I recently had the privilege of attending DataConnect 2025, a truly inspiring event focused on how data can deliver real benefits for users. The conference was jam-packed with actionable insights and practical examples that data professionals can take straight back to their roles. It wasn’t just informative, it was energising, offering a wealth of ideas ready to put into practice.
As a Data Specialist working on public sector projects, I’m aware of some of the negative preconceptions around large scale government projects.
However, my experience has shown me a far more positive, nuanced and surprisingly human story, shaped by complex design challenges, a deep focus on user needs, and cultural shifts that are redefining how data serves the public.
Smashing the stereotypes
When we consider government data projects, the public often imagine slow-moving bureaucracy, siloed departments, and initiatives mired in red tape. It’s a familiar trope but an outdated and inaccurate one. The Department of Science Innovation and Technology sees data as essential infrastructure for monitoring government, as vital as roads, railways, or broadband. It is the insight that shows Government where services are working and where they are not.
In my view, the biggest challenges in public sector data aren’t technical—they’re cultural, psychological, and strategic. At the conference, five distinctly human challenges stood out, each showing how Government Data Specialists are taking deliberate steps to improve services for users. I’ve illustrated each challenge with a quote shared during the event.
Challenge 1. The “Expert vs Everyone” Trap: Why One Size Fits None
“Serving deep expertise and broad accessibility is a strategic design challenge, not just a technical one.”
One of the hardest problems in public sector data design is creating products that work for users with vastly different skill levels. How do you meet the needs of a specialist analyst and a member of the public using the same information?
At the conference, two public health tools illustrated this dilemma. Fingertips is a comprehensive repository built for expert analysts, while Health Trends, created during the pandemic, offers simplified summaries for policymakers and the public. This “split model” forces users to self-identify their expertise, often resulting in a fragmented experience.
Rather than applying a quick technical fix, the team is pursuing a strategic redesign. They’re mapping the entire user journey across five stages: Discovery, Exploration, Interaction, Deeper Investigation, and Application. The goal is clear: move beyond the “Expert vs Everyone” divide and create a seamless experience that flexes to different levels of expertise—without sacrificing depth or accessibility.
Challenge 2. The Ethics Paradox: Experts Want Guidance, But Can’t Use It
“If the tool is unusable, the policy fails.”
When the government reviewed its Data Ethics Framework, it uncovered a paradox. Technical and data professionals were the most engaged audience, but actual uptake was low. Why? It uncovered that the issue wasn’t interest; it was usability. Professionals described the guidance as “hard to use” and “vague.” More critically, the framework, last updated in 2020, hadn’t kept pace with the explosive evolution of Generative AI. The world changed too fast for the guidance to remain relevant.
The lesson is clear: for data governance to be effective, it must be designed as a product with its own user experience, not merely as a compliance document. If the tool is unusable, the policy fails.
Challenge 3. The Biggest Barrier to Sharing Data Isn’t the Law, It’s Fear
“Risk aversion and fear of getting things wrong are more powerful than any legal hurdle.”
What stops government departments from sharing data? It’s not technical incompatibility or restrictive legislation. The most widespread barrier is cultural: risk aversion and fear of getting things wrong.
Data protection rules are often blamed, but the real issue is interpretation, not the rules themselves. A systemic lack of trust between organisations creates a chilling effect when handling sensitive data. This cultural barrier is far more potent than any technical or legal hurdle, reinforcing the silos that data sharing is meant to break down.
Challenge 4. The Mindset Shift: Data as a Reusable Asset, Not Static Evidence
“The shift from data-as-evidence to data-as-an-asset is the single most critical enabler.”
For decades, data in the public sector was treated primarily as a tool for a single, specific purpose. Reflecting on her 25-year career, the new Government Chief Data Officer, Aimee Smith, described the old model where data was treated as “evidence, intelligence, forensics”—collected for one case, one report, or one investigation and then filed away.
Today, that perception is undergoing a fundamental transformation. The modern understanding is that data is a valuable, reusable “asset” that can fuel innovation, improve services, and be repurposed for multiple uses. This shift from seeing data as a static piece of evidence to a dynamic asset is arguably the most important cultural change happening in government.
Challenge 5. Success on the ‘Hard Yards’ Relies on Courage, Faith, and Chocolate Fudge Cake
“The crazy will still be here on Tuesday.”
Large-scale programmes like GOV.UK One Login reveal the human side of data strategy. These are no quick fixes but rather a series of long, grinding “hard yards” that test resilience. Natalie Jones is Director of Digital Identity for the Government Digital Service (GDS) and is the senior responsible owner for the One Login for Government Programme. At the conference, she described a leadership model built not just on project plans but on emotional intelligence as well:
- Courage: Defending the team from pressure to “do it faster and cheaper,” and explaining why some things are better done slowly and well.
- Faith: Believing that if you do the right things and do them well, the outcome will follow even when success isn’t immediately visible.
- Chocolate Fudge Cake: Recognising when a team is exhausted and needs empathy, support, and a break, because “the crazy will still be here on Tuesday.”
These five challenges reveal a central theme: for all the talk of technology, AI, and platforms, the real work of data transformation in government is deeply human. It’s about understanding user needs, overcoming cultural fear, changing mindsets, and leading with empathy.
This raises the question of then how should we train the next generation of public sector data specialists? It’s clear to me that we need them to always consider the human element. As one of the speakers said:
“The most sophisticated data model is useless if people are too afraid to share data or too burned out to see the mission through.”
