No KPIs, no cash: Holding public sector transformation to account

In recent years, the UK government has, rightly, tied its colours to the flag of AI and digital transformation, promising significant efficiency gains and an improved citizen experience. Its action plans, roadmaps and ten-year frameworks all point to a public sector which aims to be faster and smarter by design.

There has been no shortage of commitment. In January 2025, the government pledged to take forward 50 recommendations aimed at boosting growth and transforming public services. By January 2026, the Department for Science, Innovation & Technology (DSIT) reported that 76 percent of those commitments had been met, with the remaining 24 percent in progress.

We can already see the narrative beginning to shift across the public sector. Recent ArvatoConnect data found that in 2025, more than half (56 percent) of decision makers reported undertaking Government-wide digital transformation projects – up from 37 percent reported just 12 months ago.

But activity does not equal value, and transformation cannot be judged on volume of projects or pace of rollout.

Concerningly, more than half (53 percent) of public sector organisations still lack robust KPIs for digital transformation, undermining accountability and weakening value for money. The same proportion have not actively sought insight from their end-user base to inform decision-making, and even fewer (38 percent) have gathered citizen feedback on how well changes are working in practice.

Where outcomes are clearly defined, measured and transparently reported, automation can deliver productivity gains while maintaining auditability, service quality, workforce confidence and public trust.

If public sector contracts are to be genuinely outcome-focused, suppliers must be prepared to share accountability. Risk/gainshare models are an effective way of ensuring that incentives are tied to measurable citizen benefit, not just system delivery. In short: no KPIs, no cash.

Transformation starts with your culture

Internal culture cannot be allowed to sit outside of this measurement framework.

One of the biggest hurdles to digital transformation is internal resistance. Legacy systems are certainly part of the problem, but culture is often the real sticking point. If your people don’t understand why change is happening, or feel it’s being imposed on them, progress slows. Training, honest communication and visible leadership support all make the difference. Where organisations invest in their workforce, bring their employees on the journey, and explain not just what is changing but why, adoption tends to follow.

This challenge is particularly pronounced in the public sector. Almost half (47 percent) report cultural resistance from colleagues as a significant barrier to digital transformation, compared to just 29 percent in the private sector.

Our research shows progress, but not consistency. Almost half (47 percent) of organisations have now secured cross-department input to shape implementation, up 11 percent from the previous year. The direction of travel is promising but internal collaboration is a crucial foundation – and, compared to system overhaul, a relatively easy win – so we should expect to see far sharper progress here over the next year.

Citizen experience really counts

An obvious but undervalued truth is that services must be designed around the people who are supposed to use them. While internal alignment is crucial, the true test of digital transformation is whether services become easier to use. Expectations are shaped by everyday digital experiences in the private sector, and government services are increasingly judged by the same standards.

Digital channels can improve efficiency, but digital‑only approaches risk excluding vulnerable users, increasing repeat demand and eroding trust. Our research shows that most people who self-identify as vulnerable prefer a blend of human and digital interaction. Well-designed automation should support employees to deliver faster, safer and more empathetic services, not remove human judgement from essential interactions.

AI can also help agents deliver better services for vulnerable citizens. Conversational Analytics, for example, can support the agent in identifying vulnerability, enabling intelligent triage, real-time data sharing and earlier intervention.

Many government bodies still aren’t grounding decisions in citizen needs, accessibility considerations or lived experience. For vulnerable users, even small design flaws can create disproportionate obstacles, amplifying anxiety and reducing trust in public services. Building in flexibility, accessibility and human touchpoints ensures no one is left behind as services digitise.

This isn’t just a matter of reputation. Services designed around genuine user behaviour usually cost less and achieve higher uptake. Those designed from the inside out often struggle, no matter how advanced the technology behind them is. Gathering feedback before and after launch, setting clear KPIs and monitoring performance are not box-ticking exercises, they are what turn digital spend into tangible improvements.

Pressure to deliver savings

There’s also a financial reality behind all of this. The Government is increasingly looking to digital transformation to help meet efficiency targets. Automation, better data sharing and AI tools are expected to unlock savings and improve productivity. But those benefits only materialise if the foundations are solid, with clear governance, skilled teams and departments willing to work together rather than in silos.

Technology can open doors, but it doesn’t walk through them. Real progress comes from leadership, discipline and a clear understanding of who services are for. The most successful programmes tend to treat launch day as the beginning of improvement, not the end, monitoring uptake, listening to feedback and adjusting as needs evolve.