Blog

Digital transformation in the public sector: where it goes wrong (and how to make it work)

Almost every public organisation talks about digital transformation. Yet many initiatives stall in drawn-out procurement, half-finished programmes and software nobody uses. In this article we name where it goes wrong in the public sector, and above all: how to make it work.

Why digital transformation in government so often stalls

Digital transformation in government is not an IT project, but most organisations treat it as one. A system is procured, a vendor is selected and a go-live date is set. Two years on there is a platform, but the work processes, the data quality and the people were never brought along. The result is an expensive shell wrapped around an unchanged organisation.

On top of that comes the specific context of the public sector: long procurement cycles, political decision-making, a tight IT labour market and strict frameworks around privacy and information security. Each of these is legitimate, but together they form a brake that lets transformation sink into process rather than results.

Misconception 1: starting big instead of small

The classic mistake is the multi-year programme with a big bang at the end. The larger the scope, the later the first working outcome and the greater the chance the world has changed in the meantime. In the public sector we see this in programmes with dozens of components that hold each other hostage.

It works the other way around. Cut it into small, self-contained chunks that deliver something within weeks that an employee or citizen actually uses. For a tax authority partnership, for instance, we tackled the automated processing of property and citizen-registry changes as separate, identically structured steps. Each step is complete in itself, delivers value and builds confidence for the next.

Misconception 2: steering on delivered features instead of on value

Many government programmes measure progress in delivered functionality: so many screens, so many integrations, so many ticked-off requirements. But a ticked-off requirement is not value. The real question is whether the assessment goes out faster, whether the citizen has to supply less, whether the employee does less manual work.

Best Value thinking helps here: let the vendor or team demonstrate which measurable outcome they deliver, and steer on that outcome rather than on a locked-down specification. That calls for client leadership that dares to let go of the 'how' while staying sharp on the 'what' and the 'why'.

Misconception 3: data quality and information security as an afterthought

A transformation is only as strong as the data underneath it. Yet data quality is often discovered only at the end, when the shiny new system turns out to be neatly automating the circulation of polluted data. In government, where decisions directly affect citizens, that is not a detail but a risk.

The same goes for information security and compliance. Frameworks such as BIO, ENSIA and NIS2, and the requirements of the supervisory bodies, are not boxes to tick afterwards but design principles up front. Build them in from day one and you avoid costly rework and create a platform that survives both the audit and daily practice.

AI in 2026: opportunity and pitfall for the public sector

In 2026 AI makes a great deal possible: classifying changes, summarising documents, automating quality checks, relieving staff of repetitive work. The gain is real, but so is the pitfall. A model that makes opaque decisions about citizens is simply unacceptable in the public sector.

Responsible use means a human in the loop for decisions that matter, explainability, and clear agreements about data and privacy. Start small, with a bounded application where the outcome is verifiable, and only scale up once trust has been earned. That turns AI into a reliable colleague rather than a black box.

How to make it work: in short

Digital transformation in the public sector succeeds when you treat it as a change of people, processes and technology together. Start small and deliver something that works quickly. Steer on measurable value rather than on delivered features. Make data quality and information security design principles from day one. And deploy AI responsibly, with people at the wheel.

None of these steps is spectacular. Together they make the difference between a programme that sinks and an organisation that genuinely changes.

Curious where your organisation stands?

We are DWDA: interim CMO and partner in digital transformation, with the public sector as our specialism. We combine experience in information security (BIO, ENSIA, NIS2), property valuation and supervisory frameworks, Best Value and responsible AI with a level-headed, results-driven approach.

Want to know where your organisation stands? Take our free maturity scan and receive an instant report with concrete next steps. Prefer to talk through a stalled initiative or a first bounded step? With our entry vouchers we deliver a first result at a fixed price, without a long-term commitment. Feel free to get in touch and we'll look together at where the biggest gains are.