The state won’t rewire itself – why Labour’s reform agenda hinges on delivery
At the opening of the Hippo Digital Glasgow Office, Chief Delivery Officer Mike Tattersall discusses how inclusive design can support vulnerable users
Mike Tattersall, Chief Delivery Officer
| Hippo Digital
Labour has put digital reform and AI at the centre of its plans to make the government and public sector more productive. But technology alone will not transform the state. The real test is whether Whitehall changes how it funds, owns and delivers reform.
Every government arrives with ambitious plans to reform the state. What is different now is the scale of the pressure to make that reform work. As the Labour Party gathers in Liverpool, ministers are trying to reconcile rising demand for public services, constrained public finances and an expectation that technology, particularly AI, can help close the gap and support cost of living pressures.
The potential impact is significant. The government has committed £2bn to AI in the Spending Review 2025 and said there are potential savings of up to £45bn a year from digitising the public sector. The Tony Blair Institute has similarly estimated that AI could generate £10 bn in net annual savings by the end of this parliament.
But ambition and outcome are fundamentally different things.
After ten years of building digital services across central government and the NHS, one lesson stands out to us - government rarely lacks ambition. The harder problem is turning that ambition into services that work for people and achieve the policy outcomes that ministers want.
Measuring how things change
Digital programmes are often measured through outputs: services launched, systems migrated, features delivered or projects completed. Those things matter, but they are not the outcome. A digital service can go live on time and still make life harder for the person using it. A department can introduce a new platform while leaving the underlying process unchanged.
This matters when the scale of public investment is so significant. The Government Major Projects Portfolio now represents close to £1tn in whole-life costs, yet only a relatively small proportion of projects are rated green. The lesson from previous failures, from major NHS technology programmes through to identity verification, is that the problem is rarely technology alone. It is often a failure to align policy intent, operational reality and delivery.
The question at the centre of every major programme review should therefore be simple - what has actually changed for the citizen, the frontline worker or the taxpayer? The best programmes can answer for all three.
We have seen the difference this makes in practice. In work supporting DWP on fraud and error, the focus was not simply on delivering a new digital service, but on whether it improved the ability of thousands of frontline staff to identify incorrect payments. That service scaled to 3,500 agents and helped save £69.8m in three months, contributing to a wider programme expected to prevent around £1bn in incorrect payments by 2030.
That is the kind of measure that should define success.
AI will not fix a bad service
The same principle applies to AI. The government’s ambition to become the fastest-adopting AI country in the G7 creates a genuine opportunity to redesign how people interact with the state. The Alan Turing Institute estimates that 84 per cent of around 143m complex repetitive transactions carried out across government each year are highly automatable.
But automation is not the same as reform. If organisations have not done the difficult work of identifying the real problem and what the service is actually there to achieve, AI will simply allow them to build the wrong thing faster or add yet another layer to a process that is already too complicated for the people using it. Putting a chatbot on top of a confusing process does not make that process better. Automating fragmented workflows does not remove the fragmentation underneath them.
The opportunity is to redesign services around what the policy is meant to change in people’s lives, and then determine where AI can remove friction, reduce administrative burden and support better decisions. Trust is part of that foundation too. NHS Login, which now supports 130 services and has facilitated more than 1.4bn authentications, shows what happens when identity and trust are treated as core infrastructure rather than an afterthought. That kind of foundational work is what allows digital public services to scale safely.
Three practical shifts needed
Our policy paper - Ten years of delivery, the next ten of change - identifies three areas where government can make a practical difference.
The first is accountability. Government talks constantly about breaking down silos, but budgets, data and decision-making still sit largely within individual departments. That becomes a problem because citizens do not experience government department by department. Someone leaving prison, looking for work or accessing healthcare experiences a single journey, even if Whitehall does not organise itself that way.
Cross-government priorities need named leaders with the authority to make decisions across organisational boundaries, including over data. When Hippo helped deliver the Resettlement Passport, information could move securely between the Ministry of Justice and HM Prison and Probation Service (HMPPS) because the team was organised around a clear outcome, supporting someone leaving prison, rather than around institutional boundaries. It worked because leaders across the Ministry of Justice and HMPPS were willing to take shared ownership of a problem no single organisation owned, and the next phase of reform needs much more of that kind of leadership.
The second is funding. Major digital transformation rarely fits neatly within annual budgeting cycles. Replacing legacy systems, redesigning services and building capability can take several years, yet delivery teams are often forced to optimise around short-term spending settlements. More multi-year funding tied to measurable outcomes would allow departments to invest more sensibly and reduce the familiar end-of-year rush to spend remaining budgets before they disappear. If the government wants to transform services over several years, the funding model needs to reflect the life of the problem rather than the accounting year. Longer settlements should not mean looser accountability, though. Departments and their suppliers should be held to clear outcomes across that period, not simply to spending the money.
The third is procurement. The current system can still favour short-term price over long-term capability, despite the government's wider ambition to strengthen domestic technology capacity and reduce dependency on a small number of suppliers. Value for money should take greater account of the skills, knowledge and resilience that remain within the public sector once a contract ends.
That means treating capability transfer as part of the outcome. The best supplier relationship should leave behind stronger teams, reusable technology and less dependency on the supplier itself.
Delivery is the reform
The next phase of public service reform will not be won through another round of structural changes or strategy documents alone. It will be won in the machinery beneath them - who is accountable, how money flows, whether data can move and whether the people delivering services have the authority and capability to change them, with a clear line of sight to the outcomes they are there to deliver.
Technology can accelerate that process enormously, but it cannot substitute for it. The state will not rewire itself. Unless government gets the conditions for delivery right, there is a real risk that it simply becomes much better at building the wrong things faster.
To find out more, read Hippo Digital’s new policy paper: Ten years of delivery, the next ten of change
Mike Tattersall is Chief Delivery Officer at Hippo Digital. He began his career in the civil service at 17, going on to spend more than a decade working across product and service design in organisations including the Government Digital Service, DWP, the NHS and the Department for Education. Mike has delivered scalable services through Discovery, Alpha and Beta to Government Service Standard. Since joining Hippo in 2017, he has led the delivery of national services including NHS Login and the pandemic shielding service, while helping grow Hippo’s delivery team from 60 people to more than 800.
