A decision log template for product managers, founders, and teams that need to remember why

Teams lose time when old decisions become folklore. One person remembers a technical constraint, another remembers a customer request, and a third remembers that leadership simply preferred one option. The next debate begins without a shared record of what was actually known.

A decision log is a chronological index of important choices. Each record captures the decision in the context of its time. It is not meeting minutes, a backlog, or a polished strategy document. Its job is to make a consequential choice explainable and reviewable.

What belongs in a decision log

Log choices that are expensive to reverse, affect many people, create a meaningful constraint, settle a repeated debate, or deserve a deliberate review. Examples include pricing changes, target-customer choices, architecture direction, a new distribution channel, a major vendor, a product deprecation, or a policy that changes customer experience.

Do not log every ordinary action. “Move Tuesday’s meeting” is coordination. “Stop serving a customer segment so the company can focus its product and support model” is a decision worth preserving.

The seven-field decision record

The revisit trigger is the field most logs miss. Without it, a decision either becomes immortal or gets reopened whenever someone new joins the conversation. A trigger can be quantitative (“support volume exceeds 40 tickets per week”), temporal (“review after the first 90 days”), or evidential (“three enterprise prospects require SSO before evaluation”).

# Decision: [short statement]
Date:
Owner:
Status: proposed / decided / superseded
Confidence: 0–100%

## Question
What choice are we making?

## Context and constraints
What must remain true? What deadline or limit matters?

## Evidence available now
Customer observations:
Quantitative evidence:
Known unknowns:

## Options considered
A. Strongest case / cost / risk
B. Strongest case / cost / risk
C. Do nothing / delay

## Decision and rationale
We will... because...

## Expected outcome
What do we expect to observe, and by when?

## Revisit trigger
Reopen this decision if:

Use a table as the index, not as the whole decision

A decision-log table can track title, area, owner, status, date, confidence, revisit date, and the link to the full record. Keep the nuanced reasoning on the linked page. Trying to squeeze trade-offs into one cell produces a spreadsheet that looks organized and explains nothing.

Keep customer evidence separate from interpretation

Customer-discovery notes should preserve the customer’s language, the situation they described, what they already tried, and what happened. A statement such as “I export a spreadsheet every Friday because the dashboard cannot be filtered by region” is evidence. “They want our analytics add-on” is an interpretation.

Store the raw note separately, then link it from the decision record. Look for patterns across interviews without pretending a count of anecdotes is automatically a market measurement. Justin Wilcox’s Techstars session below is useful because it focuses on what to ask during discovery rather than how to pitch an answer.

Write alternatives fairly enough to survive disagreement

If an option is described only by its weaknesses, the decision was already made before the record began. Ask the advocate for each serious option to state its strongest case, its cost, and the condition under which it would win. This reduces performative consensus and gives a later reviewer a meaningful comparison.

Include “do nothing” or “delay” when it is real. Every initiative competes with another use of money, attention, or technical capacity.

Review decisions without rewriting history

• Do not edit away the original uncertainty. Add an outcome note or create a superseding record.
• Review the process before the result. Was the question clear? Were the alternatives real? What evidence was missing but obtainable?
• Compare expectations with observations. A forecast does not need to be perfect to reveal whether confidence was calibrated.
• Close the loop. If a trigger occurs, assign an owner and date for the review rather than letting the log become passive history.

Mind Marshal can hold full decision pages and index them in a database with table, board, calendar, and timeline views. It works well as a private preparation and reasoning space. Shared decisions still belong in the team system where affected people can read, challenge, and rely on them.