“The most powerful change isn’t technical. It’s cultural. Designing surveys together, not apart.”
Jonathan Sykes, Product owner, esynergy
Data & AI
UCD
Public Services
4 min read
The Office for National Statistics (ONS) is the UK’s largest independent producer of official statistics and plays a vital role in shaping national policy, economic planning, and social understanding. Its surveys reach millions of people each year, gathering essential insight that informs decisions across government, business, and the wider public sector.
As society’s expectations around digital services, accessibility, and trust continue to evolve, ONS recognised the need to modernise the way people are invited to participate in surveys and how they experience the journey from first contact to completion. This ambition built on important groundwork already laid by leaders such as Laura Wilson, whose Responder-Centred Design (RCD) framework helped shift the conversation from data collection to respondent experience. Our work aimed to build on that foundation and provide the first steps to scaling it across the broader survey portfolio.
This ambition was not merely about improving a digital interface or replacing ageing tools. It was an opportunity to start rethinking survey participation as a modern public service, one that feels intuitive across every channel.
Rather than seeing the issues as constraints, ONS viewed the transformation as an opportunity to help create a unified survey experience that strengthens trust, improves inclusivity, and reduces friction for both respondents and the staff who support them. Survey systems, tools, and workflows had evolved over time, shaped by statistical needs and internal processes, but they had not always kept pace with public expectations around clarity, convenience, and digital accessibility.
This presented a chance to bring together service design, operational insight, digital modernisation, and emerging AI capability. The outcome would help create a survey experience that is more intuitive and accessible, while supporting higher-quality data and more efficient operational delivery.
Working across Discovery and Alpha phases aligned to the Government Digital Service model, we built an interconnected view of how surveys work in practice. Discovery focused on listening: to respondents, to field staff, to digital teams, and to operational leads. We uncovered points of friction and ambiguity that subtly eroded trust or increased burden, from confusing onboarding to fragmented case management
During Alpha, we began prototyping targeted improvements across channels. These ranged from clearer communications and more intuitive onboarding flows to more structured routing, redesigned question transitions, and improved support for accessibility needs.
Alongside service design, we introduced emerging AI capabilities to explore how technology could simplify participation. One of the most impactful examples was the SIP (Standard Industrial Classification and Standard Occupational Classification) proof of concept, which enabled respondents to classify their occupation more accurately, reducing cognitive burden and supporting higher-quality data.
A defining characteristic of the engagement was collaboration. We convened statisticians, service designers, digital teams, field colleagues, and operational leaders to co-design solutions. This cross-disciplinary approach created a shared language and shared ownership of the transformation, moving ONS away from channel optimisation toward a truly service-led view.
Data & AI
UCD
Public Services
Jonathan Sykes, Product owner, esynergy
The work highlighted opportunities to reduce operational duplication, streamline workflows, and improve the tools used by field and support teams. Small but targeted prototypes demonstrated how incremental changes could reduce failure demand, increase completion rates, and improve respondent confidence.
Through AI experimentation, ONS saw how new technologies could enhance both user experience and statistical integrity. The SIP prototype showcased a practical way to simplify a long-standing challenge of enabling respondents to classify their job or industry without relying entirely on interviewer interpretation.
Crucially, the engagement left behind a suite of reusable artefacts, journey maps, experience principles, research insights, prototypes, and design patterns, as well as proposing a more agile, user-centred way of working.
Data & AI
UCD
Public Services