Decision log: definition
A decision log is a single document that chronologically records a team's important decisions: what was decided, when, by whom, why, and based on which signals. Each entry fits in a few lines or one page. Its value comes from being consulted, not from being written: it lets anyone check, before a debate starts, whether the topic has already been settled.
Why keep a decision log
Teams decide constantly and forget most of it within months. The article Why startups forget their decisions explains the mechanism: decisions live in conversations, the reasoning never gets written down, and memory rests on individuals. A decision log is the simplest answer to that problem, available to any team today.
The concrete benefits: settled debates do not restart from scratch, discarded alternatives do not get re-proposed six months later, and new hires get access to the reasoning instead of bare conclusions.
What a decision log is not
- It is not meeting notes. Notes capture a discussion; the log captures a decision, with its rationale and the conditions under which it should be revisited.
- It is not a backlog or task tracker. The log records choices, not work to be done.
- It is not a full organizational memory. The log covers decisions; a company's memory also includes signals, experiments, lessons, and the ability to surface them.
The format descends from ADRs (Architecture Decision Records) in software engineering, extended to every structural decision: pricing, positioning, hiring, acquisition channels.
How to run one in practice
The minimal format for an entry: the decision in one declarative sentence, the date, the decision-maker, the context, the signals motivating the choice, the options considered with pros and cons, and the most valuable section of all: what would change our mind. The decision record template details each section and copies into Notion or Google Docs in minutes.
Three rules to keep it alive:
- One log, not scattered files: the value comes from consulting everything in one place.
- Write at decision time, not at the end of the sprint: ten minutes while it is fresh beat an hour of reconstruction later.
- Reread before re-deciding: if the topic was already settled, the discussion restarts from "what would change our mind," not from zero.
The limit of a manual log is discipline: someone has to remember to write, then remember to reread. That is what a living memory like Verbasil takes over, capturing decisions with their evidence and resurfacing them when the topic returns. For significant failures, the log pairs well with a post-mortem.
FAQ
Which decisions belong in a decision log?
Structural ones: decisions that commit the team beyond a few weeks or that are hard to reverse. A simple heuristic: if the topic is likely to be re-debated within six months, the decision deserves an entry.
What tool should we use for a decision log?
The tool matters less than the single-place rule: one location, sorted by date, known to the whole team. A Notion page or Google Doc is enough to start; the next step is tooling the retrieval side so past decisions come back on their own.