For anyone turning strategy into work
Plans rarely fail for want of intent. They fail because an action with no owner, no measure and no place in the business is indistinguishable from a wish, and because nothing in the plan says when to stop. The Omnigoal contributes one thing here, and it is a structural one: every action can be given an address — the core and object it changes — and an address is what makes ownership follow from the work rather than from the meeting.
The problem
When a plan lists a department beside an initiative, the initiative advances only if someone volunteers. That happens often enough to disguise the problem and rarely enough that the plan cannot be relied on.
Without an explicit method, ordering tends to reflect who spoke last or most confidently. A dozen simultaneous first priorities is not a prioritisation failure of individuals but of the process they were given.
An action detached from a goal cannot be cancelled, because no one can demonstrate it is not working. Plans then grow monotonically until the year ends and the list is rewritten rather than concluded.
The method
Seven steps from strategic intent to work that can be started, measured and — where necessary — stopped.
Open the Goal Core and name the goal this plan serves. A plan is a set of bets about how to move one goal; where the goal cannot be named, what is being written is a task list, and calling it a plan makes it harder rather than easier to review.
Locate each proposed action in the core and object it changes. This does two things at once: it makes visible which part of the business is being asked to move, and it identifies who is accountable for that part. It is also a useful filter — an action that fits no object is usually not novel but underspecified.
Score each action against the named goal, in realistic working weeks, and by what it depends on. Dependency is the field most often omitted and the one that most frequently determines what is actually possible. The scores are estimates rather than measurements; their value is that they are written down and can therefore be disagreed with.
The Time Core holds project management, market timing, golden opportunities, contingency and the year wheel. Place surviving actions against real seasonality rather than distributing them evenly: some windows close, and a plan indifferent to them will miss them without recording that it did.
Define in the Data Core what will show that the action worked, and what result would lead you to stop. Agreeing the stopping condition in advance is the point: it is decided while no one is committed, which is the only moment at which it can be decided cheaply.
Each action takes a single named owner and a fixed review moment. Shared ownership is a reliable way of producing an unstarted initiative — a finding old enough that RACI exists mainly to prevent it.
Return to the eight cores quarterly and ask what has changed. This is the iterative epistemology the framework rests on — plan, act, observe, reflect — applied to the plan itself. Plans decay because the business moves beneath them, and that is visible in the map long before it is visible in the task list.
Take this with you
Seven fields. A line missing any one of them is not unusable, but it is where plans reliably begin to come apart.
Where an action resists being given an address, the usual cause is that it has not yet been specified clearly enough to start. Rewriting it until it fits is the cheapest quality check in the method.
Standing on other work
Almost everything in this method exists already and is well documented. What the framework adds is the address — a shared way of naming which part of the business an action concerns.
Goal Core and Data Core
Objectives are goal-setting, key results are measurement. Keeping them in separate cores makes the distinction harder to blur.
Goal Core cascading into Business Core
Policy deployment: the discipline of carrying a small number of goals down into operational commitments.
Owner field on every action
Exists largely to prevent shared accountability, which is the most common cause of an unstarted initiative.
Prioritisation step
Useful as a way of making disagreement visible. Its numbers are opinions and behave badly if treated otherwise.
Time Core — Project management
The field most often left out of business plans, and the one that most often determines the real order.
The quarterly re-read
Deming’s cycle is the same iterative loop the framework’s epistemology describes.
Stopping condition and contingency
Klein’s technique for surfacing the failure modes people can see but do not raise while commitment is forming.
None of these is original to the framework, and the method is stronger for that. The contribution is the shared address, which is what lets ownership and measurement be derived rather than negotiated.
Where it lives in the framework
The Omnigoal is holistic: all eight cores are in play here, each supplying something the others cannot. Follow any of them to see the objects it contains.
The goal the plan serves — and the criterion for cancelling an action.
Learn the core
The values that decide which actions are declined, whatever their impact score.
Learn the core
The year wheel, market timing, project management and contingency.
Learn the core
The measure per action and the review rhythm that keeps the plan answerable.
Learn the core
Where most actions land: systems, people, finance, partners.
Learn the core
Campaign, channel and content actions, which carry real seasonality.
Learn the core
Actions touching promise or positioning — slow to show results, expensive to reverse.
Learn the core
Actions that change what the customer actually encounters.
Learn the coreChecklist
In fairness
A plan can be well formed and still be the wrong plan. These are the limits worth naming before relying on the method.
Questions
The questions people ask most often about action planning and The Omnigoal.
Take one goal from the strategy, list the candidate actions, and give each an address — the core and object it changes. Score them for impact, effort and dependency, sequence the survivors against real seasonality, and attach one owner, one measure and one stopping condition to each. The address step is what converts strategic language into work someone can begin on Monday, because accountability then follows from the structure rather than from the meeting.
Score impact against the named goal, effort in realistic working weeks, and dependency on other work. Dependency is the field most plans omit and the one that most often decides the order: an initiative with high impact and an unmet dependency is not this quarter’s work, however attractive it looks. Treat the scores as recorded opinions rather than measurements — their value is that they can be argued with.
Fewer than feels comfortable. A workable test is that each action has a distinct owner with genuine capacity for it; where two actions share an owner who is already fully committed, one of them is not really planned. Leaving a quarter underloaded is deliberate rather than slack — unplanned work is a normal feature of a business, and a plan with no room for it tends to be abandoned rather than adjusted.
It is the result that would lead you to end the action. Agreeing it before the work starts removes the sunk-cost argument later, because the decision was taken while no one was committed. It is worth also recording who has the authority to invoke it, since a stopping condition nobody is empowered to call is not one.
They fit together rather than compete. OKR is a goal-setting and measurement discipline, which sits in the Goal Core and the Data Core; The Omnigoal contributes the address that says which part of the business a key result actually concerns. Where OKR alone can leave an objective floating above the organisation, the address ties it to an object and therefore to an owner.
Progress monthly, relevance quarterly — they are different questions and conflating them tends to mean only the first gets asked. The quarterly review should re-read all eight cores rather than only the plan, since plans decay because the business moves beneath them, and that movement shows up in the map before it shows up in the task list.
