could affect customer demand. A buyer aware of the promotion may conclude that the AI-generated forecast needs adjustment.

The buyer then evaluates the forecast alongside the expected competitor activity and decides which inventory quantity is appropriate. The AI contributes analysis, while the person responsible for the decision determines how that analysis should influence the organization’s commitment.

This division of responsibilities also matters when decisions involve ethical considerations, conflicting organizational objectives, or consequences requiring someone to accept accountability. The organization needs to establish who can authorize an action and who must answer for its results, even when software performs much of the analysis.

Automation can extend beyond analysis. A system may also carry out parts of an approved process, so organizations must distinguish automated operations from accountability for the rules governing them. Someone must determine what the system is permitted to do and when specialist review or human intervention is required.

For technology leaders, RAPID offers a way to make those responsibilities explicit. When an AI system produces a forecast, recommendation, or risk assessment, the organization should establish who evaluates the output, what additional information must be considered, and who can authorize the resulting action.

Legal, financial, safety, and operational constraints may also require defined review responsibilities. Those requirements should be established before AI-generated recommendations enter operational workflows. Teams can then identify which decisions may proceed within approved conditions and which need additional attention.

The same question remains as AI capabilities improve. More powerful systems can increase the amount and quality of available information, but organizations still need to determine how that information affects decisions and who carries responsibility for the consequences.

Execution ownership determines whether decisions deliver results

A beverage company decided to relocate its IT center of excellence to a European city. The center of excellence was a specialized organizational unit responsible for coordinating IT expertise and practices. The company approved the relocation, establishing where the new center would operate.

But nobody was made responsible for implementing the move. Essential preparations remained unfinished because no one took charge of finding office space or determining how existing IT operations would be consolidated with the new center. The company had selected a destination without assigning ownership of the work required to operate there.

The consequences became clear when a new chief information officer arrived. The new CIO reopened the relocation decision, forcing the organization to reconsider a choice it had already made. In effect, the company had to make the same decision twice.

The failure differed from the retail bank’s repeated approvals. At the bank, unnecessary reviews delayed routine decisions before they could proceed. At the beverage company, the formal choice had been made, but implementation had no accountable owner. The decision remained vulnerable because nobody had turned approval into coordinated action.

A clearly assigned P would have been responsible for planning the relocation, identifying dependencies, coordinating teams, and addressing operational requirements. For this project, that meant arranging facilities and determining how existing IT operations would connect with the new center.

Assigning the performer during decision development would also have improved the recommendation. The person responsible for execution could have assessed the proposed schedule, identified required resources, and explained the operational consequences of the move. These considerations might have changed how the decision-maker evaluated the available options.

Once the relocation began, further choices would have emerged. Facilities, staffing, systems, and operating procedures could each require separate decisions. Consequential choices might need their own recommender, relevant Input and Agreement holders, a final decision-maker, and someone responsible for implementation.

A major organizational decision becomes executable when these responsibilities extend into the work itself. The D authorizes the direction and commits resources, while the P coordinates the activities needed to deliver it. As the project develops, follow-on decisions receive their own assignments, preventing unresolved operational questions from accumulating.

For executives, the practical test comes before final approval. The organization should be able to identify who will deliver the result, what that person must coordinate, and which decisions may arise during implementation. In the beverage company’s case, answering those questions would have exposed the missing facilities and consolidation work while the relocation was still being planned.

The bottom line

Better decisions do not necessarily require more people, more analysis, or additional layers of approval. They require clear responsibilities, access to relevant expertise, and authority assigned to the people best positioned to act. When those elements are missing, organizations can spend more time reviewing decisions than making progress on them.

RAPID helps executives distinguish between contributing to a decision and having authority over it. The framework gives recommendations a clear owner, ensures that necessary expertise informs the choice, limits approval rights to defined requirements, and establishes accountability for both the final decision and its execution.

For business leaders, the challenge is to apply this structure without creating another layer of bureaucracy. Routine decisions should move within established guidelines, while consequential choices receive the scrutiny their risks justify. Senior executives should focus their attention where their judgment and authority make a meaningful difference.

As AI becomes more involved in business decisions, these distinctions will become increasingly important. Organizations may automate analysis and parts of execution, but they still need clear rules governing who evaluates recommendations, authorizes actions, and remains accountable for outcomes.

The measure of an effective decision process is not how many people participate or how thoroughly every proposal is reviewed. It is whether the organization can make informed choices at the appropriate level, resolve disagreements without unnecessary delays, and turn decisions into results.