The transcript of the episode follows:
Vanessa Schneider: Hello and welcome to the Government Digital Service podcast. My name is Vanessa Schneider and I am Senior Channels and Community Manager at GDS. At GDS, we build platforms, products and services that help create a simple, joined-up and personalised experience of government for everyone. And as part of that work, we maintain GOV.UK, the website for the UK government. GOV.UK is used by millions of people daily. The home page alone is used more than two million times each week. We've been improving how people can navigate the site, taking a user-centred and evidence-based approach. We've previously written about this work on the GDS and Inside GOV.UK blog, and this podcast episode will be your latest instalment in documenting how we launched the new menu bar after an extensive A/B test and how we updated the GOV.UK homepage. It will also take a look at what lies ahead for making GOV.UK as simple as possible for people to use. Joining me to explore this today are Sam Dub and Jenn Phillips-Bacher, who work on GOV.UK in very different disciplines, but part of the same team. Sam, would you mind telling us a little bit about the team and then maybe what you do as part of it?
Sam Dub: We're a team of 14, which in the scheme of GDS and the scheme of government is relatively small. We bring a whole range of different perspectives and expertise to this work that includes designers, developers, content people, researchers, and our job is to make it easier for people to find things on GOV.UK, and my role as a product manager is about making sure we're working on the right problems in the right way. We're getting to the outcomes for users that we want to achieve.
Vanessa Schneider: Sam, thank you for that explanation of the team. Obviously, part of this as well is Jenn. Would you mind introducing yourself and what you do as part of the team to our listeners, please?
Jenn Phillips-Bacher: I'm Jenn Phillips-Bacher, and I'm a content strategist on Sam's team. My focus is primarily information architecture and findability. So as a content strategist in the textbook definition of it, it's all about getting the right content to the right people in the right place at the right time. And that's why a content strategist is working on navigation. It's all about improving that mode of getting users to the content that helps them achieve a goal.
Vanessa Schneider: Great, thank you to both of you. While it would be great if we could count on it, but not everyone will have been following the public journey of this work, even though we've blogged about it extensively. So would either of you mind recapping perhaps what's been happening? When did we start changing where users could find our information?
Sam Dub: One of the challenges for GOV.UK is that the amount of content published grows every year. And today it's more than half a million pages, and it might just be one page in that half a million that a user needs. And so in order to find that page, there are, kind of, multiple tactics that they'll use. They might use a search engine, they might use GOV.UK site search, or they might browse through the home page, through a menu bar, through topic pages, to find what they need. And work on that topic system, making sure that users can browse successfully is the focus of our team. There's work going on elsewhere in GOV.UK in partnership with search engines, and there is work planned to improve our own search engine. But the focus for our team right now is browse and how we get that topic system, these menus, the home page, the breadcrumbs, and related links at a content page level, all working nicely together. So users can browse to find the stuff they're looking for. What you've seen go live over the last couple of months are the first kind of public steps of some work that's been happening for around the last year and those public changes have been changed to the GOV.UK home page and a change to the menu bar. So that's the black bar that sits across the top of every content page on GOV.UK. And that's more than half a million pages of the site. And that is GOV.UK's primary navigation. And we've, we put those changes live, we put homepage changes live in December and we put several versions of the new menu bar live in the second half of last year.
Vanessa Schneider: So it's great that you've outlined what changes are taking place, but why was it necessary?
Sam Dub: So the strategy here is about making stuff easy to find, and like, like all good GDS teams, we started with a discovery, and a discovery essentially means validating a hypothesis about a problem. And here it is about understanding what was going on, why people couldn't find things or why people were abandoning journeys on GOV.UK. And in the course of that discovery, we found 3 core problems that users were facing, and they were a confusing information architecture - now this is this is more Jenn's area of expertise and confusing information architecture is not a phrase that a user will come to you and say, "Oh, your information architecture is very confusing". It will be something that you'll notice a user lost within the site and not able to find where they need to be in order to get to the service or the piece of information that they need. So that was problem one. Problem two were issues on smaller devices. So these days, GOV.UK is used most often on mobile, and last year that was, three-quarters of all visits to GOV.UK came from a mobile device, and not all of GOV.UK's pages are optimised for mobile devices. And so that presented navigational problems, and there's a real opportunity there to move from an approach that kind of worked on mobile but wasn't ideal to something that really feels like it's designed for mobile devices, which is where most users are. And the third problem we looked at was an issue of overwhelm. So a lot of users to GOV.UK feel like it's just a lot of stuff. It's the most common phrase that we hear, the most common sentiment in user research will be, "This is a lot. This is too much. I'm not, I'm not quite sure what I'm looking for here". And so in a lot of the design work we've been doing, we've been looking at how to get back to the core principles of GOV.UK that are about a simple, clear experience to make it easy to find things.
Vanessa Schneider: Sam mentioned that as a content strategist, you, Jenn, might have some experience with how to resolve confusing information architecture or sort of what kind of problems that can cause. Would you like to maybe speak to that, Jenn?
Jenn Phillips-Bacher: Sure. So information architecture is one of those phrases that you'll get a different definition of it, depending on who you talk to and what their experience is. So I think in my work within GDS, I'm thinking about information architecture as being the bone structure of the website. A good information architecture isn't really something that you point to or see. It's something that is that scaffolding that supports the entire information space so that people can find what they need. And, and it all kind of depends on the raw materials that you have. So what kind of content you have, what kind of data and metadata that you have about that content, whether you've got images, video and so on. So I always think about the UX and information architect Louis Rosenfeld. He talks about there being three tracks of information architecture, and this is where it fits in with GOV.UK. He talks about top-down navigation. So that's the things like the global menu and the user interface components you might see on a home page. And what that does is it kind of anticipates an interest or a question that a user might have when they arrive. The second thing is bottom-up navigation, so that stuff, like bread crumbs or related content links or 'you might also like'-type suggestions or any kind of contextual navigation with content - that might mean like titles and page headings, or even links embedded inside of blocks of text. And then the third thing is, is the big one, and that's search. So that's really for handling those really specific information needs. So the information architecture is kind of this interplay between top-down, bottom-up, and search. And it's that whole, that holistic information architecture that we're starting to make some significant improvements on GOV.UK.
Vanessa Schneider: Makes sense that we want to take care of this. So, yeah, menus and topic pages, they must play a big role in the user experience from what you've just shared. But making changes to how users interact with them, that would have made a bit of a difference out of the blue, no? How do you test this effectively without maybe negatively affecting users because it must be a bit challenging in a live environment?
Sam Dub: So that's a really great question, and I think something that we're quite careful about. We know that for people finding, finding what they need it is a kind of critical task and GOV.UK is part of a whole bunch of essential processes in people's lives. And whilst we want to make stuff easier, there would be significant consequences to making stuff worse. And so we start with quite an extensive process of research before we make any change on the live site. We'll develop prototypes and then introduce those to users in a kind of test scenario in usability research. And a lot of, a lot of ideas often don't make it past that stage. You your your expectations about what will work, they kind of improve over time. But still, there's a pretty high ratio of stuff that when it meets contact with a, with a user, you'll suddenly discover unexpected problems with it. And so we try and catch that stuff early. And then when we do introduce changes to the live site, we want to use a kind of experimental method. We want to make sure that the change doesn't look simpler, but it actually works better. And that's where we'll use a technique called A/B testing or multivariate testing. And what that essentially means is comparing the performance of two different designs. We do that by users opting in to measurement on GOV.UK. When you accept cookies on GOV.UK, you have the option to accept measurement. And what that means is that at scale we can see how the site is being used. When we introduce a new design, we can compare how the new design is being used versus how the old design is being used. What that then allows us to do is look at certain key metrics and, for example, a new menu bar, we would want to see that a new menu bar is being clicked on, and that would be one very simple metric to see whether it was being recognised as a menu and whether it was being used. We would want more sophisticated stuff than just being clicked on. We would want to see a good user journey across the site, so users opening the menu bar, selecting an option and then successfully navigating down to a piece of information or a service. And we can look at those patterns of journeys across the new design and the old design and see which is more successful with users. So this process removes the, the choice out of this process, the, the bias out of the process, we just see what works for users. That's what we go with.
Vanessa Schneider: I was wondering, were there any adjustments to what you were testing with users based on how they were responding to your A/B testing?
Sam Dub: The way we will work on a design for an element as important as a site wide menu bar: very rarely will it be once and done, so it'll be a process of continuous iteration where we're looking at data. And sometimes the changes are significant and sometimes they're smaller. But there are quite significant differences from the first version of the menu bar we put live, changes to the content, changes to the interactions. And I would imagine it's continuing to refine that over the next few months. What tends to happen is that at the beginning, the changes are quite major. And then over time, you move into a process of polishing the design and you're making smaller and smaller changes and smaller optimisations. And we're getting that with the menu bar, I'd say we're at a point where there might be a couple of little content tweaks we want to try, but we're pretty much there we think.
Vanessa Schneider: Jenn, how does your insight as a content strategist feed into the menu bar?
Jenn Phillips-Bacher: What I'm thinking about is the understandability of the system as a whole, so can people understand the information that is being presented to them? One thing we look at is understandability of labels. And one of the ways that we can do that is a quantitative piece of research method called tree testing. And what that effectively does is allows us to look at a user's click journey through a hierarchical structure. So our, our existing topic system as an example is hierarchical, you use it to narrow down to a set of more granular categories. So we can use tree testing to understand whether the entry points are understandable for people. Are they going down the right path to begin with? Did they get lost once they're inside of that structure? And in that way we can identify where, where people are getting lost, where we might need to make changes to the language that we're using and where we need further qualitative investigation because we can't, we can't know everything from the quantitative. Often tree testing opens up a whole new set of questions that we actually want to ask humans face-to-face. So it's good fodder for ongoing research, which again feeds into the iterative development of things like the menu bar and the topic system. So it isn't, it isn't right until we know that users can find what they need where they expect to find it.
Vanessa Schneider: And does the tree testing run in the background of the A/B testing or is it something that needs to be set up separately?
Jenn Phillips-Bacher: Tree testing is set up separately from A/B testing, and the way that we are doing it at the moment is through a banner across certain sections of the website. So when we have a tree test that's live, you're likely to see it within the topic structure or other pages within the website.
Vanessa Schneider: Great, so that sounds like a really non-invasive way for you to get data on how people are progressing through their journeys without them needing to reveal anything about their personal details. And obviously, as big proponents to agile, it makes sense that we're having a very iterative approach. I was wondering if that was transferable to other parts of our mission to improve navigation because we've talked extensively now about the menu bar, but obviously, that's not been the only area of activity.
Sam Dub: So another area where we're making these kind of evidence-based changes is to the design of core components of the website. And one place where you can see a change that we're really proud of is on the GOV.UK home page, where you see all the topics listed out, the area of the site that you as a user need to click on is now substantially larger. And you might think, 'Well, it's a small change, and it's just, maybe, a design tweak to make the page feel more, kind of, better spaced out'. The change is much more significant than that and actually comes from a different set of user needs. This is about using the website with touch. So users, particularly on smaller devices, I'm sure we've all had the experience of, like, struggling to tap on a link on certain websites. We've all had that experience, maybe where we clicked on the wrong thing, because when you're trying to tap on a link, it's difficult to tap on the right one. Those problems are magnified if you are on a smaller device or maybe have a tremor or a motor impairment. And that, that can be the difference between being able to successfully complete a task on mobile and basically having to abandon and either use a different device or ask someone else to do it for you. So it's quite an important change for us across the site to make these targets so that you have to tap on or click on in order to navigate through the site significantly larger. And we feel that is a good thing for everybody. It should make the website easier and quicker to use on mobile and more accessible in the process. So accessibility, mobile usability, are core principles of the pages, including the home page that we're going to be redesigning as part of this process.
Vanessa Schneider: Sounds like we're on a roll. This kind of work, we've mentioned it before, is never done. So what is next on the agenda for us?
Jenn Phillips-Bacher: So a key part of simplifying our information architecture is working toward improving our topic system, making sense of what we call topic browse. So in GDS lingo, we often talk about mainstream browse, and that's what you see on the home page at the moment. And it's the topic-based part of GOV.UK that includes the top categories you might use to browse to content like 'passports, travel and living abroad', or 'working, jobs and pensions'. So they're quite broad categories that contain all of the content that supports top tasks, so the things that most users are going to want or need to do. For example, like applying for a passport or checking your state pension age, so very task-oriented. But in parallel with that, we have something that we often call specialist topics or specialist sectors. And they're similar, similar to mainstream browse in a way, but it doesn't have a predictable home in the same way that mainstream browse does. So that means that users, who are looking for more in-depth information relating to their business or industry or they're working in some kind of advisory or professional capacity, can't navigate to that content without using search. And what's interesting about these two systems is that they kind of do some of the same things, they're hierarchical and there's an element of curation. So content designers or publishers are deciding which of the content in those topic areas is most important and should be prioritised for individual user needs. So what we're looking at doing at the moment is consolidating these two topic structures into a single browse layer for GOV.UK. So that will allow people to get to a broader range of content, and it will do a better job at reflecting the full breadth of content published on GOV.UK. We currently have about 650,000 content items, or pages, which is huge. It is huge [laughs]. And we're only really surfacing a small fraction of that within mainstream browse.
Sam Dub: What we're working on over the next couple of quarters is to combine these two topic systems into a single definitive browse system that will allow users to find anything that they need on GOV.UK. And that's a real design challenge because what we don't want to do is overwhelm users. What we want to do is make sure there's a route to everything the people need without it feeling like there is this huge volume of content for them to wade through. And so we're working very hard on design patterns, using some of the latest thinking within GOV.UK Design System, like accordions, like the grid patterns that you have seen on the new home page. And we're working to make sure that there are simple routes to information and services. But there are also the people who need it, ways they can dig a bit deeper and they can get to the specialist content as they need it.
Vanessa Schneider: So if users have become interested in the work that you're doing, is there any way for them to engage with it? And I mean users, whether they are the end-user, as in a citizen that is going through the process of navigating a journey, or even teams that are running services in government that might be more of a middle person, and they want to improve how their content appears to users. What can they do to get in touch with you and how can they help?
Sam Dub: So if you're a user of GOV.UK and you're inside government, if you're a civil servant, you can get involved and and talk to us through the cross-government content community, so there's a cross-government Basecamp that you can join and we'll be doing any call-outs for participation and collaboration across government using that channel. And if you're not a content designer, you can talk to your nearest friendly content designer within your department or the managing editor. Each department and agency, I believe, has a managing editor. And they're a great, kind of, point of contact between you and GDS and our work. You can also keep up to date on Inside GOV.UK. We're doing our best to work in the open on this and blogging about forthcoming changes through there. And if you're a user, and you are listening to the GDS podcast, you, the blog should be a really good sense of what's what's coming up. But we hope also, for you. That this is a pretty seamless experience, that we don't expect these changes to disrupt anybody. They will be iterative over the coming year and gradual. And because of the evidence-based way that we're working in terms of the user experience, it should be one of iterative improvement because we realised that for so many people, GOV.UK is critical to their jobs, to their livelihoods. And we're careful about how we're iterating. We want to make consistent progress, but we don't want to disrupt everybody in the process.
Jenn Phillips-Bacher: Over the last quarter, we've been collaborating with content teams from HMRC and DWP to test some of our assumptions. And those have been really fruitful and helpful discussions because it's demonstrated to us some of the areas where government publishers might have a slightly different perspective to GDS in terms of how people are navigating the site. And we've learnt a lot about users who use HMRC and DWP services. We're also aware that some departments with active user research communities are also doing tree testing, and that's somewhere where we could really learn from each other. Understanding the mental models of users who use content that is being produced by each of these departments. And over the coming months, we will also be reaching out to departments and agencies via managing editors to review the specialist topics or specialist sector pages that you have on GOV.UK. And we will support those teams with making any changes that are necessary in order to get them to fit into the new topic system. If you're currently working in government and working on things like topic systems or tree testing of your own content, get in touch. We'd love to hear what you're working on and what you're learning.
If you're a person who works in content or works with content, whether you're a content designer, UX writer, content strategist, information architect, content architect: GDS is a great place to work and develop your career and we're building up our content strategy practice a bit more. So even if you don't see a job title that looks quite right, please have a look at the job description and think about transferable skills. Our content strategy team is made of people who've worked in fintech, libraries, social media, journalism. So even if you think you don't fit the title, do have a look at the job description.
Vanessa Schneider: So it doesn't get more straightforward in terms of calls to action than what you've both shared with us. We'll have the contact details on Jenn's research with other departments in the blog post, and you can subscribe to the Inside GOV.UK blog by visiting [insidegovuk.blog.gov.uk] and on the right side bar, you can find options to subscribe to the blog. And if you're more in the mood for listening, you can find all episodes of the Government Digital Service podcast on Spotify, Apple Podcasts, and all other major podcast platforms, while the transcripts are available on Podbean, goodbye.
Jenn Phillips-Bacher: Goodbye.
Sam Dub: Goodbye.
To leave or reply to comments, please download free Podbean or
To leave or reply to comments,
please download free Podbean App.