Building confidence in digital and AI transformation for the public sector

Every year, the UK public sector spends £26 billion on digital technology.¹ Although digital has long been championed as a vital driver of increased productivity and improved policy outcomes, the reality is that most initiatives fall short of their intended outcomes. As the National Audit Office (NAO) starkly puts it, there is “a consistent pattern of underperformance” in the UK government’s attempts to deliver digital change.²

This isn’t sustainable, especially as the rapid advancement of artificial intelligence (AI) and emerging technologies reshape work and the world around us. Government has an incredible opportunity to harness digital and AI to become more productive and responsive to citizen needs. According to the recent State of Digital Government Review, the public sector stands to realise over £45 billion in annual savings and productivity benefits through full potential digitisation of services.³ The true opportunity for growth and wider productivity for UK Plc could be much higher. Digital is no longer just about delivering services more efficiently, it is about harnessing insight from each interaction by which we can learn, improve and create new ideas and opportunities.

A brighter digital future is within reach. But the public sector will need to overcome significant barriers to get there. They include complex legacy systems, a culture of risk aversion, lack of digital skills, and a disconnect between technology delivery and business goals. Altogether, they create a challenge of confidence. Because of past underperformance, public sector leaders and civil servants don’t always believe that digital transformation programmes will deliver and sustain their expected benefits.

What if the public sector had confidence that digital change would not only succeed but provide platforms for growth and learning? What if there were a way to be sure that every digital and AI initiative had the right objectives, was set up to deliver effectively, and achieved lasting impact?  

At Baringa, we believe gaining this confidence is possible. It hinges on three key dimensions – clarity, certainty, and continuous optimisation of outcomes. With this framework, the UK government can maximise the return from its digital investments, align technology with strategic goals, and deliver better services to citizens. Here, we summarise the challenges and opportunities associated with applying these three dimensions to drive successful digital and AI transformation.

1. Clarity of outcomes

The first crucial step in successful digital and AI transformation involves defining clear outcomes. Well-defined outcomes enable teams and leaders to prioritise work and ensure that their efforts remain strategically aligned during the delivery of a project. To achieve this clarity, we recommend prioritising three elements:

• Addressing "disco(very) fever". Each new government digital project starts the discovery phase from scratch, leading to delays and duplication of effort. Discoveries have been conducted as standard for all digital initiatives for many years, but the collective insight and learning is not harvested from this investment. A solution lies in establishing a central knowledge repository that acts as a foundational resource for new initiatives. It would summarise relevant insights and focus new discovery phases on genuine knowledge gaps, saving time and resources. At the same time, each project should be thinking more clearly about what government wants to learn and improve on from each project as this will define the data approach. A programme like ‘Making Tax Digital’ isn’t just about automation, it is about building new depths of understanding into UK commerce, growth and challenge areas, tax yield, employment and many related topics. Discoveries should be focused less on things that have already been discovered and more on new boundaries of knowledge and experience.
• Implementing dynamic outcome mapping targeting growth as much as cost. Traditional approaches to defining business cases, aligned with the Green Book business case structure, tend to result in inflexible, long-term projections that conflict with the dynamic nature of digital programmes. To address this, government departments should adopt a dynamic outcome mapping approach, which transforms static business cases into living documents where outcomes can be defined more clearly and updated dynamically as project requirements evolve. Moreover, typical Green Book-aligned IT business cases necessarily focus on cost avoidance or reduction when far greater upside often exists in growth potential for UK Plc. The potential learning that AI enables multiplies this upside potential by orders of magnitude.
• Taking a legacy-aware approach alongside a future-aware approach. Outdated legacy systems remain a key source of inefficiency and a major barrier to modernising services. Government departments should adopt legacy-aware outcome mapping that integrates technical debt reduction and legacy system migration into the core objectives of every digital initiative. This ensures that digital investment contributes to a more agile, efficient, and future-proof technological landscape. The costs of carrying legacy infrastructure should be measured in more than just the directly observed maintenance costs and cyber risks. The barriers created by legacy infrastructure represent opportunity cost and inflexibility that limits national growth.

2. Certainty of outcomes

With the average ICT project in the Government’s Major Projects Portfolio (GMPP) taking seven years to deliver,⁴ the public sector needs ways to improve certainty even as technology and requirements evolve. Instead of trying to predict the future of long-term digital and AI transformation programmes, project teams should focus on what can be controlled. This includes:

• Investing in procurement strategy. Research from the NAO reveals that poor supplier relationships have contributed to over £3 billion in cost increases across large digital change programmes.⁴ To improve this, government should focus on creating aligned incentives with suppliers, developing long-term partnerships, and blending digital and commercial expertise in procurement teams. This approach enables more collaborative relationships and better value for money. Procurement strategy cannot be expected to have 20:20 foresight, but should be expected to map out implications from well-established trends already in play. We do not know where generative AI’s impact on software development will be in seven years, but we know the marginal cost to deliver a line of code will have been reduced substantially. Public procurements need to cater for such trends to the advantage of citizens and businesses alike. Public sector procurement can drive faster adoption of improvements in the UK IT industry, or it can act as a trailing anchor to the past.
• Building internal delivery management capabilities. Strong delivery management plays a crucial role in ensuring the smooth progress of complex digital change programmes. Yet, delivery management skills are often in short supply within public sector organisations. Government must prioritise attracting and retaining skilled delivery managers who can facilitate collaboration between cross-functional teams and manage risks effectively. As AI leaps forward, so will the gap between where many public sector teams are and where they need to be.
• Leveraging cross-government experience. When government departments share their experiences and knowledge, it can have far-reaching impact. However, this is currently the exception, rather than the rule across the public sector. The creation of the Digital Centre for Government offers a way to turn this tide. We recommend seizing this momentum to establish formal, cross-government mechanisms for sharing data, insights, and lessons learned, so public sector organisations can approach digital and AI transformations with greater confidence. The journey with AI will multiply the importance of this. AI is forcing us all to reconsider how we learn and what we learn, whilst compounding the value of structured, relevant data and learnings.

3. Continuous optimisation of outcomes

Only 8% of major government change programmes are rated as highly likely to deliver their intended outcomes.⁵ To improve this track record and create lasting impact, the public sector should:

• Adopt a service lifecycle approach. Government departments often fail to make incremental improvements to systems and decommission them at the right time, instead continuing to use them past their lifespan which reduces their effectiveness while driving up cost. The cost to replace is usually clear, but the opportunity and risk costs of keeping aging systems are usually opaque. Government departments should consider how services will be moved through to obsolescence right from their inception, optimising the services along the way to maximise their impact. This reduces risk, makes it easier to adopt new technology, and delivers better value for money for taxpayers. Systems shouldn’t continue to be evaluated just for the limited functions they perform, they should be evaluated for how well they meet the emerging need and opportunity in their domain.
• Cultivate a transformation and learning culture. Successful digital and AI transformation is as much about mindset and behaviour as it is about technology. In particular, it requires effective collaboration and making transformation a collective endeavour, not just the responsibility of departmental digital and data functions. By fostering knowledge-sharing initiatives across departments, supported by industry experts and academia, government can accelerate the transfer of effective use cases and best practices for the implementation of emerging technologies. Developing long-term partnerships with tech providers and academic institutions will ensure access to cutting-edge skills and solutions that can be implemented, from private sector use cases for public sector benefit. Consider every digital transformation as the start of a learning journey. The data the systems can and will generate are the basis of the future knowledge of the organisation.
• Develop sustainable skills and generational compounding of learning from data. With only 6% of civil servants in digital and data roles,⁶ government needs to address its skills gap and build its digital capability so it can identify and respond to new trends, efficiencies, and service improvements as they arise. This can be achieved through a three-pronged approach to skills development – establishing a digital skills academy for continuous training, implementing a comprehensive knowledge retention strategy embodied in the data you collect, and creating platforms for cross-departmental expertise sharing.

Approaching change with confidence

Success in digital and AI transformation requires attention to all three dimensions. Clarity provides direction, certainty drives effective delivery, and continuous optimisation ensures long-term impact.

By focusing on these elements, the public sector can move beyond the pattern of underperformance that has characterised many digital and AI initiatives and deliver meaningful innovation and growth.

If you’re interested in learning more about how Baringa can help you unlock the potential of digital and AI to deliver better public sector outcomes, get in touch with us today.

^{1. State of digital government review - GOV.UK}

^{2. The challenges in implementing digital change – National Audit Office}

^{3.State of digital government review - GOV.UK}

^{4. Major projects in government – Institute for Government}

^{5. Government's approach to technology suppliers: addressing the challenges - NAO insight}

^{6. The challenges in implementing digital change – National Audit Office}

^{7. State of digital government review - GOV.UK}