At HMRC Digital we look to put user needs first, designing and creating digital services that are shaped around what works for users of the services.
Three weeks into my time at HMRC Digital I wrote about working in a team in government, which you can [read here](/hmrc-nearly-three-weeks-up/. I dealt in broad strokes previously.
Now approaching three months on the job, I thought it might be useful to explore how we are approaching creating the inheritance tax digital online service with a user first approach.
I joined the project a couple of months back and the team had already done some sterling work creating theoretical design approaches, going through these with users in research sessions, and adapting our work based on those responses. As design approaches are approved by users our developers create the digital service.
Inheritance tax itself is hard. I compare it to papier mâché.
The original rules were created. Think of them as the first layer, neat, and on recorded on paper. And as time has passed and little bits have been stuck on top, more and more little bits responding to situations, scenarios, where the rules have had to be created or changed. Inheritance tax is not alone on this. A lot of British policy and law is like this. It adapts to circumstances. This means though that this is hard.
Creating the ability for citizens to submit inheritance tax information online is a long process. To help us bring the digital inheritance tax online we are carrying out two phases of work: one this year (2015), the second next year (2016). It is a big programme.
(And remember laws and policy are always capable of change.)
This year’s work is focused on bringing online what is currently centred around the 205 inheritance tax paper form.
What this means as simple as possible: Anyone who fills in a 205 form and while doing that isn’t at any point directed away to any other inheritance tax form doesn’t have any inheritance tax to pay. A lot of the work being doing in this first phase is groundwork for any later phases.
Research is important to the way we work.
We want to know. As we work in government we have a lot of great insight to draw upon, and we work closely with the inheritance tax teams within government who have a lot of experience and knowledge. We also talk with our colleagues in other government agencies, including DWP and GDS.
If we don’t know, we care, we want to find out, digging deeper. There are plenty of ways we can get this understanding.
If we don’t get answers we will be working with assumptions, which is not a preferred way of designing and making.
If we have to work with an assumption to keep our work going, we look to validate (prove/disprove) that assumption as soon as we can, in the most appropriate way. Slight correction: We make sure we validate that assumption.
We do – after all – work with user needs.
From research we know that the majority of people who fill in the current paper based 205 form will be doing this as a relative of someone has passed away.
Collecting and organising this information is not the easiest process. This is amplified if you are looking to submit inheritance tax information for a close family member.
It is our goal to make submitting the inheritance tax information as painless as we can. We know it is not easy for the user for emotional reasons.
We also know collecting the information that is needed is not easy. It can take time to find. It can take time to look through. And it can take time to understand what exactly is needed for inheritance tax.
We don’t want to make it any harder. If anything, we want to make it easier.
Inheritance tax is policy driven: government has its needs. This, for the most part, means we need to acquire information that is needed to form a conclusion against that policy. At the lowest level a citizen will provide information to find the answer to the following question:
Do I need to pay any inheritance tax on this person’s “estate”?
While there are government needs (it is a tax), the citizen’s – the user’s – needs are at the centre of the work we do. We want to make sure the digital service we create is accessible, is usable by the general public. And we need to make sure the information we collect is stored securely and is the right information for HMRC.
That last point - that we provide the right information for HMRC - has come through as a strong user need during our research. Users want to make sure they provide the right information to HMRC the first time. They want to make sure they get it right, for themselves, for the deceased, for anyone else.
At its simplest level inheritance tax needs the user to provide information about the deceased’s assets, debts, and any gifts the deceased has given away.
As we have it in the service at the moment:
An asset is anything the deceased owned that has a value.
A debt is anything owed by the deceased.
A gift is anything given away by the estate.
There are different types of asset, debt, and gift.
We have designed the service so the user can provide information as simply as possible. In many cases we employ a simple ask a question and get a “Yes” and “No” response. And we provide supplemental and guiding information wherever we can, to aid the user.
The information provided by the user allows the service (and therefore HMRC) to reach a conclusion – which you might remember is “Do I need to pay any inheritance tax on this?”
The assets and debts parts of the online service are in a position where we feel we can push the service into ‘beta’. Beta means we would make the service available online so the public can use the service – and by using it helping us understand user needs even more, so we can adapt the service to meet those needs. (Because we’re driven by research.)
Gifts are not so easy to explain and understand. Recently we have been spending a lot of time in user sessions looking to find out how we can explain to users gifts in a way that isn’t hard. We have been spending time focused on this because previously the user response hasn’t been too positive.
To be fair, gifts is hard.
For the user research sessions we tend to provide the user with documents that mimic real life documents – such as bank statements, emails from relatives saying they received some money, a share certificate, and so on. We then ask the users to review those documents and show and tell us how they would extract information for the inheritance tax online service.
The week before last we reviewed the previous user sessions and found that when presented with an example of an inheritance tax our participants were taking the figures provided and working backwards to see how those figure were arrived at.
It was fascinating. Our users said they wanted to be sure they were working their figures out correctly and the examples illustrated a correct way. Our users also found that understanding gifts was hard.
We also had user feedback about the tone of the language, with suggestions there was a lot, it was “the usual government tone”, and it could do with being “more human”.
Because user research sessions are about exploring we decided we would use the Friday to pull together a prototype that provides several pages of simple, clear explanations of “gifts”, written in a slightly more informal tone. These pages would come before the user enters any gift information.
A prototype is something that tests an idea. (I have blogged about this before.) The prototype we looked to make is a “fake” version of the service that users could use in a web browser, as if it was the real thing.
Although we moved quickly to make this prototype, as a team – myself, Andy the content designer, and Sara our user researcher – we huddled and worked nimbly, first sketching out some wireframes, understanding what content we need, how we will divide it, and then setting to work on making this around my Macbook.
I code using the GDS prototype kit using our sketches as reference, while Andy pulls together the copy we need as content. Sara, as always, keeps us on track making sure we meet user needs, as well as helping shape some of the detail of the upcoming research sessions.
On Saturday I took a couple of hours to review the work (always good to sleep on something new), and did a couple of tweaks.
Sunday morning, thanks to people’s generosity, I did some pop-up research near where I live to see how the tone and size of the pages felt. If there were any issues from this quick Sunday session we could resolve anything major on Monday as part of our team review ahead of the day of user research on Tuesday.
Doing some pop-up research with the prototype
Sunday was a quick five minutes each with four people, and half an hour in-depth with one person. Tuesday we went into the user research sessions looking for more focused, more detailed user response – each session was 90 minutes, one on one in a room at a computer, listening and watching the users interact with the fake service we’d pulled together. It’s always fascinating seeing how our manifested theories act as stimulus, whether the response backs up our approach or not.
The outcome to Tuesday’s sessions was really favourable.
Examples pages from the prototype created
Users found the “mini manual” pages effective, helped them understand, albeit not completely. Several users told us if they were doing this application for real they would actually print off one of the pages, take it away to read in their own time, so they could understand – and know they understood.
This was a really unexpected response but one that made sense. We had divided a lot of bulky information into smaller pages, and altered the tone. Once it got to the page where the user really needed to understand some new concepts this made sense. And printing it off made sense as well, especially given the large amount of “desk work” required by any inheritance tax application.
The person filling in the inheritance tax application online will receive information in emails electronically, on letters through the post, from phone calls and their own written notes. There is no escaping that at this moment in time. It made us super-aware we need to make sure the pages print alright!
What didn’t work so well? While we made in-roads on gifts, users need further help to understand exemptions that can be applied to gifts. Users struggled here – and we know that this is an issue with the wider service. We could use this as our cue for What To Do Next.
Overall, we could view our research sessions as a “success”. They showed we were moving in the right direction, and also helped us guide where to concentrate next.
Earlier this week we had a workshop exploring how we can make the process for the user simpler. We immersed ourselves in the full service, and put ourselves in the eventual users’ shoes.
Where would they get this information, where would they find out what to get, how would they build it up, how would they split it up, how would they “record” what the figures were, and so on.
We have some ideas which we need to put in front of users, see how they respond. Because, at the end of the day, it’s not about what I think is right, or anyone else in the team thinks is right – it what is right for our users, it is what works for our users.
We’ll be continuing to research this in the coming couple of weeks. Hopefully I will be able to update you in the near future on how we get on.
If you want to read more on how we design in government there’s a great post by Rebecca Cottrell on the GDS design notes blog.