What the model agreed, with a name, a date and a budget

What a plan needs in order to be executed

An action plan agreed by a whole group arrives with all its force for about six months. That is the window for turning it into work with names, figures and dates, while the agreement is still warm and everyone remembers why they signed it.

The leadership model does something almost no organisation does on its own: it puts in writing what is expected of anyone running a team, measures how leadership is exercised today, and produces a set of proposals the leaders themselves have agreed. That is where its remit ends (and where it should end). What comes out of that room is an agreement on what has to be done. What is still missing is who, when and with how much.

The distance between agreeing and executing is always much the same, and it is almost always covered by the same four things.

• The plan needs an owner and a budgetThe proposals come out of a collective session, and a collective agrees: executing is always the work of named individuals. Each action gets a person with a name, a figure beside it and a date, and with that it stops being a well-drafted intention and becomes a commitment.
• The visible actions go firstAmong the things the model identifies there are actions that cost almost nothing, in time or in money, and that relieve something the whole place has been noticing daily. They are the stone in the shoe: sometimes it is literally that the light switches in a toilet have gone two years without being fixed and everybody knows it. Taking it out costs nothing and is appreciated at every step. They go first because they do what no organisational plan can do yet: show, without having to announce it, that things are changing and changing for the better.
• The organisational decisions are taken earlyReviewing the job-classification framework, assessing performance or redesigning a process all have consequences for named individuals, and that is why they are accepted so easily in the abstract. Actually taking them is what changes the rules of the game, and what turns the model into the way the organisation works.
• Don't buy before you designWhen a gap appears, the quick answer is to buy: a new post, a new machine. Looking first at whether that role still makes sense, whether the process is well designed or whether the work can be recombined turns an urgent purchase into a decision with judgement behind it. Sometimes the conclusion is to hire, and then you hire knowing what for.

The starting point

This service does not invent actions: it executes the ones the model already identified. Anything that did not come out of there does not get in.

It is a deliberate limitation and it is worth saying first. An action that has not come out of the diagnosis and the assessment brings with it neither the evidence that justifies it nor the consensus that makes it executable. You can propose it, you may even be right to propose it, and the organisation will still experience it as the consultant's idea. The ones that came out of the model are theirs, and that difference decides whether they get done or not.

For the same reason the earlier road has to have been travelled: a leadership model defined, assessed and with a closed action plan. If your organisation has its own — whoever produced it — and that model got as far as the assessment and the plan, it works as a starting point. If there isn't one, it has to be done first. This is not a commercial condition: without it there is simply nothing to deploy.

The two lines that get deployed

The action plan that comes out of the model always contains three kinds of thing, and two of them are deployed here. The third, training, has its own route and its own method.

• LINE 1Visible actionsLow-cost initiatives that embody the new style and are noticed straight away: getting to know what the service next door does, secondments between units, reshaping the management meetings, ring-fencing a percentage of time for improvement projects proposed by staff. They cost little, and that is why they go first.
• LINE 2OrganisationThe decisions that change the rules of the game: the talent development plan, process re-engineering and structures. They are slow, they have consequences, and they are the only ones that stop the model from being voluntary.
• LINE 3TrainingIt is not deployed here. When the plan calls for an in-house programme for current and future leaders, that is Learning for Leadership, a service with its own method, its own length and its own way of being commissioned.

The talent development plan

It is the biggest piece of the organisational line and the one that appears most often in action plans, though almost never under that name. An onboarding plan, a performance assessment system, a map of critical knowledge and a career path that doesn't force you to move up a grade are not four loose initiatives: they are four stages of the same cycle. Treating them separately is what stops any of them from quite working.

1. AttractionWhat kind of talent this organisation needs, which of it is available inside and outside, and what it can genuinely offer in return. Three questions almost never asked before a job advert goes out, and they decide whether the process will be of any use.
2. SelectionIt starts from the strategy and the processes, not from the vacant post. You identify the critical processes, the over-resourced ones and the invisible but key ones; you separate the formal post from the real role and from the profile needed; you look at workloads; and only then do you decide whether to fill it from inside or outside, with a rule written down before the candidate appears.
3. Onboarding, pay and performanceWhat a person has to understand and what they have to contribute at thirty, ninety and one hundred and eighty days. What is paid, what is not paid and on what criteria. And what is assessed: what is expected of each role, what the person has contributed, under what conditions they did it and — the question usually missing — what decisions follow from it.
4. LearningThe move from a training plan to a learning system: who translates the strategy into a learning agenda, how people learn while working, and which learning is expressly decided against. It is the hinge with the training programme, if the plan called for one.
5. Development and successionWhat growing here means without having to move up a grade, what real options exist and what is expected in order to reach them. And the map of critical knowledge: which roles cannot be left empty or improvised, what would happen if that person were not here tomorrow, who could take it on with the right preparation, and what has to be done today to make that possible.

Re-engineering before hiring

The problem is usually less a lack of talent than the design of the work.

Faced with hiring decisions, you have to measure where the time of the critical roles goes, split into three categories: tasks of strategic value, necessary but non-strategic tasks, and friction — rework, bureaucracy and badly designed decisions. The split is always a surprise, and almost always in the same direction.

With that figure on the table, hiring stops being the first option and becomes the fourth. Before it there are three decisions to take, one by one:

• Remove a roleWhen it answers to a process that no longer exists or that another one absorbed. It is the decision nobody takes, because it has no visible beneficiary and it does have a named casualty.
• Redesign the processWhen the role is saturated because the process forces rework that adds nothing. Hiring here buys time and locks in the bad design for another ten years.
• Automate a taskWhen the load is real but repetitive and traceable. It is the natural door into the conversation about artificial intelligence, and it starts here — in a specific process with measured hours — and not in a catalogue of tools.
• Recombine rolesWhen two posts each do half a function and neither does the whole of it. It is usually the cheapest decision and the one that generates most resistance.

Each decision is justified in writing with three things: impact on the strategy, impact on people, and the risk being taken on. Hiring remains a perfectly legitimate option. What stops being legitimate is hiring first.

Formal post, real role and required profile

The org chart says one thing, the real work says another, and what it takes to do it well says a third. Putting the three in columns for each critical process is a couple of hours' work and it always uncovers the same things: duplicated roles, orphan roles — tasks nobody owns — and star roles, people who hold up an entire process without their post recognising it or their pay reflecting it.

In public administration that table is the raw material for touching the job-classification framework with arguments, instead of doing it with the inertia of whichever posts happen to fall vacant.

How the work is done

Three horizons, with a decision at the end of each one on whether to carry on, adjust or stop. The total length depends on how many actions the plan has and on whether the organisational line includes the job-classification framework or not.

1. 0 – 3 monthsNow: turning the list into a planEvery action in the original plan gets an owner, a cost, a date and an indicator, or it is dropped in writing. Two or three visible actions are launched and the model and the plan are communicated to the whole body of leaders, not just to the committee. By the end of this stretch the organisation has to have noticed something concrete.
2. 3 – 9 monthsShort term: the organisational decisionsThe talent development plan starts with the stages the diagnosis flagged — usually onboarding, performance and recognition — and the re-engineering of the priority processes begins. If the plan called for a training programme, this is where it is designed and the first edition piloted.
3. 9 – 18 monthsConsolidation: aligning what gets decidedPerformance, career and recognition are tied to the same key functions the model defined. Without that coherence, the model coexists with an incentive system that rewards the opposite and ends up losing. A periodic review of the model is also established, so that it doesn't age in silence.

What comes out of the deployment

• The executable planEvery action with an owner, a cost, a timetable and an indicator. It is the document that turns a consensus into a commitment, and the one that makes it possible to ask for account without the conversation becoming personal.
• The talent development planThe five stages written and decided by the organisation: critical processes, the post-role-profile table, workload analysis, criteria for filling posts internally or externally, onboarding plan, pay policy, performance assessment system, career path and succession plan.
• The re-engineering decisionsWhich role is removed, which process is redesigned, which task is automated and which roles are recombined, each decision with its strategic justification and the risk taken on. In public administration, the raw material for reviewing the job-classification framework.
• The performance assessment systemWhat is assessed, at what level — person, role or process — how, and in order to decide what. With its risks written down beside it, which is the part usually missing and the one that makes these systems get abandoned in year two.
• What is decided againstThe proposals in the original plan that are dropped, with the reason. Without this document the full list survives indefinitely as a reproach, and any progress looks insufficient.

What makes it different

Where the method comes from

It is not a method thought up in the abstract. It comes from having had to deploy a leadership plan from inside an institution, with its budget and with its people.

At the Universidad Pública de Navarra, where I directed management between 2016 and 2025, the leadership model ended the way they all end: in a set of proposals agreed by the group that was going to live with them. What came next is what this page describes, and it was deployed along the two lines.

• Visible actionsA programme for each service to get to know the work of the others, secondments between units, a common way of running meetings with its protocols, and an onboarding plan for new joiners.
• Organisational decisionsA review of the structure of services, sections and units; the definition of results and objectives for each unit; project teams that cut across units; and the rules for remote working.

The third line, training, became an in-house programme with its own faculty and its own qualification, and it is told in Learning for Leadership.

Where this service ends and the ones next door begin

It is the reasonable question if you have read the other pages: several services here talk about changing how an organisation works. The boundary is sharp and it is worth saying out loud.

• Leadership ModelIt comes before, and it is a requirement. It defines what is expected of anyone who leads, measures it with the whole group and produces the action plan. Without that plan, this service has nothing to deploy.
• Learning for LeadershipIt is the third line of the same plan. When what is needed is an in-house programme for the people who lead, it is done there and not here.
• Service Redesign LabIt looks outwards: it redesigns a specific service from the experience of the person who uses it. Here you look inwards, at how the work of the people who deliver it is organised.
• Embedding user-centred strategyIt installs a permanent capability to redesign services, with a core team and an annual programme. This service does not install a new capability: it executes a specific plan that already exists, and then it ends.
• AI StrategyAutomating a task appears here as one of the four re-engineering decisions, on a specific process and with measured hours. Deciding which tools the organisation prioritises and how it governs its data is another conversation, with its own service from October 2026.

Who it works for

• UniversitiesWhen the leadership model is already written and assessed, and management knows that what comes next touches the job-classification framework and the professional development of administrative staff.
• Public administrationWhen the job-classification framework has spent years describing an organisation that no longer exists, and it has to be reviewed with arguments instead of with the inertia of vacancies.
• Public health and professional servicesWhere critical knowledge sits with very few people and the retirements you can see coming in five or seven years still have no succession plan.
• Large organisations and industryWhen headcount has grown, processes rest on named individuals and the default answer to every gap is still to ask for another hire.

How it is commissioned

What your organisation puts in

• An action plan that existsDefined, assessed and agreed with the group. It is the raw material of the assignment; without it there is no assignment.
• Senior management, decidingOrganisational decisions are taken, not proposed. If senior management is not prepared to sign whatever comes out, the work produces one more report.
• The people team, insideThe talent plan is written by whoever will maintain it. If the people area only comes in to validate at the end, the plan doesn't survive the first year.
• The owners of the critical processesTwo or three people who know how the work is really done, not how it is written that it is done.
• A willingness to look at the job-classification frameworkNot necessarily to change it: to look at it. If it is untouchable by a previous decision, half the organisational line falls away, and it is better to know that before starting and not in month six.
• Accepting that some actions get droppedThe original plan usually contains proposals that don't survive the cost or impact analysis. Dropping them in writing, and explaining why, is part of the work.

How it is closed

The scope depends on how many actions the plan has, on whether the organisational line includes a review of the job-classification framework and on how many processes go into the re-engineering. It is set in a prior conversation about the specific action plan, and each assignment is quoted on that basis. If the leadership model isn't done yet, the conversation is a different one and starts there.