Salem J. AbabnehHow I workMethod 01

Building a roadmap a government actually executes

A strategy that names an ambition is easy. The hard part is the line from that ambition to a specific project, with an owner, a budget and a place in the queue. This is how I get there.

Seven workstreams phased across a four-year roadmap Each workstream starts at a different point and runs for a different duration, so that dependencies are respected and delivery capacity is never all committed at once. Year 1Year 2Year 3Year 4 Operating modelsService operationsData & emerging techAI agentsHuman resourcesPolicyBehavioural science

The method

  1. Diagnose the current state honestly

    Not a maturity score. What works, what is duplicated across entities, and what has already been attempted and failed, because that history decides what is politically possible now.

  2. Define the workstreams before the projects

    Seven of them on the national roadmap. Workstreams are how a portfolio stays coherent: every project has to belong to one, and a project that belongs to none is usually somebody's pet.

  3. Build the longlist, then cut it against a defined bar

    Everything that could be done, gathered across every entity. The value is in the cut. A candidate survives only if it is independently deliverable, has its own finish line, and can carry a named owner. If it fails that test it is a task inside another project, not a project, and treating it as one is how a portfolio inflates to a number nobody can run.

  4. Sequence against capacity and dependency, not enthusiasm

    Most roadmaps fail here. They get sequenced by what leadership wants first rather than by what the delivery organisation can absorb at once, and by what must exist before the next thing can start. One workstream is named the dominant priority and the rest are sequenced around it, because a portfolio where everything is priority one has no priorities.

  5. Make blocked a visible state, not a quiet delay

    A project waiting on a decision is not in progress and is not finished. If blocked is not its own state on the board, blocked work reads as active work and the first honest picture of the programme arrives months late.

  6. Cost it, own it, and label every estimate honestly

    Every project carries a number and a named owner; an unowned project is a slide, not a commitment. Where a figure is indicative it is marked as indicative on the page it appears. A placeholder that is allowed to look like a real number is the fastest way to lose a programme's credibility, and you only lose it once.

  7. Stay for activation

    The roadmap is the beginning of the work, not the deliverable that ends it.

Where this comes from

A centre-of-government national modernization roadmap: four years, more than $200M, 111 projects across 20 government entities, structured into seven workstreams covering operating models, AI agents, human resources, policy, data and emerging technologies, behavioural science, and service operations.

The other methods

Changing how a large organization works? Let's talk.

[email protected] LinkedIn Download CV