What is decision memory?
Decision memory is a record of what an organization decided, what outcome it expected at the time, and what actually happened, kept together so the lesson can be reused later. The defining feature is the expected outcome: because it is written down before the result is known, the decision can be graded honestly afterwards rather than explained away.
It is a narrower idea than the term institutional memory, which usually covers everything an organization knows. Decision memory covers only decisions and their outcomes, which is what makes it possible to grade.
How it differs from writing things down
A wiki holds what someone wrote. A decision log holds what someone decided. Enterprise search finds either of them again. None of the three tells you whether the decision was right, because none of them recorded what was supposed to happen. Decision memory closes that loop: expectation first, result second, lesson third, and only then reuse.
The five steps
Record the decision
What is being decided, by whom, in what situation. This is the part most organizations already do, usually in a document or a meeting note.
Record what you expect to happen
Before the result is known. This is the step that makes everything after it meaningful, and the step almost nothing else asks for. An expectation written after the fact is not a prediction.
Grade it against what happened
When the outcome lands, compare it with the expectation. A decision that went badly is as useful as one that went well, provided the record survives it.
Turn it into a lesson a person approves
A draft lesson is reviewed and approved by someone other than the person who made the decision. Separating author from approver is what makes the lesson trustworthy to a colleague who was not in the room.
Reuse it, with the evidence attached
Later questions are answered from approved lessons, each citing the decisions and outcomes behind it, so the reader can check the reasoning instead of taking the answer on trust.
Why this record is different
A decision recorded with its expected outcome before the result is known, then graded against reality and approved by a person, cannot be scraped from the web and cannot be reconstructed after the fact. That makes decision memory more than documentation. It is an asset your organization accumulates and owns outright: reuse it, share it, sell it, and train your own models on it, always on your own terms.
Sharing lessons without sharing data
A lesson can be useful to an organization that did not live through it. Through the Almanexa Hub, an organization publishes approved lessons as signed packs, and another organization installs them. Only the approved lesson travels: the decisions, outcomes, documents and identities behind it stay where they were created. That is what makes sharing possible between organizations that would never share their underlying records.
Common questions
- How is decision memory different from institutional memory?
- Institutional memory usually means everything an organization knows, including processes, relationships and undocumented know-how. Decision memory is deliberately narrower: decisions, the outcome expected, the outcome that occurred, and the lesson drawn. That narrowness is the point, because it is what allows a decision to be graded rather than merely remembered.
- Is decision memory just a decision log?
- A decision log is the first step of it. What a log usually lacks is an expected outcome recorded before the result is known, a grade once the result arrives, and a rule that the person who made the decision is not the person who approves the lesson. Without those three, you have a record of what was decided but no way to know what was learned.
- Can decision memory be self-hosted?
- Yes. Almanexa's core is source-available under the Business Source License and runs inside your own environment, so decisions, outcomes and lessons stay where your other records are. You can read how approvals are enforced and how the record is sealed rather than taking it on trust.