---
title: How to record a decision
url: https://tideswell.xyz/docs/decisions/record-a-decision
description: Record what was decided, why, and what you rejected, on the purchase order, supplier or record it affects.
---

> Documentation index: fetch https://tideswell.xyz/llms.txt to discover every page before exploring further.

# How to record a decision



A decision is recorded **on the thing it affects**, so it stays where the next person will look.
The dialog is called **Record a decision** and it takes about a minute.

## Before you start [#before-you-start]

You need the right to record decisions on the record in question, and the **Decisions** module
has to be on for your workspace.

## Where can you record one? [#where-can-you-record-one]

From the record itself: a purchase order, a supplier, a wholesale account, a wholesale order, or
a custom record.

You can also record one straight from a chat message, which is often where the decision actually
happened. When you do, you pick what it affects from a list: Purchase Order, Supplier, Campaign,
Task, Wholesale Account or Wholesale Order.

That last route is the one worth knowing. Decisions get made in conversation and then evaporate,
and capturing one from the message that settled it is the cheapest moment there will ever be.

## What do you fill in? [#what-do-you-fill-in]

Four things.

* **What was decided.** One plain sentence.
* **Why**, which is optional and is the whole point. Fill it in.
* **What we didn't do**, also optional, also the point.
* **When this was decided**, since that is often before you got round to writing it down.

## How should you write each one? [#how-should-you-write-each-one]

**What was decided** should read as a sentence somebody could act on: "Moved the merino base
fabric from Portside Mills to Kaimen Knitwear from the autumn order."

**Why** should carry the facts that drove it, not the feeling: "Portside's lead time went from
45 to 70 days across two revisions, and their on-time rate is 61% of 18 orders."

**What we didn't do** is where the value hides: "Did not renegotiate with Portside, because the
lead time was the problem rather than the price."

Without that last line, somebody next year reads the decision and asks why nobody tried
negotiating. With it, they know it was considered.

## When is it worth recording one? [#when-is-it-worth-recording-one]

When you would be annoyed to find the reasoning missing in a year.

* Changing supplier, or agreeing new terms.
* Placing an unusually large order, or deliberately not placing one.
* Giving a wholesale account terms you would not normally give.
* Deciding to write stock off rather than discount it.

Not for routine work. A decision log that records everything is as useless as one that records
nothing.

## What happens after you record it? [#what-happens-after-you-record-it]

It appears on that record, and in the workspace decision log, per
<Ref to="decisions/decision-log" />.

The next step, on decisions where you can say what you expect to happen, is committing to an
outcome so it gets measured. See <Ref to="decisions/outcomes" />.

## Where does the analyst come in? [#where-does-the-analyst-come-in]

It can read your decisions. "What did we decide about Portside Mills last season, and did it
work" is a question it can answer, alongside what was measured.

It needs the decisions view permission to do so, and it will not claim the decision caused the
result. See <Ref to="analyst/stock-suppliers-and-marketing" />.
