Decision debt

Every organization accumulates decision debt: the compounding cost of choices that were never properly recorded.

The what survives: the contracts, the code, the policies, the reporting lines, all still there, all still in force. What evaporates is the why. Why this architecture and not the one that was nearly chosen? Why this vendor, at this price, on this contract length? Why the scope stopped exactly there and not one feature further?

It is called debt because it behaves like debt. You take it on for free, in the moment, by not writing something down. You pay it back later, with interest, every time someone has to reconstruct reasoning already done once.

Decision debt is what an organization owes its own future, and it accrues in two ways: calls nobody makes, and calls nobody records. The second causes the first. A hard call stays unmade because nobody can find solid footing to make it on, and solid footing is exactly what an unrecorded decision fails to leave behind. Unrecorded decisions are the principal; undone decisions are the interest. Deciding faster is an interest-only plan: it doesn't touch what's owed. Memolok pays down the principal. It will not make the hard call for you, but it makes sure the next one has something to stand on.

How it accrues

Nobody chooses to take on decision debt; it arrives quietly through ordinary, reasonable events:

• People move on. The person who held the context in their head changes teams, or leaves. Nothing was hidden, it was simply never recorded.
• The premise expires quietly. A choice made because of a headcount freeze, a contract term, or a library that had no alternative at the time. The constraint lifts. The choice stays, now resting on nothing.
• What got written down was the outcome, not the reasoning. A ticket says migrate to X. It does not say what else was on the table, what ruled the others out, or what everyone expected to gain.

What it costs

Five failure modes, and most organizations can name an example of each within about thirty seconds:

Cargo-culting. The choice gets copied forward into the next project because it is there, not because anyone has checked that it still fits. Its original justification is unavailable, so it is treated as a house style.

Paralysis. Nobody will touch it, because nobody knows what it was load-bearing for. The safest available action is to build around it, which is how systems and policies accrete. This is decision debt's other face: the hard call stays unmade because the fence around it can be sensed but not located.

Zombie revisitation. A decision that was made, explicit alternatives ruled out for explicit reasons, comes back from the dead years later with new people re-litigating it from scratch. Sometimes the option that lost the first time even wins this time, for reasons nobody can compare against the original round.

Phantom decision. An issue that was never decided, but that everyone treats as settled. There is no verdict to look up, because none was ever reached, only an unspoken assumption that somebody, somewhere, must have handled it. It resurfaces every planning cycle as though for the first time, because in a sense, it is.

Decision decay. The world the choice was made for changes. Because no premise was recorded, there is nothing to check the choice against, so nobody notices it has quietly stopped being right.

See the glossary for quick-reference definitions of all five.

Why a wiki does not clear it

Most organizations have somewhere that decisions could be written down. The gap is not storage; it is structure.

A document store can find you a page. It cannot tell you whether that page is still in force, what replaced it, who was accountable for it, what it deliberately left unsettled, or what was believed to be true at the time. Prose in a search index answers what does this say. Decision debt is paid down by answering is this still the decision, and what was it resting on.

That is a difference in kind, not in diligence. You can write beautifully and still be in debt.

It is not only a large-organization problem

A solo project accrues decision debt too. Context that is obvious today is gone in six months, and a single stakeholder is a team of past-me, present-me and future-me who cannot talk to each other.

Memolok exists to make this debt payable. See the Decision Intelligence hub for where this fits alongside Decision Automation and Decision Support, what a decision record is, and why AI agents need them.

Try Memolok today

Memolok is in early testing, and we're inviting people to try it now. Sign up today and start recording your decisions. A person reads that list, so nothing lands in your inbox automatically.