Periscope about user research for GOV.UK

Tara Land holding an smartphone with a mirror of herself on it

Watch Naintara Land, Interim GDS User Research Lead at GDS answer your questions live on Periscope about user research on GOV.UK on Wednesday 8 September 2015.


Interviewer: We are Periscoping from Aviation House, from GDS, and our guest today is Tara. Tara, please can you introduce yourself to the internet?

Tara Land: Okay. Hello, I’m Tara Land and I lead the user research team here at GDS.

Interviewer: Fantastic. We know that Periscope isn’t the most perfect tool in terms of accessibility, so can you just tell everyone what we’re going to do about that?

Tara Land: Yes, so we’ll be putting up a video, which is subtitled, and also a transcript of this session – hopefully tomorrow, definitely by the end of the week.

Interviewer: Fantastic. We’ve got a bunch of questions to ask you. Of course, if you’re watching this live on Periscope and you’ve got questions, then you can send them in to us through Twitter, and we will put them to Tara and get you some answers. In the meantime, we’re going to get started with the basics. Tara, what does user research actually mean?

Tara Land: Okay. User research means spending time with the users of our products and services to understand who they are, what they’re doing, how they live their lives, and create a picture of these people who will be using the things that we’re trying to design and develop so we can bring that into our teams and make sure that we’re designing/developing the right things for people. We’ll use a set of established research methods and a range of both quant- and qualitative approaches.

Interviewer: Why does user research matter?

Tara Land: It’s simple, really, because it’s about understanding the problem. If we’ve understood our users and we’ve understood what it is that they’re trying to do, we have defined the problem that we’re trying to solve. That means that we’ve got something really robust to build upon. It means that we’re enabling better service design, and that means higher digital take-up, higher compliance, more effective policy outcomes, reduced user errors, and that means saving time and money.

Interviewer: Lovely. User research is all about understanding user needs, so can you just describe what a user need is? How do you know what one is when you’ve got one?

Tara Land: Okay. I think there’s a kind of acid test, which is very much it needs to be something that a user themselves would say – or certainly, if you were going to read it back to them, they would recognise. We do sometimes come across needs that are a business need but kind of in disguise, and so there’s a real test to do that.

I think a need is very much a necessity arising from a certain problem. That’s how Tomer Sharon describes it, and I really like his description. They’re important because our guiding principle is users first; start with the user needs. That’s the differentiating thing about what we’re doing here at GDS.

Interviewer: Lovely. If I’m working in a government department and I want to do some user research, perhaps for the first time, how do I know that I’m doing the right thing? Where do I even start?

Tara Land: Okay. We have lots of tips and tools, very practical things on our user research blog, so just reading and seeing what’s available is a really good starting point in terms of knowing how to equip yourself. In terms of knowing whether you’re having an impact or whether you’re doing the right thing, I think it’s very much about having an impact on your team, seeing that when you’re doing research that it’s actually affecting design decisions – decisions that the team are making, that it’s changing the language and the thinking within the broader multidisciplinary team that you’re working in.

I think it’s really important that user research isn’t seen as something that happens off in a corner on the side, that research isn’t something that sits on a shelf. Research is something we deliver, and that’s the real measure of whether you’re doing the right thing.

Interviewer: How do we make sure, though, that it’s integrated, that it’s part of everything else? How do we do that?

Tara Land: Okay. Firstly, we believe that it’s really important that the user researcher is embedded so that they are part of the agile team; they’re not somebody who occasionally comes in, or user research isn’t outsourced and happens [in] an entirely different organisation [all] built-in.

Then we bring the team; it’s our responsibility, as researchers, to bring the team on a journey with us. It means getting them to come, and whether we’re in a lab to observe testing or come out on field visits, to be there when we’re doing analysis so that we create a shared understanding of what we’ve learned from our users and what that means.

Again, user research isn’t something that the researcher reports back and that they own; it’s that our understanding of our users is a shared kind of experience. Everybody has a responsibility for understanding their users, but we facilitate that through research.

Interviewer: User research is a team sport?

Tara Land: It is a team sport, very much so.

Interviewer: Yes, okay. We’re going to be doing some more stuff about user research on Twitter for the rest of this week, because we’re having a whole week of user research stuff. Can you just tell our viewers today what’s coming up tomorrow?

Tara Land: Tomorrow is really exciting because there are two very lovely video clips of my colleague John Waterworth, who is great, showing people around our user research lab here. So, talking both about the observation facility, so very much touching on some of the things I’ve just said about involving the team in terms of very much being part of the research, and then also talking about the research studio where our participants come in and discussing the various ways in which we use that.

Interviewer: Fantastic. We’ve just talked about the lab, then. Can you talk a bit more about the lab? You’ve spent a lot of time sitting in it; tell us a story of one of the favourite things that you’ve seen in there or something that’s hit home while you were in there.

Tara Land: Okay. I think it’s probably important to say that we’ve got a lab. We do lots of other things apart from just lab-based research. Often in the lab, or just doing research in general, the experiences and stories we hear can be often very moving, sometimes quite harrowing in terms of what real people tell us about their lives and where their needs aren’t being met.

Certainly through the lab-based testing, I think we have countless stories here across GDS about where, because we’ve been iteratively feeding research findings into the design, we test something and we see something clicks and shifts entirely. Somebody speeds through a process and sort of says, “That was really easy.” I think that’s incredibly inspiring and motivating, not just for the researcher but for the entire team that’s come along on that journey.

Interviewer: What do those moments feel like?

Tara Land: There’s a real sense of achievement and fulfilment to feel that you’ve done your job in terms of representing the user, being the user’s voice, that you have made a difference, that they have been able to do something that often at the very start they don’t think they will be able to do.

Interviewer: Fantastic. Don’t forget, if you’re watching, you can send in your questions via Periscope or Twitter, and we will pose them to Tara right now. This is your one chance, internet; if you want to ask Tara a question, now is your chance to do it, because we’ve only got a few more minutes of this Periscope session. I’m going to ask a couple more; then we’ll dive into some that have already been submitted. Tara, what does a user researcher need to be good at?

Tara Land: There’s the kind of technical stuff, and by ‘the technical stuff’ I mean research skills, so things like moderating, usability testing, being able to ask open, non-leading questions, being able to analyse. Particularly with qual data, it’s incredibly challenging because there’s a vast, and very rich, and deep body of data, and to be able to pick out and see the patterns in it, and make sense of it all, and bring these stories back.

But then I think a really key part of our job is to create empathy in the teams that we work in, so it’s not just that we’re reporting back; we’re very much bringing these narratives and creating genuine empathy. It’s not a functional sense of “This is what users need,” but a sense of why and where this is really going to enrich their lives or add value. So, as a user researcher, being able… Having those skills to tell those stories with the emotion that often we hear them in, so to accurately be able to pass them on.

Interviewer: We’ve had a question in from Ian Callaghan. Ian says, “What’s a good way of getting into user research if you are changing career?”

Tara Land: Okay. It’s a slightly tricky one to answer, because it depends on where you’re coming from and what other kind of experience you have. I think that we have researchers in our team who come from a whole host of different backgrounds. We have people who’ve been product managers; we’ve had people who’ve been accountants, [come from a] finance background. We’ve got people with a more traditional research background, with psychology degrees or digital anthropology, so I don’t think there’s necessarily an academic kind of prerequisite. I think it’s very much those attributes of being curious and insightful, being able to see patterns, being able to think and see the big picture, as well as hone in on the finite detail. To me, those are really key skills, and then being able to communicate them well.

Interviewer: We’ve had another question in from Andrew. Andrew says, “How can you practise openness in user research, without suggesting premature conclusions or giving away user identity?”

Tara Land: I think there are a few questions in that question, so I’ll take what I’ve kind of understood. In terms of practising openness, to me that sounds like sort of transparency and sharing findings. I think we believe very strongly here – and it makes a big difference in terms of achieving the impact that I was talking about before – of bringing everyone with us, bringing our teams with us when we’re doing research. We’ve had the Director of GOV.UK going accompanying contextual field visits to people’s homes, to listen to them telling their stories about their childcare needs, bringing people with us so that everybody feels like they have a rapport with their users.

Premature conclusions: again, the way that we do analysis here, firstly we do it together with the team. Everybody should have participated in research or been in a lab and observed what’s happening, and we get together and discuss what happened. We use a particular kind of approach; we use affinity sorting, which allows us to remain really grounded in what we actually observe, what we saw users doing, rather than immediately thinking, “This is what we do now.”

So, by affinity sorting and using that particular approach, it means we build up that kind of rigour; we’ve got a really clear rationale. The findings throw out the solutions, rather than we try and fit the findings into a solution that’s already been decided.

Interviewer: Brilliant. We’ve had another question. This is from Tariq, who says, “I’d like your view on how wide you scope your UR context, where wider equals more work but potentially deeper, better service.”

Tara Land: I think Tariq is asking about whether we do contextual research. We’ve talked quite a bit about the lab [we’ve mentioned there], which we use primarily for usability testing, which is the particular methodology, but we say “We have agile projects, but we have discovery at the start.”

During that time, we go out into the field and we learn about our users. We do proper contextual research. We go to people’s home, or place of work, or whatever the environment/the relevant context of use is, and we very much focus on both – what they’re doing, but pick up on all the tacit clues that you only get by going out into the field. That’s really, really important and that’s what creates a really solid foundation that we can then begin to design on.

Interviewer: We’re zipping through our time and we’ve got a whole pile of new questions that have just come in. This one is from another Andrew; this one is, ‘How can you know your sample is representative of the wider population and not just laboratory behaviour?’

Tara Land: Okay. Often we will use core methods, so our lab-based testing tends to be with a smaller number of users, but we are testing in every sprint, so across the project lifecycle we’re building up an incredibly rich understanding, but with numbers of a particular group of users.

We also very much work with our performance analyst colleagues to look at the data that we have. There are various other sources of information. On GOV.UK, which is where I’ve been for most of the last two years, we have the anonymous feedback that’s sent in, as well as Google Analytics. We have other sources of data – our satisfaction survey, for example; so, we can really triangulate what we know about audience, using both quant and qual sources.

Interviewer: Lovely. Here’s another one. This is from someone on Twitter called @Mutopia: ‘What design change did you make which felt counterintuitive?’ That’s an interesting one.

Tara Land: That’s a really interesting one. I don’t know if I can think of a… I can think of one now. Recently – it’s not a design change that I made, but it was somebody in my team – my colleagues Cath and Dipa, who were working with the team, looking at the work that we need to be doing to GOV.UK to get ready for the election. We were looking at how we represent policy, particularly with the new government coming in. In the past, it’s just been deleted; we delete the entire legacy of the previous government, and so we did something radically different this time and looked at how we could have that up there and archive it.

That research also threw up lots of questions about who are using these kinds of pages and how we were grouping them, so we actually made some quite… Based on that feedback, we made some changes to how we grouped and presented policy and topic areas that felt like it wasn’t necessarily how we modelled, how we were modelling or thinking how people were using it, but the research showed that there wasn’t a clear user need for what we originally designed for.

Interviewer: Okay. We are already way over our time, but we’re going to steam ahead and do some more of these questions. Thank you, internet, for sending in your questions; these are lovely. Another Twitter user – this is from @Mirso – says, “What would be a good first step to get started. We are quite daunted; how do you get over being daunted?”

Tara Land: I think you can start small, and I think it’s important to understand that it’s better to do something than not do anything. If you don’t have a lab, that doesn’t mean that you can’t talk to users; there are ways to get access to real people – so, phone interviews.

We often do remote user testing, usability testing, because we’ll have user groups who are accountants or lawyers and cannot come to our office and come and sit in our lab for an hour in the middle of the day. There are lots of ways that you can reach users and actually have conversations with them, get feedback on the stuff that you are building, without feeling like it’s a massive investment in terms of money and time.

Interviewer: Briefly, though, how do you find those users in the first place? Where do you reach them?

Tara Land: It depends on who they are. Often for the users that we would describe as ‘mainstream citizens’ – so, people who are just trying to do things that everybody has to do, so whether it’s renewing their tax disk or passport – we’ll often use an external recruitment agency to ensure that we’re not biasing the sample. That feels important to do that we’re not doing it through our own networks or contacts, that it’s an externally recruited group of people.

Sometimes when we have very niche users who are a certain kind of professionals, we might use contacts or networks that our colleagues across government in departments and agencies might have. I’m trying to think of an example for that. Last year, when we were transitioning all the arm’s-length bodies, we relied heavily on some of the more niche agencies, like MHRA, to put us in touch with people to find people who regulated medical devices. That was quite a niche user group.

Interviewer: Brilliant. Let’s move on. We’ve got loads of questions coming in now. This is from Anna: ‘How do you source or find’ – we’ve just done that one, in fact – ‘How do you source/find customers, do you use research?’ Sorry, Anna, we’ve just done that one; we’ll move on. @Mutopia again; Tanny says, “I’ve learnt from this that one can be a digital anthropologist. What is that?”

Tara Land: I’m not a digital anthropologist; I’m probably not the best person to answer that question. There are courses, Master’s courses, at various institutions which do digital anthropology. We have a number of people here with that kind of background who are great.

Interviewer: Okay. Here’s one from Hugh Prichard: ‘Is the user need sometimes analysed too late, like when the project has already been designed?’

Tara Land: Yes, and it’s really frustrating, but that doesn’t mean that you give up; it means that…

Interviewer: You iterate.

Tara Land: Yes, exactly, it means that you throw it away. That’s one of the great things about working in agile here – that we should be able to make something and be willing to throw it away if it’s not right. It sounds like a radical thing to say, but yes.

Interviewer: Okay. Another Twitter user, @Salamisa, says, “How can you retain insight for future use by your projects and others across government?”

Tara Land: This is something, actually, I feel really strongly about, because often research… I think it’s really important that what we learn as researchers isn’t limited to the particular thing that we’re building, that we’re feeding in. In government there’s very much a strong case for that where we’re not making… The user needs themselves aren’t disposable.

Part of our approach is to focus, to scope, to make sure that when we’re scoping research we’re scoping it around the user and the user needs, not the product. We’re learning about the users, because their needs are stable; users will be continuing to get married and have children 15/20 years from now. government and the way that it delivers its services and policies will change, of course, so scoping research by the user, rather than the product, is a kind of fundamental way of doing that.

Then we’ve been working very closely with our colleagues across government to share knowledge and understanding. We’ve started to build some maps of key life stages, key transitions – so, maps around approaching retirement or experiencing a bereavement, so something again that’s a stable experience or need that’s not defined by the particular kind of socio-economic climate of a particular point in time.

Interviewer: Okay, I think we are going to have to wrap this up pretty soon, so just one or two more. This is from Twitter, from @JazzElevator, who says, “How do you deal” – I love this one – “How do you deal with internal voices that think they know better than the user?” Those voices in your head.

Tara Land: Yes, and in your team. I think I hear where @JazzElevator is coming from. This is very much what Alan Cooper talks about: the notion of the ‘elastic user’ and the user is whoever the loudest person in the room is. It’s really annoying and you end up just doing design by committee, which is never a good thing.

It’s one of the reasons why we feel so strongly about user research being a team sport, which is why we pretty much demand that everybody gets involved. We have a sort of mandate here where we say, “People should have two hours of exposure to users every six weeks.” It feels really important to have that expectation set of everybody, to say, “Have you had your two hours? What are you doing to make sure that you’re keeping yourself in touch with the reality of the users, rather than seeing things through whatever particular lens you’re bringing.”

Interviewer: Fantastic. Let’s do one last one, and then we’re going to stop. I hope your voice isn’t going. This is another one from Andrew Palmer: ‘What’s the shelf life of research data? Is it more than six months?’ Sorry, not six months, six sprints.

Tara Land: Six sprints. It depends. This goes back to what I’ve just been saying about working with government, scoping research around the user and user needs rather than the product. That kind of stuff, if we’re doing research around something that is a life event, that is a thing that will continue to happen, I think that it has a longer shelf life. Of course, we have to do lots of interaction, level focused things too, to make sure that the kind of services and products we’re building are usable and will work.

There is lots of work going on by my great colleagues Caroline Jarrett and Tim Paul, in particular, around design patterns, which are very much research informed; they’re all based on research to establish these patterns. Even though you might have learnt something on one particular project, your findings, your learnings live on because we’ve got them at pattern level too.

Interviewer: Okay, so just one last reminder: what are we doing about the accessibility?

Tara Land: We are subtitling and transcribing this session, and it will be available hopefully tomorrow, definitely by the end of the week.

Interviewer: Lovely. Tara, thank you for your time.

Tara Land: Thank you.

Interviewer: Internet, thank you for joining us; thank you for your fantastic questions. Let’s say, “Goodbye.”

Tara Land: Goodbye.

Leave a comment