Decision Log vs Decision Record: What Is the Difference?

A decision log is a chronological list of what was decided and when. A decision record is a structured, self-contained document capturing one decision in depth: the resolution, the reasoning, the alternatives rejected, who decided, and under what authority. The log is the index; records are the substance behind each entry.

The terms get used interchangeably, but conflating them is why so many "decision logs" turn out to be useless six months later. A list of one-line decisions tells you that something was decided without telling you anything you would actually need to act on it.

The difference in one sentence

A log captures that a decision happened; a record captures why it happened and what it means. The log is optimized for breadth and chronology, so you can scan everything decided in Q2. The record is optimized for depth and self-containment, so a reader who was not present can fully understand a single decision without asking anyone.

Side-by-side comparison

Dimension	Decision log	Decision record
Structure	Chronological list	Structured document
Answers	What and when	Why, who, authority, reversibility
Optimized for	Breadth, scanning	Depth, self-containment
Granularity	One line per decision	One document per decision
Failure mode	No reasoning preserved	No overview across decisions

What a decision log is good for

A decision log shines at orientation. New hires can scan it to see the major calls made over the last year. Leaders can review it to confirm nothing important was missed. It functions as a table of contents, an index into the organization's decision history, and that indexing role is genuinely valuable.

Its limitation is that a one-line entry strips out everything that makes a decision actionable later. "Adopted Postgres over MongoDB, March 2026" tells you the outcome and nothing about why, which means when someone proposes switching back, the log offers no defense.

What a decision record is good for

A decision record is where the reasoning lives. It states the alternatives considered, why the chosen path won, who made the call, and whether it is reversible, a distinction we explore in the context of a system of record for decisions. Because it is self-contained, it is the unit that prevents relitigation and the unit AI can ground an answer in.

The self-containment is the key property. A record that requires you to already know the backstory is just a log entry in disguise. A real record stands alone, which is what makes it durable when the people who wrote it are gone.

Why you need both

The log and the record are not competitors; they are layers. The log gives you the map; records give you the territory. Without the log, you cannot find the record you need. Without records, the log points to entries that have no content. A mature decision system of record provides both: a scannable index backed by deep, self-contained documents.

The trap most teams fall into is building a log and calling it done. They end up with a tidy chronological list that preserves none of the reasoning, which means they pay the full cost of undocumented decisions despite feeling organized. The list of decisions exists; the understanding does not.

Common Questions

Is a decision log the same as a decision record?

No. A decision log is a chronological list capturing what was decided and when. A decision record is a structured document capturing one decision in full, including why it was made, by whom, under what authority, and whether it can be reversed. The log indexes; the record explains.

Do I need both, or can I pick one?

You need both. They are layers, not alternatives. The log is the table of contents that lets you find decisions; the records are the chapters that contain the reasoning. A log without records preserves no "why"; records without a log are hard to navigate.

Why isn't a one-line decision log enough?

Because a one-line entry strips out the reasoning, the alternatives, and the authority. When someone later proposes reversing the decision, the log offers no defense because it never recorded why the original choice won. That missing context is exactly what causes teams to re-argue settled questions.

Which one should we build first?

Start with records for your highest-stakes decisions, then index them with a log. Reasoning is what is irrecoverable once people leave, so capturing depth on the decisions that matter most yields more value than a complete but shallow chronological list.