Who Owes You a Decision? The Case for a Deferred Decision Ledger
Photo by Photo by EFFYDESK on Unsplash on Unsplash
There is a specific kind of silence on a UK project that everyone recognises and almost nobody logs. It sounds like: "Let's park that for now." Or "we'll pick it up after half-term." Or the particularly British "leave it with me." None of these are refusals. Each one, in the moment, is entirely reasonable. Taken together over a nine-month programme, they are how a project ends up in February still waiting on a question first asked in June.
The decision itself is rarely the problem. The problem is that the deferral has no address.
The deferral is sensible — the invisibility isn't
I've sat in plenty of meetings where the right call was genuinely to wait: the landlord hadn't responded, the funding round wasn't closed, the product owner was on leave. Deferring was correct. What wasn't correct was that the open question lived only in a call, a thread and somebody's memory, with no date attached and no sense of what waiting was costing.
Most project tools handle tasks badly suited to this. A risk log is about things that might happen. An action log is about things somebody must do. A deferred decision is a third thing: a choice that is currently unmade, sitting between two parties, quietly accruing interest.
Four columns and nothing more
A deferred decision ledger doesn't need software. A shared page will do, as long as it has these four columns and no more:
| Decision | Owner | Trigger date | Cost of delay |
|---|---|---|---|
| Landlord consent for partition wall | Client FM lead | 12 Jan (partition lead time) | £600–£900/week in re-mobilisation |
| Brand sign-off on the printed pack | Client marketing director | 3 Mar (print slot) | £1,200 one-off reprint risk |
| Confirm whether phase two is in scope | Client sponsor | 28 Feb (supplier resourcing) | £400/week held capacity |
| Agree data migration approach | Supplier technical lead | 20 Jan (build start) | £0 until 20 Jan, then £1,100/week |
The decision is a question phrased as a question. Not "landlord consent" but "does the landlord consent to the partition, and if not, what's the fallback?" Precision here saves you a second round of emails.
The owner is one person. Not a team, not "the client side." One name, with the authority to actually settle it. If no single person can settle it, that's a governance problem worth naming in the ledger itself.
The trigger date is the heart of the thing. It isn't the deadline for the decision — it's the last date on which making the decision still avoids the cost. You set it by working backwards from the consequence, which is why it usually sits a fortnight or a month before the date everyone would instinctively write down.
The cost of delay is a band, not an accounting entry. "£600–£900 a week, mostly standing time on site" is honest and useful. False precision, down to the pound, invites an argument about the number instead of the decision.
What belongs in it — and what really doesn't
One test: if this stays open for four weeks, does something stall or get more expensive? If yes, it goes in. If no, it doesn't.
That single filter keeps the ledger at ten or twelve lines. Every ledger I've seen fail had forty items on it, half of which were because someone felt the register should be comprehensive. Comprehensiveness is what killed it. A ledger nobody reads on a Friday is just another document.
A worked example
A fit-out client I worked with had a partition wall question that ran for five weeks. The sequence was familiar. It came up on a Tuesday call, went to the client's facilities manager, who needed the landlord's view, who needed a drawing, which needed the designer, who had moved on to another job.
Once it went into the ledger, three things changed in a fortnight. The facilities manager was named, which meant she knew she was named. A trigger date of 12 January appeared, driven by a six-week partition lead time, which made the question urgent in November rather than January. And the cost of delay — roughly £700 a week in trades standing around drinking tea — sat in a column next to her name, visible to her director.
It was settled in eight days. Nobody was shouted at. The ledger did the persuading.
The ritual matters more than the artefact
A ledger reviewed once is a document. A ledger reviewed weekly is a control.
The pattern that works: ten minutes, same slot every week, one named keeper — usually the project manager or the account lead — and a short walk through every line. Anything that needs chasing gets chased that morning. Anything settled gets struck through with the date it was settled.
That strikethrough habit is worth a mention. Six months later, the ledger doubles as a record of what was decided and when, which is useful when someone asks why the second floor went a certain way. The decisions you made are as important to capture as the ones you're waiting on.
The bit people find uncomfortable
A ledger makes the cost of delay visible to the person who owns the decision, not only the person who bears it. In client–supplier work that's a mild, healthy shift in the room. It can also feel like scorekeeping if it's introduced badly.
The framing that lands: this is here to help you decide, not to record that you were slow. Which is why I'd keep the cost column factual and the tone flat. No red text, no exclamation marks, no naming people in the weekly update who haven't been spoken to first.
Keep it small
Ten lines, four columns, one owner per line, one ten-minute review a week. That's the whole method. The aim isn't that every decision gets made quickly — some genuinely shouldn't be. The aim is that nothing quietly goes missing, and that when you choose to wait, you know what the waiting is costing and who has agreed to pay it.
