Why I am excited about the Performance Review of Digital Spend

By Giulia Cirri

I get it. Funding models are not the hottest part of working in digital services. But funding is how you get stuff done. And how the funding process is structured, the hoops you and your team have to jump through, affect how quickly you will be able to get your thing done, how long you will be able to sustain it, whether you will be able to improve it, and how success is measured.

But the impact of funding models is even larger. They contribute to shaping what is considered  feasible. Cost-benefit analyses influence what is accepted into the realm of the possible at the very early stages of a programme or project. Funding models shape (and reflect) culture, risk appetite, and ambition. Still, the current Green Book methods do not bring out the best of digital delivery, and neither do they scrutinise it well enough.

Current funding methods are inadequate for digital services. 

This might be pretty evident to anyone who has worked on a digital project in the UK Government. I stumbled upon it early in my career: one of my first roles in a government team had to do with defining costs for the Outline Business Case of an ambitious digital programme to improve cross-government data exchange. I vividly remember trying to sketch out the programme’s costs for the next 5 years while we were still in discovery. We did not really know what shape the programme would take yet, let alone the tech, roles, or subscriptions we would need for the next half decade. What we knew was that it was a topic worth exploring, something that had the potential to significantly improve users’ lives. And the pathway to get it done was a 5 year business case based on Green Book guidance¹The Green Book is guidance issued by HM Treasury on how to appraise policies, programmes and projects and the recommended framework for developing business cases – see ‘The Green Book and accompanying guidance’ and ‘The Green Book: appraisal and evaluation in central government’.

At the time, trying to assess a digital programme through Green Book methodology felt like trying to fit a square peg in a round hole. With the years, that feeling grew. Digital is iterative: it needs flexibility and the ability to pivot as teams learn and iterate. It needs accessing relatively small amounts of funding easily, closing the tap as quickly when results are not met, and ongoing funding for continuous improvement. 

Green Book methodology instead requires you to outline full costs, benefits, and requirements for several years very early in the process with a level of certainty that is unrealistic for agile delivery. The volume and certainty of the evidence requested is particularly prohibitive for projects that use technology for which the benefits are still speculative, like – dare I say it – AI. Not only: Green Book processes are cumbersome and can slow down delivery – especially when they are applied to small changes to existing programmes or at early stages, such as discovery or alpha. Green Book guidance also underestimates the importance of ongoing funding for people (with persistent teams that ensure continuous improvement and maintain feedback loops with users) and tech (to avoid risky legacy systems).

Don’t get me wrong: I am a fan of the Green Book. I just do not think that it works well enough for digital services. We all want funding to be assigned based on evidence, but current funding methods focus a lot on cost-benefit analyses before a project is started, and not enough on continuous assessment against metrics as the project progresses. Although Green Book guidance points to creating monitoring and evaluation mechanisms, ‘most business cases for digital transformation projects do not include agreed outcomes metrics for tracking benefits realisation’²Performance Review of Digital Spend, p.10.  I want to see teams investigate a problem, try out a solution quickly and receive more investment if it is working, instead of spending ages collecting evidence and crafting an Outline Business Case for a project that might change and evolve in the near future. 

This is not about letting digital teams off the hook, and going ahead on a mad spending spree financing anything that has an ‘AI’ label slapped on top. It is about designing funding models that suit how digital projects actually work and are able to scrutinise them at the right time.

This government seems to be paying attention to the divergence between digital programmes and funding methods. 

The State of Digital Government Review and the Blueprint for Modern Digital Government published in January 2025  set out the challenges of current funding models and the intention to reform the government’s  approach to funding digital and technology³At the time I summarised both documents here.. 

The Performance Review of Digital Spend, which came out last month, has kept the effort going. It sets 4 funding models to test with a ‘portfolio of pathfinders’ from Digital, Data and Technology (DDaT) projects funded through the current Spending Review (SR25).

Summary of the four funding models, Performance Review of Digital Spend, p.5

There is also a considerable amount of effort into improving training and guidance for civil servants developing DDaT business cases, and a plan to improve outcome metrics and evaluations for DDaT projects. 

Why am I excited about the Performance Review of Digital Spend?

I really like the spirit of this Review: it focuses on reducing the details, evidence, and time required for business cases, especially smaller ones or those that use innovative DDaT initiatives. Instead, it introduces more flexibility to move funding across programs and, and this is exciting, focuses on outcomes and metrics as a way to understand whether funding should progress. 

Currently, every digital investment is put through the same business case approval process, regardless of scale or risk, leading to inefficiencies and delays in delivery

Performance Review of Digital Spend, p.15

In particular, there are five themes that I was really happy to see:

1. Staged funding and increased attention to outcomes

At the cost of sounding like a broken record, business cases often require a level of certainty which is unrealistic and can slow
down service delivery.  At the same time, there is not enough attention on what has happened after the money has been spent. This imbalance limits the opportunity to test things quickly and reduces the scrutiny over services once they have been delivered, which in turn lowers incentives to properly maintain and improve services. 

What to do about this? The first two of the four funding methods are all about staged funding: 

• Staged funding for innovative technologies focuses on those cases where high levels of uncertainty make the current evidence requirements even more unrealistic. The first round of funding is assessed based on a more hypothetical analysis, and funding is then released in stages as and if the work demonstrates progress. Otherwise, the tap is closed and funding  is reallocated.
• Staged funding for live services instead focuses ‘on robust and transparent performance management’⁴ Performance Review of Digital Spend, p.17, linking together change and run funding, and using shorter agile business cases and regular reviews to unlock further funding.

These funding methods can help bring out the iterative nature of digital. In new services, it is about pushing the art of the possible, pivoting, and failing fast and cheap. In live services, it is about maintaining open the feedback loop with users and continuously improving the service. The reliance on performance management and metrics ties in with the rest of the Review, which includes several initiatives to focus on outcomes metrics and evaluations, such as working with departments to develop new metrics for tracking the benefits from DDaT investments, and the development of evaluation plans for a few high value DDaT investments.

2. More flexibility through portfolio based funding

The third funding model blends programme and portfolio approaches, allowing departments to create one single multi-year business case for their portfolio, and then shift funding between smaller projects in the portfolio. This would be really helpful to decrease the barriers of creating large business cases for small projects or small changes to existing projects, allowing departments to move money around as their projects progress and priorities might shift. 

3. Risk
reduction in
technical
debt and
cyber
security

While as any person in government, I am aware of the ubiquitous presence of legacy tech across government, I am going to come clean and confess that this is not my area of expertise. But the State of Digital Government Review really drove the point home about just how big of a presence and a risk legacy tech is across government. So it is really good to see the government not forgetting about the level of tech debt that we currently have, although  this fourth funding method is the one with the least amount of details in its implementation milestones⁵ Performance Review of Digital Spend, p.18 . 

4. More and clearer guidance for creating business cases for agile projects

It won’t fix everything, but clear, accessible supplementary guidance can help civil servants make quick progress in adapting Green Book rules to agile delivery. The Review also sets an aim to publish an example
of a lifetime
appraisal for a DDaT
investment, and strengthen the requirements for departments to publish business cases. Sharing the good work that is already being done in government will at a minimum diminish the time cost of business cases and its impact on delivery. 

5. More work to investigate the RDEL – CDEL imbalance

Last, the Review announced more work to investigate the increased use of RDEL rather than CDEL for digital projects⁶ The UK government’s funding model works around two types of expenses: capital expenses (CDEL) mostly covering one-off spending on assets, such as infrastructure; and resource expenses (RDEL) covering day-to-day, recurring expenditure, such as salaries. . The move towards Anything-as-a-Service (XaaS), meaning buying software as a subscription rather than purchasing physical tech, has driven up the use of RDEL expenditure. This is creating budget problems for departments, because CDEL budget is usually easier to secure⁷ State of Digital Government review, p.39.

There is another reason why I am personally invested in the government investigating the RDEL-CDEL imbalance: permanent staff is cost under RDEL, while consultants are under CDEL. This means that departments sometimes resort to consultants because capital funding budgets can be easier to secure versus budget to hire permanent staff. I do believe there are situations where consultants are useful: we can bring an outside perspective, help navigate internal politics, bring expertise that is not available in-house, and complement in-house build teams for short delivery periods and then handover to civil servants for maintenance and running. But as a consultant, taxpayer, and user of public services, I am fully okay with being employed where I am actually useful, and not just because a department has more CDEL than RDEL budget. So I am really happy to see further investigation into an issue that is creating barriers to funding while costing the government a lot of money. 

This is not a full review of the Review. This is merely a list of things that made me excited, and it is also definitely biased by my experiences of working in government. There is a lot more stuff in the Review that can make someone else happy, or furious (and I am keen to hear about either). Mostly, I am happy to see interest in moving towards funding models for digital services that feel a bit more like fitting a round peg into a round hole. Or at least an oval one. 

As a professional in the field, and as a user of public services, I am very keen to follow the developments. Next stop: GDS’ multi-year Digital and AI Roadmap this summer. 

• 1
  The Green Book is guidance issued by HM Treasury on how to appraise policies, programmes and projects and the recommended framework for developing business cases – see ‘The Green Book and accompanying guidance’ and ‘The Green Book: appraisal and evaluation in central government’ ↩︎
• 2
  Performance Review of Digital Spend, p.10 ↩︎
• 3
  At the time I summarised both documents here. ↩︎
• 4
  Performance Review of Digital Spend, p.17 ↩︎
• 5
  Performance Review of Digital Spend, p.18 ↩︎
• 6
  The UK government’s funding model works around two types of expenses: capital expenses (CDEL) mostly covering one-off spending on assets, such as infrastructure; and resource expenses (RDEL) covering day-to-day, recurring expenditure, such as salaries. ↩︎
• 7
  State of Digital Government review, p.39 ↩︎