Wrapping up the project
Sharing everything necessary to take this work forward
Project details
- Length: 11 weeks (January — March 2021) with a two-week extension period in late April / early May.
- Funder: Catalyst
- Charity partners: Hyde Foundation, National Ugly Mugs, Turn2us (with help from Citizens Advice and Work Rights Centre)
- Digital team: Doug Belshaw (Project Manager), Dan Mosforth (Service design), Hannah Belshaw (UX & user research), Ivan Minutillo (UX & front-end dev), Tom Broughton (technical feasibility)
- Contact details: hello@dynamicskillset.com / daniel.mosforth@bay.digital
Problem
The challenge given to us by Catalyst was:
What if we prototyped, tested and identified a way to provide seamless remote support for online Universal Credit claims, so the 25% of people currently failing to register can make a timely and successful claim?
This work should rapidly generate, prototype and test ways of providing remote support for making online Universal Credit claims. The outputs should enable organisations who provide benefits advice to best support those who struggle or fail to claim Universal Credit online.
By the time we began the project, indications showed that the percentage referenced could be as high as 33%. We needed to work quickly and produce results.
Full details of the project kick-off are available in this post.
Definition of done
We agreed the following with Catalyst:
- Prototypes that have been tested with service users and/or the
organisations supporting them. A prototype here might be a product,
service, experience, document or a comparison of existing tools. - Insights from research and testing
- Documentation of the process and learnings to be openly published
and shared with the sector.
Approach
We used a double-diamond approach to dig deep into the problem area during the first five weeks of the project. See this post for more details on the convergent and divergent thinking that we used during the process..
Our findings from user research (see this post) were that there were six main problems to solve:
- Lack of awareness of what’s necessary
- Lacking of visibility of claimant activity
- Don’t know what certain questions mean + consequence on app
- Lack of English language skills
- Losing passwords / login details
- Not sure how much they will get
We decided to focus on the first two of these (see this post), with our rationale being:
- Language skills — although language skills have been highlighted as an issue preventing people from registering for Universal Credit , our research has flagged that helping those with poor English language to navigate the online form could lead to unintended consequences. For example, the Department for Work and Pensions (DWP) will be unaware they require further assistance at the next stage (managing a claim).
- Digital skills — we are using whether a user currently has access to their own email account as a diagnostic tool to decide whether they are able to register for Universal Credit independently.
- In both of the above cases, we would direct users to the phone helpline, or to support from charities such as those involved in this project.
After much discussion and work on personas (see this post) we settled on three potential prototypes:
Prototype A — Visualisation of steps: service map showing overview of application process. This would help claimants understand what is involved in the Universal Credit process.
Prototype B — Checklist: interactive check-box list of documents and other resources required to fill in Universal Credit form. This would help claimants prepare the information required for the Universal Credit application.
Prototype C — Real time support from a real-life professional: document comparing options for screensharing between claimant and adviser. This would help advisers who cannot see claimant activity.
A fourth, ‘in-context help’, was mooted but never fully explored. This would be something like an AI assistant making suggestions when people are stuck with the application form.
Prototypes (main project period)
The three initial prototypes the project team came up with are detailed in this post. We presented these at an online show-and-tell session (see this post) attended by representatives of the DWP, other government departments, and the charity sector.
We continued to perform user research and testing as we iterated the prototypes, refining our personas (see this post) as we did so.
Prototype A — Visualisation of Steps
Prototype B — Checklist
This prototype was iterated upon and superseded by Prototype E (see below).
Prototype C — Real-time support from a real life professional
Prototypes (additional funding period)
As the main funding period drew to a close we published this post sharing experiences from the representatives of charities involved in the project.
The digital team then received 10% uplift from Catalyst, which was used to fund two more weeks of prototype iteration. We decided to work directly with the Hyde Foundation based on charities’ interest and engagement over the initial 11-week funding period.
Prototype D — adding checklists to the overview
We tested two variants of Prototype D, one using a ‘drop-down’ approach, and the other continuing to use the ‘overlay’ approach from Prototype A.
The overlay approach was much preferred by both Hyde Housing and NUM, as well as the end users we tested it with. One reason for this was that it seemed to work better with older devices.
Based on the feedback we received, we colour-coded each circle once complete, as well as adding an option for users to call a helpline if they require some assistance.
Overlay approach
Drop-down approach
Prototype E — adding an overview to the checklist
We have worked openly and transparently from the inception of this project. Any information or files not listed above, should be available via the links below. If you do not see what you require, please get in touch.
Handover docs
Presentations
Image assets created
- Visual thinkery from Bryan Mathers (CC BY-ND license)
- Initial personas
- Revisited personas
Approaches used
- Big reveal
- Defining the cast
- Fast 5's
- Must / should / could
- NOISE model
- Pre-mortem
- Traffic light categorisation
