CASE STUDY 05 / PUBLIC SECTOR WORKFLOWS
Problem Register
for state institutions
Sharing digital service problems so others can learn from them
Read the case study
CENTRAL DESIGN QUESTION
When a plan stalls in review, how can everyone see why, and who can move it on?
The Problem Register is a shared space for Hungary’s state institutions. When a problem comes up in one of their digital services, for example a step where citizens keep getting stuck, the institution records it here and every other institution can see it. Others can learn from it, and problems that appear across many services can be handled the same way everywhere instead of being solved again each time. Institutions then send the developments they plan in response through a companion tool, Submissions, where they are reviewed before work begins. I worked on the research behind both, focusing on what happens when a plan stalls in review.
MY ROLE
UX designer contributing to the interview research and the interpretation of organisational workflows.
SCOPE
Responsibilities, information hand-offs, review dependencies and the design implications for problem lists, plan submission and review.
DELIVERY
Research and concept direction. The interface ideas discussed here are proposals arising from that work.
What was moving through the system
Four kinds of item moved through the two tools: a problem someone had found, a development proposed to solve it, an annual plan that bundled several developments, and a procurement request to buy the work. They were related, but each answered a different question and represented a different level of commitment.
The Problem Register (Kataszter in Hungarian) held the problems, grouped by life event, such as having a baby or losing a job, and by severity. That is the problem map on the opening screens. Submissions (Bejelentések) held the developments and plans that institutions sent in for review. Keeping the two apart mattered: if a problem and a proposed solution are treated as the same thing, nobody can ask whether the proposed work actually addresses the underlying need.
I approached the research through the people involved and the information each of them needed to act: the institution requesting a development, and the teams responsible for coordination, technical assessment and procurement. They often saw the same plan differently, because each was responsible for a different decision.
FIG 1 · MORE INFORMATION REQUESTED
Looking beyond a simple sequence of statuses
Reviewing a whole annual plan and reviewing a single procurement request were different processes, and parts of them ran in parallel. Showing them as one straight line of statuses would have hidden what was really happening.
A typical case: a hospital submits a plan for an online appointment system. The plan sits with the coordinating team, so its status reads “pending”. The real blocker, however, is a missing data protection assessment that only the hospital can supply. The label tells everyone there is a delay but not why, and knowing which team holds the plan does not tell anyone who can unblock it.
My design interpretation was that useful status information needed to explain the dependency. Which review was in progress? What input was still required? Who was responsible for supplying it? What could the person looking at the record do next?
Those questions changed the focus of the interface discussion. A status needed to show what the item was waiting for, not only where it currently was.
FIG 2 · CONCEPT
Each milestone shows what it is waiting for, not only where it sits in the process.
Following information across organisational boundaries
The research followed how information moved between institutions: hand-overs, checks, requests for more information and approvals. It also looked at the same data being requested again and again, information scattered across channels, and how hard it was to trace earlier decisions.
These issues are connected. When it is unclear how two records relate, people retype information or reconcile versions by hand. When a request for more information arrives without context, the institution knows something is missing but not what the reviewer actually needs.
The design implication was to keep the context with the item as it changes hands. A returned plan should show, in one place, what is being asked for, why, and who needs to respond. FIG 1 shows this direction: the requirement that needs correcting is marked, and asking for help is one click away.
Translating the research into interface questions
The design brief that followed covered three areas: a list of problems and development ideas, building and submitting a plan, and reviewing it. I used the workflow questions to define what each area had to communicate.
For the list, the useful distinction was between the underlying problem and a proposed development. For plan preparation, it was the relationship between the plan and the items included in it. For review, it was the relationship between the current status, an outstanding dependency and the action needed to proceed.
Problem record
Keep the need distinguishable from a proposed solution
What issue is this development intended to address?
Plan preparation
Make the included developments and their relationships clear
What is being submitted together?
Review
Show the relevant dependency and responsibility
What is preventing progress, and who can act?
Request for more information
Retain the reason and the requested response
What needs to be supplied or corrected?
Decision history
Preserve the context of changes and decisions
Why is the item in its current state?
These directions fed into the next phase of design: a problem and idea list, a plan-building interface and a review workspace.
05 · WHAT THE WORK ACHIEVED
When a plan stalls in review, how can everyone see why, and who can move it on?
06 · TAKEAWAY
This work developed my ability to investigate a service across organisational boundaries. A screen can accurately display a status and still leave the user unable to act. Understanding the responsibility and dependency behind that status is what makes the information useful.
NEXT CASE STUDY · 06 / 06
BettAir
Turning office air-quality data into one score anyone can read