Your strategy deck is already stale

The deck took three weeks. It is genuinely good — the analysis is sound, the choices are argued, the numbers are sourced. It was presented on a Thursday, and by the following Thursday two of its inputs had changed and nobody noticed, because a deck has no way of noticing.

This is not a criticism of decks. It is a criticism of using a static artefact as the connective tissue between deciding and doing.

What a deck is good at

Worth saying first: decks are excellent at what they are for. They compress an argument, they force sequencing, and they are the right medium for persuading a room. A strategy that cannot be presented in twenty slides is usually a strategy that has not been decided.

The failure is downstream of presentation. The deck becomes the reference document, and a reference document that cannot update is a liability that grows with every week it stays authoritative.

The three couplings that break

Coupling to state. The deck's analysis rested on facts as of a date. Pipeline, capacity, burn, hiring, competitive position. None of those hold still. The deck asserts them permanently, in a fixed font, with no timestamp on the individual claims — so a reader in week six cannot tell which parts have moved.

Coupling to execution. After the deck, someone transcribes it into work. Tickets, projects, owners. That transcription is lossy in a specific way: the reasoning does not survive it. The ticket says what to do. It does not say which strategic bet it serves, or what assumption made it worth doing. So when the assumption changes, nothing connects the change to the work that depended on it.

Coupling to outcomes. The deck predicted things. Six months later, some of those predictions can be scored. Almost nobody scores them, because the deck has been superseded by a newer deck and the prediction was never registered anywhere a grading step could reach it.

Break all three couplings and you have an expensive artefact that describes a company that no longer exists, referenced by work that cannot explain itself.

Living strategy, concretely

"Living" gets used loosely, so here is the concrete version. A strategy is alive when three properties hold.

It reads current state rather than restating it. The analysis references the state model, so the figures behind the argument are the current ones and carry their own freshness. A state model rather than a dashboard is the precondition; without it there is nothing to reference.

Its alternatives are preserved, not discarded. A deck presents the chosen path and buries the rejected ones in an appendix. That is backwards. The rejected paths are what you need when conditions change, along with the condition under which the ranking would flip. Brainis models alternative futures before recommending a path, and the alternatives stay addressable afterwards.

Its work traces back to it. Every piece of execution points at the objective it serves. Not by convention — structurally, so the link survives reorganisation and can be queried in both directions.

Compilation, not transcription

The interesting move is treating the gap between strategy and work as a compilation step rather than a copying step.

Transcription is a human reading a strategic intent and typing tasks. It loses the reasoning, it loses the dependencies that were obvious to the author, and it produces work items whose acceptance criteria are whatever the person typing thought of at the time.

Compilation means an approved strategy becomes dependency-aware missions with owners, budgets, and acceptance criteria — generated from the strategy rather than retyped from it. Three things fall out of that.

The mechanism is on the Strategy Engine and missions.

What to keep

None of this argues for abandoning the deck. Keep three things from it.

The forcing function. Building the deck is what makes a team actually choose. The compression is the value.

The narrative. People commit to a story, not to a dependency graph. The narrative belongs in a presentation, and it should be presented.

The dated snapshot. There is real value in a fixed record of what you believed in July, precisely because it cannot update. That is a prediction record, and it is worth keeping for grading.

What changes is the deck's job. It stops being the reference document that execution reads from, and becomes an argument that gets made once, in a room, on a Thursday. The reference is the live strategy, sitting on live state, with the work hanging off it.

The deck was never the problem. Asking it to stay true was.