RAPID · Decision Authority Design
Recommend · Agree · Perform · Input · Decide
What is RAPID?
- Five roles: recommend, agree, perform, input, decide
- Exactly one decider, and as few agree roles as possible
- RACI divides tasks; RAPID divides decisions
What actually changes in an organization that applies RAPID is not how meetings are run but the unit of management. Before, you managed a forum called the 'weekly management meeting.' Afterward, each individual decision, such as 'approve the shipping price for the new product' or 'replace an outsourcing vendor,' becomes something to manage. Once a decision has a name and a person’s name beside it, you can see in a list which decisions have been stalled for weeks. It also becomes clear that a delayed decision was a design problem, not someone’s personality problem. That lets you ask afterward, not who made the decision, but who was supposed to make it.
Of the five roles, the line that is hardest to call in practice is the one between Agree (A) and Input (I). The test is simple. If the company would bear legal or financial liability when that person objects, they are A; if the outcome would merely be disappointing, they are I. That is why A is limited to functions where regulation or contracts are at stake, such as legal, finance, safety and quality, and when a single decision has more than three A roles, it has effectively reverted to a consensus body. There must be exactly one D, and the moment you want to write down two is a sign the decision should be split in two. Sequence is also often misunderstood: RAPID does not flow in the order R→A→P→I→D. In practice, R gathers I and drafts the proposal, coordinates with A, D decides, and P takes over; the letters are only a memory aid.
The most common point of failure in practice is assigning D by rank. Sending every decision right to the CEO or division head makes the chart look tidy, but it ends up identical to the existing approval line, and about two months later you hear, 'We made the role chart, so why is it still not working?' D belongs to the person who answers for the numbers resulting from the decision, not to whoever holds the highest rank. Another common misunderstanding is to treat R as 'the person who prepares the materials.' In fact, R gathers the input, persuades A and brings the proposal forward, so R often has the greatest influence in practice. Finally, leaving a role chart untouched for more than six months is a typical path to failure. Whenever there is a reorganization or a new business, roles should be reassigned at that point.
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum Consulting
Ara LimSpecialist in global leadership, AI, and business communicationA framework introduced by the global consulting firm Bain & Company to improve how organizations make decisions. Its name comes from the initials of the five roles.
Company A is an industrial parts manufacturer with 320 employees and annual revenue of about KRW 180 billion (≈ US$134,670,100). The ex-factory pricing decision for three new products had been delayed for four months, and the company lost the chance to supply two major customers. The pricing item was brought back to the weekly management meeting 11 weeks in a row and left the table each time without a conclusion. (Hypothetical example)
- The head of strategic planning went through the last six months of meeting minutes and pulled out 14 agenda items that had come up more than once. Nine of them were decisions about pricing, investment or staffing.
- On pricing, the head of sales said, 'We track market prices,' while the finance lead countered, 'Approving the margin rate is our call,' and it turned out that both had written themselves down as the decision-maker.
- The CEO named a single business unit head as the D for the shipping-price decision and kept only the finance lead, who protects the minimum margin, as the A. The head of sales and the head of quality were moved down to I, giving input only.
- The product planning manager, acting as R, submits the pricing proposal and cost basis every other Friday, and an internal document now states that D will reply with approval or rejection within 5 business days of receipt.
- Three months into operation, the average time to decide pricing items fell from 47 days to 9, and weekly management meeting agenda items dropped from 12 to 7. Of 14 resubmitted items, 11 were closed.