You send a draft, get approval, revise it, and then hear a familiar question: “Why are we doing it this way?” The answer may be buried in a meeting, a chat thread, or somebody’s memory. Soon, a settled choice is reopened—not because it was necessarily wrong, but because its reasoning has disappeared.
A small decision log solves this particular kind of workplace drag. It is not a project plan, a meeting transcript, or a place to defend every move. It is a short record of meaningful choices: what was decided, why, what was deliberately not chosen, and what would cause the team to revisit it.
Used well, it gives new collaborators a way to catch up, helps reviewers give relevant feedback, and makes changes feel intentional rather than chaotic.
What a decision log is—and what it is not
A decision log is one shared document or section of a project workspace containing brief entries for decisions that affect later work. Each entry should be understandable months later by someone who was not in the room.
The key distinction is between information and decisions. “The homepage needs new copy” is information. “We will lead with the customer problem rather than product features because first-time visitors lack context” is a decision.
Do not turn the log into a diary of everything that happened. A diary creates more reading than clarity. Record choices that narrow the path forward, such as:
- which audience or use case comes first;
- which scope items are postponed;
- which approval standard applies;
- which format, workflow, or owner the team selects;
- how competing feedback was resolved;
- what assumption the work depends on.
The point of a decision log is not to prove that the team was right. It is to make the reasoning visible enough that changing course becomes a conscious choice.
Start when the cost of forgetting is real
Not every task deserves documentation. If you are choosing a meeting time or making a minor wording adjustment, the overhead will outweigh the benefit.
A log earns its keep when a decision has at least one of these features:
- Several people contributed. No one person holds the full context.
- The work will last longer than the conversation. A campaign, rollout, redesign, recurring process, or multi-stage deliverable is likely to outlive a chat exchange.
- There were credible alternatives. Future reviewers may reasonably ask why another route was not taken.
- A tradeoff was accepted. For example, speed was prioritized over customization, or consistency over local flexibility.
- The choice will constrain later work. A decision about audience, file structure, measurement, or ownership can quietly shape dozens of later choices.
If your team frequently circles back to the same debates, that is a stronger signal than project size. A modest project with shifting stakeholders may need a log more than a large project with a stable, experienced team.
Use a format people can actually maintain
The best decision log is usually plain. Put it where the people doing the work already look: in the project brief, shared workspace, or a clearly named project document. A sophisticated system that requires a separate login and elaborate fields often becomes another abandoned tool.
Give each entry a date and a descriptive heading. Then capture five pieces of context.
| Field | What to write | Why it matters |
|---|---|---|
| Decision | The specific choice, stated as a completed action. | Prevents vague entries such as “discussed onboarding.” |
| Reason | The main evidence, constraint, or goal behind the choice. | Preserves the context that is otherwise lost first. |
| Alternatives considered | One or two realistic options and why they were not selected. | Stops old ideas from returning as though they are new. |
| Owner and date | Who confirmed the decision and when. | Shows that the entry reflects a real conclusion, not a stray suggestion. |
| Revisit trigger | The event or evidence that would justify reopening it. | Makes the decision firm without pretending it is permanent. |
Here is an illustrative entry:
Decision: Publish the employee guide as a searchable web page before creating a polished PDF.
Reason: The content is expected to change during the first few months, and staff need to locate individual answers quickly.
Alternatives considered: A PDF was deferred because each update would require redistribution and version control. A printed booklet was not needed for the current audience.
Owner and date: Operations lead, April 8.
Revisit trigger: Reconsider a PDF if the content stabilizes and a downloadable reference becomes a repeated request.
Notice what is missing: a transcript of everyone’s opinions. The entry records the conclusion and its useful context, not the full negotiation.
Write decisions in language that can be checked later
Weak entries tend to be too soft to guide anyone. “Focus on users,” “simplify the process,” or “make it more engaging” may sound agreeable, but they do not tell a colleague what was chosen.
Instead, make the object and the tradeoff concrete:
- Instead of “Keep the page simple,” write “Limit the first screen to one primary action so new users do not have to compare three equally prominent options.”
- Instead of “Use a flexible process,” write “Allow teams to choose their own planning tool, but require one shared weekly status update in the central workspace.”
- Instead of “Prioritize quality,” write “Delay the launch until the help content covers the three most common setup errors; cosmetic refinements can follow.”
This does not require false certainty. If the decision is provisional, say so. Words such as “for this phase,” “until the pilot is reviewed,” or “given the current deadline” are useful because they state the boundary honestly.
Capture the decision at the right moment
Trying to write polished documentation during every discussion can make people self-conscious and slow the meeting. Waiting until the end of the week is also risky: details blur, and a tentative idea can be remembered as a firm conclusion.
A simple sequence works better:
- During the discussion, identify the fork in the road. Say, “It sounds as though we are choosing between a broad launch and a smaller pilot.” Naming the choice helps the group decide rather than merely talk.
- Before the meeting ends, confirm the conclusion aloud. Ask, “Can I record that we will run the pilot first, with these two success conditions?” This gives people a chance to correct the summary.
- Write the entry soon afterward. Keep it short enough that this takes a few minutes, not an hour.
- Share the link or location with the affected people. A quiet link is often enough; for consequential choices, call out the decision in the next update.
If you are not the formal project owner, you can still use this approach without sounding territorial. Frame it as a service to the group: “I wrote down the decision and the condition for revisiting it so we do not lose the context. Please correct anything I missed.”
Make the log useful during feedback, not just after meetings
The highest-value moment for a decision log is often when feedback arrives late. A reviewer may object to a choice that was already considered. That does not mean their input should be dismissed. It means the discussion can start at the right level.
Try a response like: “We chose the shorter version because the first audience needs a quick overview, and we planned to revisit that if the detailed questions remain common. Does your feedback suggest that condition has been met, or is there another goal we should weigh?”
This response does three things: it acknowledges the feedback, supplies missing context, and leaves room for new information. Contrast that with “We already decided,” which may be factually true but usually ends the conversation without resolving the concern.
In review meetings, bring up the few entries relevant to the work being reviewed. Do not ask people to read an entire project archive. The log should reduce cognitive load, not create homework.
Set revisit triggers so the log does not become a museum
Teams sometimes resist documenting choices because they fear decisions will become rigid. The solution is not to avoid a record; it is to record the conditions under which the choice should change.
Good revisit triggers are observable. They might include:
- a pilot produces a repeated type of user confusion;
- a required dependency is unavailable by an agreed milestone;
- a new audience is added to the project;
- a change in scope removes the original constraint;
- feedback reveals that the team optimized for the wrong outcome.
“Revisit later” is not a trigger. It places a burden on memory and makes reconsideration unlikely. “Revisit after the first round of support questions is categorized” gives the team a recognizable moment to look again.
When you do reverse a decision, do not delete the old entry. Mark it as superseded and add a new entry explaining what changed. The older reasoning may still be useful, especially when someone asks why the team did not take the current route initially.
Avoid turning documentation into a political shield
A decision log can be misused as a way to win arguments, assign blame, or make colleagues afraid to revise a choice. Those uses damage the collaboration the log is meant to support.
Keep the tone factual and collective. Record “The team selected” or “The project owner confirmed” rather than “Jordan insisted.” Describe constraints without editorializing. If disagreement matters, state the substantive difference: “One concern was that a single template would not fit local teams; the group accepted that tradeoff for the initial version.”
Also avoid claiming a decision was final if the person with authority has not actually confirmed it. Label it “proposed” or “pending approval” until it is settled. A log should clarify authority, not quietly manufacture it.
Give the habit a modest rhythm
You do not need a documentation meeting. Add a two-minute check to an existing project rhythm: “What did we decide this week that someone might need to understand next month?” If the answer is nothing, write nothing.
For an active project, scan the log before a major review, handoff, or new teammate’s first meeting. Remove nothing essential, but make sure superseded decisions are clearly labeled. The goal is a short path to context, not an impressive archive.
Start with one live project that has already generated conflicting feedback. Create a heading called “Decisions,” add the most recent meaningful choice using the five fields, and send it to the people involved with one question: “Is this an accurate record of what we decided and when we would revisit it?” That small step is enough to make the next round of feedback more useful.