It’s been a while since I last wrote a blog post but if you’ve been reading our weekly notes, we’ve been quite busy lately!
It is time though to tell you a bit about one prototype we worked on in Sprints #3 and #4. As said before, one key finding from discovery was that people care about issues that matter to them, their family and friends. At the moment, the Parliament website does not cater for easily finding information on a subject, a topic or an issue (we are still unsure about the terminology to use here). Besides, our analytics for the past years have constantly shown that the most visited pages on the website are Bills and legislation, Members, and Visiting information. Given that one key task when visiting the website is to find information about Members we decided to tackle two key aspects of Parliament and user needs: Members of Parliament and issues (we will stick with the term issues for the time being…).
We started with a broad idea of what we could do with Members – in conjunction with what we are currently working on with the domain model of our parliamentary data. But being too broad (i.e. all the information available about a member) was not and still is not tailored for the type of experimenting we are doing in Alpha: start small, iterate and scale up. So we decided to start small and define our goal around the starting point of a user journey: I have an issue/I am interested in an issue that I want to know more about. Our big assumption at that stage – based also on the top tasks completed on the existing site - was to say that one thing one might find useful is to find members interested in the given issue or find the members who are working on this issue (illustrated in Parliament by their parliamentary work, their contributions).
We started in Sprint #3 with sketching a low fidelity paper prototype that would allow users to find members who have listed an issue as an interest and/or have contributed on this issue – by contributing we mean having submitted oral questions, written questions, made contributions in debates, tabled Early Day Motions, etc. Let’s call this prototype the “People’s finder” prototype.
We conducted some guerrilla testing with staff of the House of Commons and House of Lords to get a quick idea of whether we were on the right track. We learned a lot about how people might want to filter their results or how the importance of difference contributions could be ranked. You can read about Ed’s experience of this in a previous blog post.
Jack then went off and built an HTML prototype based on the findings of the guerrilla testing.
In Sprint #4, we quickly agreed that we needed to further explore and test this idea so we started by conducting usability testing on the HTML prototype. We agreed that the personas we would focus on were Lakshan, the specialist user working as a lobbyist, and Eric, the interested member of the public.
With this in mind, we recruited participants for usability testing and scheduled sessions with two public affairs specialists, one member of the public and MP research assistants.
In parallel, we wanted to explore another way of finding members interested in an issue through their membership to groups. By groups, we mean committees, all party parliamentary groups (APPGs: informal cross-party groups of members of the Commons and Lords interested in a particular topic), government and party roles. Let’s call this prototype the “Groups’ finder” prototype.
Ganesh, Jack and Ed conducted further guerrilla testing with members of staff in both Houses with the Groups’ finder paper prototype while I conducted most of the usability testing for the People’s finder HTML prototype.
The findings of the usability testing on the People’s finder taught us a few things:
- There are existing user journeys and behaviours that we need to make sure we maintain. For example, the name of a member being hyperlinked means for all participants that you will land on the profile/biography page of the member with their contact details and biography information.
- Allow filtering of issues on results page to sub-categories within issue (i.e. right to buy is seen as a topic within the housing topic)
- Content needs to be worked on, especially for the 'no results found' page as not particularly useful to participants
- Further work needs to be done to explore the issue of ‘my MP as a representative who can help me vs. MPs in general interested in an issue’. This point was raised by the member of the public who did not see why she would get in contact with other members of Parliament as her assumption was that only her MP could help her.
- Activities/contributions vs. group/membership
- Members’ membership to committees, parliamentary and government roles were suggested by participants as a useful indication of a member’s involvement on a specific issue. This point validated our assumptions that groups or memberships would be a useful way to get to members interested or involved in an issue.
As for the guerrilla testing of Groups’ finder:
- Participants considered that members' membership to a group was a sign of their definite interest in that specific topic
- Ordering of groups was good. Departmental roles may be more important than APPGs.
- Party spokespeople are seen as valuable
- Contacting: title link goes to committee landing page and name goes to biography.
After that, we took the lessons learned of our testing and iterated the HTML prototype:
We were left with a number of questions and explorations at the end of Sprint #4 that we want to take further in the next sprints:
- Do more work around 'groups'
- What happens if there isn't any content around an issue?
- How do we define an issue/topic/subject?
- Work further on a more rounded-up journey from looking for a member on a specific issue to potentially taking an action (i.e. contacting a member or a parliamentary group) and when this action should take place?
- We will take a focus on the general public, namely around our personas Eric and Patricia, as we need to understand better the limitation around the assumption that ‘only my MP can help me with my issue’ as opposed to contacting other members of the Commons or the Lords. We need to understand and manage expectations of contacting members of Parliament.
- Adding real data and test it
- It will also be the opportunity to expand on the domain model, but I will leave it to Giuseppe and Rebecca to tell you more about this!