Choose a Decision Framework by Where Decisions Break

The request usually arrives after an offsite. A leadership team has spent two days agreeing that it decides too slowly, and someone proposes a decision making framework for the team to adopt: RAPID, DACI, a consent rule, a decision log. Within a few weeks the letters are on a slide, every agenda item carries a named decider, and the team feels it has dealt with the problem.

One adoption, two quarters later

One payments company we worked with did exactly this. The team adopted DACI, assigned a driver and an approver to every item, and ran it faithfully for two quarters. The meetings changed shape. The outcomes did not. The same pricing decision came back three times, each time with its approver clearly named, and each time reopened by a member who had not been in the room when it closed. The framework had answered a question the team was not failing on. Its decisions had always had owners. What they lacked was any rule for staying shut.

This is the ordinary result of choosing by reputation. A search for a decision making framework for executive teams returns roundups: five or seven models set side by side and compared on their acronyms. None of them asks which part of the team's decisions is breaking, and that is the only question that settles whether a framework will help.

The four break points

Followed from the moment it is raised to the moment the organisation acts on it, a decision can fail in four places. In the teams we see, almost every stalled or reversed decision traces to one of them:

• unowned, where the decision reaches the room and nobody is authorised to close it, so it waits for agreement
• under-prepared, where it closes on a single option, with costs nobody stated, and the work exposes what was missed
• unheld, where it closes properly and is reopened by whoever lost, or by whoever was absent
• uncarried, where it holds in the room and arrives downstream in several versions, or not at all

The break points are sequential, and a decision that fails at an early one never reaches the later ones. A team whose decisions have no owner will not yet see a reopening problem, because nothing is closed firmly enough to reopen.

What each model was built for

Read against these four, the familiar frameworks stop competing with one another. RAPID and DACI are ownership tools: they settle who closes. Option briefs, pre-mortems and the habit of stating what each option costs are preparation tools. A decision register with written reopening conditions is a holding tool, set out in how to stop a decision from being reopened. An agreed sentence and an outbound step at the end of the meeting are carrying tools. Compared this way, team decision making frameworks are not alternatives to one another but repairs for different faults. An ownership framework installed over a holding failure ends where the payments team's DACI did, correctly run and beside the point. A second comparison, on what each model costs to run across a quarter of decisions, is set out in team decision frameworks differ in what they cost to run.

Teams diagnose themselves badly in a workshop, because each member describes the failure that costs them personally. The commercial lead remembers the reversal. The technology lead remembers the decision that never closed. Both are real, and neither tells the team where most of its decisions break.

Tracing the last quarter

The more reliable method is a trace. We take the team's last quarter of minutes and list every decision it attempted, usually somewhere between fifteen and thirty. For each one we ask four questions in order: did someone close it, was it closed on more than one costed option, did it stay closed, and did the organisation act on the version that was agreed. The first question answered no is that decision's break point.

The distribution is the diagnosis

Laid out together, the results rarely spread evenly. Most teams have one dominant break point, and it accounts for most of their stalled decisions. A team that finds ownership failures everywhere needs a rights framework first and nothing else yet. A team whose decisions close cleanly and come back a month later needs a register, and would gain nothing from another round of decider assignments. The trace also shows where the team is stronger than it believed. That matters, because a framework installed over a working habit tends to replace it with paperwork.

The roundups imply that a framework is something chosen whole. In practice, the framework that holds in a leadership team is assembled: one rule for each break point the trace found, written on the same page, and nothing for the break points it did not find.

Four lines on one page

The fullest version, for a team that fails at every point, is four lines settled before an item reaches the agenda: who decides and by when, which options were considered and what each costs, what evidence would justify reopening it, and the sentence that will leave the room. Each line maps to one break. The first is the territory of the RAPID framework. The last is the territory of an outbound segment at the end of the executive weekly.

Most teams need two of the four lines, not all of them. Every line a team does not need is a line it will fill in without reading, and a framework filled in without reading teaches the team that the form is the decision. In a large company the page has a second reader as well, the committee above the team, and a practical decision framework for corporate leaders is written so that its output survives that forum too.

Every line a team does not need is a line it will fill in without reading.

When the break point moves

The break point moves once it is fixed. A team that installs clear ownership usually starts closing decisions faster, and some of those faster decisions are then reopened, because closing more of them exposes the holding problem the ownership problem had hidden. The trace is worth repeating after two quarters, and the page changes when the distribution does.

Our engagements on this begin with the trace rather than with a model. A quarter of minutes, sorted by break point, usually settles in one session what the offsite spent two days debating, and it gives the team a reason to adopt the rules it needs and decline the ones it does not. Where the trace points to ownership, the starting point is who decides what in a leadership team.

From there our consulting work inside the decision architecture pillar writes the one-page framework with the team and sets the date for the second trace, and our training programmes build the habits each line depends on, from preparing costed options to agreeing the sentence that leaves the room. A decision making framework for a leadership team is chosen by where its decisions break, not by which acronym the team has heard of.