Strategy Manager — CEO Strategy Layer
Company strategy as reviewable data, not slides — every objective, owner and dependency in one graph.
In build — canvas at 17 of 19 tasks; runs locally, not yet deployed
- 143 — Strategy nodes captured as structured data — missions, objectives, initiatives, constraints and the people accountable for them
- 155 — Declared relationships between them, 130 approved and 11 rejected on the record — a reviewed graph, not a diagram
- 0 — New databases. Strategy lives as YAML in git, so a change to the company's direction arrives as a diff someone can approve
Most organisations have a strategy. Very few can say what it currently says.
It lives in a deck from a planning offsite, and the deck is already out of date.
Nobody can answer what does this initiative actually serve? without asking the person who wrote it. Objectives and the work delivering them sit in different systems, so nothing tells you when the two have drifted apart. Strategy spanning several businesses gets quietly conflated, and a decision made for one ends up applied to another. And rejected ideas vanish, so the same argument is re-litigated every quarter.
Strategy as something you can question
- Nodes, not prose. Missions, drivers, constraints, objectives, initiatives, capabilities, stakeholders and metrics are each a typed record with an owner and a lifecycle.
- Links that are argued. A node declares what it contributes to, depends on, or must precede — and each link carries a status, so a rejected idea stays visible as a decision instead of disappearing.
- Several businesses, no bleed. Every node carries the business it belongs to. One capture pattern and one vocabulary, but never one scoring model, so strategy for one part of a group is never silently read as strategy for another.
- A canvas for the chains. An interactive board for what has to happen before what. Goals and outcomes are kept deliberately distinct from events, because they are different kinds of claim.
- It creates the actual work. Something ready to be real becomes a project and a milestone in the tracker the team already uses, with the link recorded on both sides.
Status is resolved, never copied
A node points at the work delivering it, and that status is fetched live when somebody asks.
Nothing is written back into the strategy files, because a copied status is wrong the moment it is written — and a plan that quietly disagrees with delivery is worse than one that admits it does not know.
Confident enough to be provisional
A node that is only a guess is marked as one and kept out of the surfaces that would make it look settled.
That is deliberate. The failure mode of every strategy tool is that writing something down makes it look decided; a low-confidence idea should be capturable without appearing on a rendered map as established fact. Nothing is hard-deleted either — superseded and archived are distinct recorded states, so the history of a decision survives the decision being reversed.
Where it fits
- Groups running several businesses, where one strategy conversation keeps contaminating another.
- Anyone tired of a strategy that exists as a deck nobody has opened since the offsite.
- Teams who want the link between a stated objective and the work actually funded to be inspectable rather than asserted.
- People comfortable treating strategy the way engineering treats code — reviewed, versioned and diffable.
Follow one thread
The fastest way to judge it is to follow a single initiative up to the objective it serves, and down to the work delivering it.