Digital transformation of government ministries, built to be repeated.

Ler esta página em português

The problem

Government digital programmes are procured in multi-year blocks and judged at the end. That is the whole problem in one line: a long horizon hides failure. A decision that was wrong in the first months does not surface until the money is spent, everything else has been built on top of it, and everyone who made it has moved on.

The large contract is the failure mode, not the safeguard. The NHS National Programme for IT was abandoned at a cost of roughly £10bn. Healthcare.gov was sunk by its systems integrator and then rescued by a team of under twenty people. Size buys the appearance of seriousness and very little else.

The other standard failure is building identity and infrastructure before any service a citizen actually wants. Kenya spent roughly US$100m on Huduma Namba, a national identity number that arrived ahead of the services that would have given people a reason to register; eCitizen, which led with services, is the counter-example. And a programme that survives both of those still tends to leave dependency rather than capability — the service runs, and it runs because the supplier is still there.

FAROL exists because all of this is documented, and it is built backwards from the record: what worked in Rwanda, Cape Verde, Togo and Estonia, and what failed in Kenya, the United Kingdom and Australia. Every rule in the method is charged to a programme that already learned it somewhere else, at someone else's expense.

Five principles, non-negotiable

• F

  Focus on outcomes / Foco no resultado

  Every phase begins with a target metric, never with a requirements list. A requirements list is how people agree on activity without agreeing on a result.

• A

  Abolish before automating / Apagar antes de digitalizar

  Cut the process first. Digitising a twelve-step process is automating waste. A typical service goes from twelve steps to four before a line of software is written.

• R

  Rhythm of 90 days / Ritmo de 90 dias

  No phase goes longer than 90 days without delivering visible value. The programme can be long; the wait for evidence cannot.

• O

  Ownership from day one / Operação local desde o dia 1

  Ministry staff operate the system from the start. Capability transfer is the method itself, not a final phase bolted on at the end.

• L

  Ledger of evidence / Lastro de evidência

  Every phase produces measurable proof that funds and justifies the next one. No proof, no next phase.

Non-negotiable means something checkable here: any project choice that violates a principle is escalated to the sponsor together with an alternative. There are no silent exceptions.

The phase engine

1. P0 · 3 weeks

   X-Ray

   Where is the money and the pain? Week one follows the money: revenue and cost by service, where the fees sit, where the leakage happens, where paper still rules. Week two follows the queue — volumes, waiting times, points of discretion, citizen pain. Week three prioritises and sells: one beacon service chosen, with the business case and the 90-day plan ready for signature.

   Exit gate: a business case signed by the Minister. No signature, no P1 — which is what stops months of interest without commitment.

2. P1 · 90 days

   Beacon

   Can we deliver visible value? One service end to end — application, workflow, decision, payment, verifiable document — redesigned before it is coded, so every field, signature and stamp justifies itself or dies. The legal basis is a ministerial order rather than legislation: weeks instead of years. The dashboard is public from week four, and the service ships with an assisted channel, a counter and agent network that files on behalf of citizens without a smartphone.

   Exit gate: the target metrics are met — or the programme stops or pivots, publicly.

3. P2 · 6–12 months

   Foundation

   Can the ministry sustain scale? A clean master registry of whatever the ministry actually is: people, licences, cases, assets. Minimum viable interoperability — a simple gateway, not a national platform. Pragmatic identity, using the civil registry, SIM registration or bank KYC rather than waiting for a national ID. A legal package that settles electronic acts and basic data protection, and an internal academy that trains the trainers.

   Exit gate: three services in production, operated exclusively by the ministry's own staff.

4. P3 · Ongoing

   Factory

   Does the machine replicate itself? Service by service from the playbook, with a shared platform team alongside the service teams, and benchmarking across services and across ministries on the same standard metrics: time from application to delivery, cost per transaction, revenue collected against potential, digital transactions per month, and what citizens say about it.

   Exit gate: a pipeline prioritised and executed without the architect on site.

The gate at the end of P1 is a kill gate, and that is the point of it. The targets are set on day zero, in writing, and if they are missed on day 90 the programme stops or changes direction — publicly, with the numbers that say so. Almost nothing in this field is built so that it can be stopped by its own evidence, which is exactly what makes the evidence worth anything afterwards. If the targets are missed, the programme stops at no cost to the next phase.

The artefact library

Each phase produces documents rather than a deck. Twelve are indexed, and the status column is the honest part of the table.

An artefact enters the library only after it has been used on a real case. Four of them are written as documents today — A2 lives inside A1 — and the rest are not. Saying that here is cheaper than being found out later.

The diagnostic canvas is published in full, because the questions a ministry is asked in the first three weeks are the fastest way to judge whether any of this is serious. It prints to A4 and is meant to be filled in.

Field rules

Twelve rules are distilled from public programmes that worked and public programmes that did not, and each one is charged to the case that paid for it. Breaking a rule is escalated, exactly like breaking a principle. Six of the twelve:

What makes it different

Why this is an asset

FAROL is written down and numbered. The phases have codes, the gates have criteria that can actually be failed, and the artefacts have an index with a status column that admits what is not written yet. Someone other than its author can pick it up and run a phase with it, which is the only test that means anything.

It is domain-independent. The artefacts do not change between a health ministry and a finance ministry; what is written in them does. A diagnostic canvas asks the same questions of a licence renewal as it does of a tax refund.

And it compounds. Every phase adds to the ledger of evidence, so a fourth case starts from everything the first three learned rather than from a blank page.

Which is what makes it licensable. The same sequence, the same artefacts and the same metrics in any ministry means results can be compared across ministries and across countries, and local teams can be trained to run it — rather than selling hours, over and over, always starting from zero.

Talk to us

Write to info@simplesmoothsafe.com. Say which ministry or organisation you are writing about and which service is hurting — not what you would like to buy. You will get a direct answer, including “this is not a fit” when that is the honest one.

Email info@simplesmoothsafe.com