What is a Decision Room?

Most companies make their most expensive decisions in their least governed place: a meeting. Slides are prepared, opinions are exchanged, someone decides, and the reasoning evaporates. Six months later, the business remembers the outcome, but not the reasoning. Nobody can say what was known at the time, which assumptions proved false, or why one option was chosen over another.

A Decision Room is the opposite of that.

It doesn’t replace human judgement. It ensures human judgement has complete context.

The anatomy

A Decision Room holds five things, permanently:

• An objective. The thing being decided: commit the expansion, recover the margin, renew the contract. Stated, owned, dated.
• Options, ranked by reality, not optimism. Every option carries the earliest date it clears every binding constraint, not the date someone hopes for. An option that commits to seats the market cannot staff is marked blocked, and the room says which constraint binds.
• Evidence, not opinions. Each business function publishes facts with their evidence state, measured, modelled, inferred, stated or unmeasured. The room shows which is which.
• The decision, and the override. A human always decides. If they overrule the recommendation, the room records who, when and against what evidence. Overrides are allowed, and remembered.
• An outcome, scored. When reality lands, it is compared with what was recommended and what was decided. That score becomes institutional memory.

A meeting is an event. A Decision Room is an object.

One concrete example

No customer required. This is the shape of the thing:

• Revenue commits to a 300-seat expansion.
• Operations believes mobilisation is achievable.
• Technology shows the CRM integration won’t complete until 19 September.
• Hiring shows only 60% of the required German speakers will be available by the requested date.
• Finance confirms the option as sold falls below the minimum margin.

Rather than producing another dashboard, a Decision Room opens with four options (commit as sold, delay, split the ramp, reprofile the supply), each ranked by the earliest date it clears every constraint, each showing exactly which constraint binds it. The executive still decides. They just decide seeing everything.

Why rooms open themselves

The most useful property of a Decision Room is that nobody has to ask for one. Detection runs daily across every operating function, looking for compound conditions, a contractual go-live that precedes platform readiness, demand being created ahead of the team that would serve it, an account renewing on eroded economics. No single function can see these conditions in isolation. Only when Revenue, Operations, Finance, Technology, People and Legal publish evidence into one shared model does the condition become visible at all. When it warrants a decision rather than a dashboard, a room opens with the options already priced.

Three kinds of room

Pattern rooms open from an Intelligence signal that crossed a threshold: delivery confidence below the line, a contract breach accumulating, a margin gap widening. The room shows what Legal, Finance, Operations, People and Account Management each say about the question (Clear, Watch, or Risk) and presents options with a recommended path.

Commitment roomsare published by Finance when a commercial commitment is about to be locked. The financial picture is stated in plain language (revenue, margin, mobilisation cash, Finance’s verdict), module gates show whether the commitment is safe to proceed, and the full Finance workings sit one click away.

Readiness rooms check the gates before a commitment is approved. STOP gates must be resolved, RISK gates need an owner, and the days-to-go-live countdown runs live, with every gate deep-linking to the module that owns it.

The full walkthrough of all three lives on the Decision Rooms page.

Why deciding is permanent

When a decision is made, it locks: decision, reason, person, moment. It resolves the signal that opened the room, and commitment calls mirror into the Finance Decision Ledger. After 14 days the outcome can be classified: correct, correct override, wrong call, or not enough information. That classification feeds the learning spine, so the platform gets better at knowing which warnings to trust.

This is not just an audit trail. It is how the firm gets better at deciding.

What changes when decisions have rooms

Three things. First, decisions get faster, because the option analysis is prepared by the system, not by a week of slide-building. Second, decisions get honester, because every option shows the constraint that binds it. Third, and this is the compounding part, the organisation starts learning, because every outcome is scored against every recommendation. The second hundred decisions are better informed than the first hundred.

A meeting can still happen. It just happens inside a structure that remembers.