Every recurring decision needs one named owner, in writing

By Meet Patel · 2026-10-03 · 6 min read

Bain & Company surveyed more than 350 organizations about how they make decisions and found that only about 15 percent practice effective decision making. Paul Rogers and Marcia Blenko, the Bain consultants behind the RAPID framework, traced the most common bottleneck to ambiguity over who gets to decide what, in “Who Has the D?” (Harvard Business Review, January 2006). A company of 30 people can see the same pattern in a single pricing exception.

What a missing owner looks like on a Tuesday

Take a hypothetical 30-person SaaS company where a prospect asks for a 25 percent discount on a 12-month contract worth $48,000. The account executive believes the head of sales approves discounts. The head of sales believes anything above 15 percent goes to the CEO. The CEO believes pricing belongs to product, and product believes margin belongs to finance. Four people each hold a plausible claim to part of the decision, and nobody holds all of it.

The prospect waits five days while the question moves between inboxes. The discount that finally gets approved is 20 percent, which costs $9,600 in first-year revenue, and the approval comes from whoever was least busy that week. The story contains no difficult person and no slow process. The only thing missing is a written answer to the question of who decides.

The mechanism is arithmetic. A group of n people has n(n-1)/2 possible pairs, so 30 people have 435 and 50 people have 1,225. Each pair is a place where an unowned question can bounce. With 8 people the bouncing is cheap, because everyone knows who would decide. At 30 it becomes expensive, and the cost is paid in days of delay that never appear in a plan. This is the territory of the decisions your company is not making, which tend to be the expensive ones.

Two frameworks, one shared core

RAPID names five roles: recommend, agree, perform, input and decide. Rogers and Blenko say the roles are not carried out in that order and that they took liberties to get a usable acronym. The role that carries the framework is the D, a single point of accountability who commits the organization to action. The framework also contains a useful rule about vetoes: anyone who exercises one must offer an alternative or escalate to the person with the D (Bain, “Who Has the D?”).

DACI, which Atlassian documents in its Team Playbook, uses four roles: Driver, Approver, Contributors and Informed. The Driver gets the decision made by the agreed date. The Approver is, in the playbook’s words, “the one person (yes: one!) who makes the decision.” Contributors advise without a vote, and the Informed hear the result afterwards.

The frameworks differ in two ways that matter at 20 to 50 people. RAPID gives a formal veto through the Agree role, which suits constraints that sit outside any one function, such as legal or cash. DACI separates the person who runs the process from the person who decides, which stops the decider from also doing all the chasing. I would borrow both: one approver per decision, a driver who owns the date, and a veto only where a named external constraint exists.

The one-owner rule

My rule for a decision-rights table has three clauses.

1. One name. The owner is a person. A team, a committee or a role with two holders is a group of people each assuming another will decide.
2. Consequence follows ownership. The owner is the person whose own numbers move when the decision is wrong. Discount authority belongs with whoever carries the revenue target and the margin floor, and if those are different people, the table needs a second row for the cases that cross the line.
3. Disagree, then commit. Contributors can object on the record before the date. After the date they carry out the decision. Jeff Bezos put the norm this way in his 2016 shareholder letter: “Look, I know we disagree on this but will you gamble with me on it? Disagree and commit?”

The owner also carries a duty that people forget: decide by the date. A decision that slips past its date has still been made, by default, and the default usually favors whoever benefits from nothing changing. How quickly the owner must move depends on how reversible the choice is, which is the argument of the sibling post on one-way and two-way door decisions.

Build the table in an afternoon

Building the table takes a list and two honest answers per row, with no workshop required.

1. List the recurring decisions. Pull them from the last quarter: approval requests in Slack, meetings titled “sign-off”, expense exceptions, hiring requests, release go or no-go calls. Aim for 15 to 25 rows.
2. Ask two people independently. For each row, ask two people who work near the decision who decides it, and write the answers down before comparing them.
3. Compute the ambiguity rate. Count rows where the two answers differ, plus rows where someone answered “I’m not sure”. Divide by total rows. Suppose 20 rows produce 7 disagreements and 3 shrugs: the ambiguity rate is 10 of 20, or 50 percent. That number is your baseline, and I would expect it to persuade a leadership meeting more than any framework does.
4. Assign and publish. Fill one line per row with owner, driver, who must be consulted, who holds a veto and over what, who is informed, and the decision deadline in days.

Here is how a few rows might read in the hypothetical 30-person company. These figures are illustrations, not recommendations.

• Discount up to 15 percent: owner, account executive. Veto: none. Deadline: same day.
• Discount from 15 to 30 percent: owner, head of sales. Finance has a veto only on the margin floor, and must propose an alternative price if it uses it. Deadline: 2 days.
• Vendor spend above $5,000 a year: owner, head of operations. Consulted: the budget holder. Deadline: 5 days.
• Production rollback during an incident: owner, the on-call engineer. Nobody has a veto. Deadline: immediate.
• Headcount inside the approved plan: owner, the function head. Informed: finance. Deadline: 3 days.

Each row has a person, a limit that tells you when the decision moves up a level, and a deadline measured in days.

Where decision-rights tables fail

Four failures are the ones I would check first when a table gets written and then ignored.

The first is an owner with no authority over the resources the decision spends. If the discount owner cannot see the margin floor, they will escalate every case, and the table collapses back into email. The second is a contributor list that keeps growing. A decision with eight consulted people has no owner in practice, because the owner is spending their week managing eight opinions. I would cap consulted roles at three and move everyone else to informed.

The third is a veto without an alternative. RAPID’s rule is a good one precisely because it forces the person blocking a decision to do some of the work of the decision. The fourth is no review date. Rights drift as the company changes, and a table written for 20 people will be wrong at 45. Put a quarterly review on the calendar and compute the ambiguity rate again, so you can see whether the table is working.

One more case deserves a line in the table itself: the owner leaving. Rights belong to the role, so write the role beside the name and update the name in one place when the person changes.

What the table does to speed

When the 30-person company above publishes its table, the discount question gets answered by the account executive in minutes for the common case, and by the head of sales within two days for the larger case. The CEO leaves the loop entirely for decisions that used to pass through them. That is a hypothetical, and actual numbers will vary, but the direction follows from the mechanism: fewer pairs of people need to talk before an action happens.

The principle I would keep from all of this is that decision rights are a property of the decision, set before the argument starts. A company can survive a disagreement about what to do, and it has a much harder time surviving a disagreement about who gets to say.

Written by Meet Patel — startup operator and growth strategist in Dubai.

Read on themeetpatel.com