Digital transformation

Digital transformation is the only phrase in this masthead a reader might reasonably object to. It has been attached to website redesigns, redundancy programmes, enterprise resource planning migrations and the purchase of laptops. The objection is fair. The response is not to abandon the term but to insist that anything filed under it names what changed, for whom, and at what running cost.

There is a usable public measure of how this goes at scale. In its Major Projects Annual Report published on 13 July 2026, the National Infrastructure and Service Transformation Authority reported 189 projects on the government's major projects portfolio at the end of March 2026, carrying a combined whole life cost of £924.2 billion. Twenty-nine were rated green for delivery confidence, 109 amber, 34 red, and 17 were exempt from rating.

Those ratings are assessments of confidence rather than verdicts, and the portfolio contains projects whose scale would make any assessor cautious. Even so, the distribution is worth sitting with. Fifteen per cent green, on the programmes that are examined most closely, assured most heavily and reported to Parliament annually. Private sector transformation is not assessed to that standard by anybody, and there is no reason to think it does better.

Pilots clear. Rollouts do not.

The consistent pattern is not failure at launch. It is a first deployment that works and a second that does not, because the first ran with volunteers, executive attention and tolerance for breakage, and the second runs in a department that did not ask for it, has its own reporting line, and had its objections recorded and overruled.

That is an organisational problem wearing a technical costume, and it is where most of the money goes. The piece on why transformation stalls at the second department works through the mechanics.

The remedy proposed at that point is almost always more communication, and it almost always fails, because the objection was usually accurate. Nobody embraces a change to how they are measured. They agree to try it, on terms, and the terms are the negotiation that most programmes skip, which is the case made at nobody embraces change.

The vocabulary is not a side issue

Two words get used interchangeably in board papers and mean entirely different amounts of money. Putting an existing process onto a screen is one thing. Changing what the process is, so that the organisation does different work in a different order, is another, and it is the one that produces both the savings and the resistance. A programme that funds the first and promises the returns of the second is the single most common way British transformation budgets are lost. The distinction is worked through at digitising, digitalising, and the difference that decides a budget.

A related habit is picking the system first. The solution should be the last decision, not the first, and a surprising proportion of programmes invert that because the procurement timetable demanded a named product before anyone had watched a user try to finish a task. The solution is the last thing you should pick is the argument, and it is not a design fashion. It is a sequencing rule with cost consequences.

The National Audit Office made an adjacent point about capability in its March 2023 report on the barriers to efficient digital transformation in government, observing that many business leaders do not yet have the expertise required to comprehend and tackle the challenges the civil service faces in a digital age. That was written about Whitehall. Anyone who has sat in a large private board meeting about a platform migration will recognise it, and no comparable audit body publishes the equivalent finding about FTSE boards.

Four public signals that a programme is in trouble

A delivery confidence rating that slips two years running, rather than one bad year. A contract variation, because a variation is where the plan first meets the estate and somebody had to write down which one gave way. The quiet departure of a senior responsible owner, which is often the earliest thing visible from outside. And a post-implementation review that is commissioned and then not published. All four are matters of public record for government programmes, all four are watchable without a single briefing, and all four are systematically ignored in favour of the launch announcement.

Automation is the thing most often sold as transformation, so it is covered separately and at greater depth under intelligent automation. The organisational leadership question, which has quietly changed shape as the chief technology officer role split, sits at the job that replaced the transformation CTO.

The position taken here is that transformation is mostly a question of who has to change their working week, and that the technology is usually the least contested part of the plan. Any account of a programme that does not say whose job changed is describing a purchase, not a transformation. Public sector programmes are covered in more procurement detail under public sector technology.