A short(ish) history of the NHS design system, Part 2: 2017 — a website redesign, and an app

Dean Vipond
6 min readMay 9, 2022

With all this activity going on, it was decided that the NHS Choices name was to be retired, and the website as a whole be redesigned. It was clear that as more of the system was undertaking ‘digital transformation’, that the main website would increasingly become a starting point, or waypoint, in longer service journeys for people. I started at NHS Digital in the middle of 2017, and joined the people looking at this.

We did research with people to understand what their views were of the NHS, and also how they felt about doing more things online. Unsurprisingly, some people welcomed the convenience of being able to manage more of their health and care digitally, whilst others expressed concern that they would lose the human contact which is at the core of the NHS (simplistically, some people viewed ‘more online stuff’ as meaning ‘fewer nurses’). A previaling theme though, was of people’s opinion of NHS staff. Even when people said they had had a poor experience of the NHS, they would still say things like “but the nurses were amazing”.

Of course, a view like this isn’t universally held, and varies significantly across different cohorts of the population. But it was clear to us that as we approached the redesign of the site, it needed to express the humanity at the core of the NHS constitution, and feel like it had been made and run with the same care and attention people expect from in-person health services. This would be inherent in every aspect of the site — the graphic design, content design, frontend development and engineering.

Making it accessible

As we established the ways of working for this, we put accessibility at the centre of our design approach. On the NHS Choices website, the level of accessibility was inconsistent. We set up training for the whole team (not just the designers and devs), so everyone understood what we were aiming for, and why this was a critical success factor.

Whilst the services being trialled were much more reflective of new ways of working (by being research-led, and building on the foundations of the GOV.UK design system), what they lacked was consistency. Each team, quite rightly, was working out good solutions for their users. But as these services, and the NHS website as a whole were beginning to converge, it was time to do the hard work to align everything.

Three different web page templates, showing the different styles being used before the design overhaul

Blue rinse

The design of the NHS website would not be successful, if it were just GOV.UK with a different font and blue buttons. The NHS is a brand in its own right, and people’s relationships with it are different than with government. And we needed to reconcile the broad range of content on the NHS website with the proven, functional approach modern digitally-enabled services require.

Then followed a long, iterative period of art direction, research, consistency exercises, difficult decisions, negotiation with stakeholders, negotiation with service teams, and so on.

While we knew a re-skin of GOV.UK would not result in a successful design, we did know and value the huge body of testing and iteration that had gone into the GOV.UK design system. We were keen to build on their foundations in order to:

  • accelerate delivery
  • be consistent with other public services
  • have a solid foundation of accessibility.

We spoke extensively with colleagues from GDS about our intentions, and they were extremely helpful and supportive.

My users versus your users

What has always been the case with user-centred design, is there is never one, unequivocal way of solving a problem. But when two different service teams have spent a lot of time and effort researching and refining a design solution (for example, a method of navigating between pages), being able to arrive at a single solution that could be used by both (or more), was challenging. Part of my job was brokering that change, and helping people understand why their very successful solution needed to change for the greater good.

Service teams would often say, “Your users might need that, but our users responded better to this.” It takes time to bring everyone around to the idea that there is no such thing as “your” or “my” users, but that we all have the same users, who may be moving in and out of multiple services and areas of content at any one time.

This is where decision-making based purely on user research can paint people into a corner. Usability testing, for example, will never prove unequivocally that a design solution is perfect. You will always be limited by multiple factors, even if your participant base is diverse. What user research does, is equip people with enough insight to make a good decision. You cannot hide behind user research as a way of abdicating responsibility for a product decision. Research is an indispensable tool to understand a problem space, and evaluate a solution. But at some point, someone needs to make a decision, and be accountable.

How to make friends and influence stakeholders

I learned a lot in this time about socialising work, and your intent, within an organisation. This work would impact a lot of national-level products and services, and (at the time, we were hoping) at local level too. We aimed to show what could be done, and ensure that it was meeting the needs of our users, but was also satisfying the aims of our stakeholders. These things aren’t always necessarily in sync, in any large organisation, and it takes time to win the trust of people whose necks (and budgets) are on the line.

On reflection, I would have done more to allay anxieties on progress, and to reconcile things stakeholders would like to see, with things the team needed to get on with. At the time, trust was low (we were a new team), and visible progress was in short supply in the early days of the project. We knew, for example, that just 2% of NHS.UK’s traffic came via the homepage, so that was much lower on the team’s priorities, than getting the typographic scale just right. But a mocked-up homepage would have probably expressed the overall design vision to people, and secured the team more trust in advance. Likewise, such a visual would have given stakeholders an artefact to use in presentations to people they were accountable to.

A smaller team (including myself) had previously released a new homepage concept on the now-retired NHS beta website, which focused more on search. But we could not progress this at the time, as the on-site search functionality was poor, and couldn’t be addressed without significant work (which has now happened — on-site search on the NHS website is now very good).

An early design of the NHS website homepage, presented on an iPhone 5. It has a blue background, and focuses on a prominent search box, with the label ‘Search the NHS site’.
An early version of the NHS beta website homepage

Anyway, the work to retire the Choices brand, and redesign the website began mid-2017 and would continue through to our launch in September 2018. More on that later.

While all that was going on, the Secretary of State announced there would be an NHS App, that would open more people up to health and care services.

Read Part 3 — 2018: Standardisation and launch!



Dean Vipond

Design leadership and human-centred stuff in general. Previously lead designer @NHSDigital and lots of other things. He/Him. http://www.deanvipond.com