Budgeting System

Three parts of the shipped system carry the heart of the design.

China Post Group Corporation is one of China's largest state-owned enterprises. It wanted to build its own budget management system in house, and EY's tech consulting team was brought in to design it.

The starting point matters. China Post had never used a shared digital system for budgeting at all. Each unit filled in its budget tables by hand and emailed them up — department to subsidiary to corporate. This project wasn't redesigning a product. There was no product. It was a digital transformation: moving 900,000 people off paper and email, onto one system, for the first time.

China Post's budget management was slow, manual, and impossible to see into. Three problems defined the cost budget planning and review experience — and each one had to be solved a single time, in one interface, for every kind of unit at once.

• 01No real-time budget tracking.

  Users had no live view of expenses or budget performance. Timely, informed decisions were nearly impossible.

• 02Slow, error-prone data processing.

  Budget tables were buried under hundreds of irrelevant account items, and historical reference data was hard to reach.

• 03No tools for budget review.

  The review process had no visibility. Once a plan was submitted, no one could see where it was or who held it.

Four principles guided every decision — and each one was really an answer to the same question: how does one system fit an entire, varied enterprise?

• 

  Seamless Adoption

  Design intuitive, easy-to-navigate interfaces that require minimal training, ensuring quick and smooth adoption across departments and subsidiaries.

• 

  Customization

  Allow users to customize views and filters based on their department or subsidiary, making the system adaptable to various roles and needs.

• 

  Transparency

  Ensure clear visibility at each stage, allowing users to easily track progress and understand who is responsible for each step.

• 

  Data Security

  Implement robust measures to protect sensitive financial data, ensuring compliance with industry security standards and maintaining user trust.

I was one of two junior designers on a four-person EY team, and I owned roughly 60% of the design work — co-leading design with the other junior designer. The project also ran across China Post's own development, data, and finance departments, so coordination between organizations was constant.

• User Research

  Facilitated 7 stakeholder interviews and documented 40+ interviews, identifying key pain points and aligning design solutions with user needs.

• UX & UI Design

  Designed the user flow for the Cost Budget Planning section — 20 interfaces prioritizing usability and smooth navigation.

• Key Artifacts

  Independently developed the cost budget planning process map and organizational chart that informed design decisions.

• Cross-functional Coordination

  Wrote 5 product requirement documents and worked closely with the development teams to align technical execution with design intent.

• User Acceptance Testing

  Co-led UAT with 100+ users, turning feedback into design iterations.

• Stakeholder Presentation

  Co-presented updates at 7 key milestones, integrating feedback to keep the project on track.

Understanding a system with many faces

A system for 900,000 people across many kinds of unit could not be designed from one vantage point. Research came in three parts.

Interviews with stakeholders. I interviewed department heads across corporate, subsidiary, and provincial levels to understand how budget planning actually worked at each tier — and mapped China Post's full organizational structure to see exactly which unit needed what. That map led to the phased rollout: starting with Express & Logistics, the subsidiary with the most complex account structure.

Document analysis. I analyzed China Post's existing manual budget documents — the hand-filled tables themselves — to understand exactly what data fields and metadata an automated system would have to carry.

Interviews with end users. I spoke with the people who fill budgets in day to day, turning what they said into a clear set of insights about where the manual process broke down.

• Insight 01

  Current quarterly budget reports created a significant time lag between financial activities and management awareness, hindering the ability to respond promptly to budgetary issues.

  • "We need to see what's happening now, not three months ago." — P3
  • "Instead of canceling projects in Q4, we could be smarter about spending throughout the year." — P12

• Insight 02

  Budget planning was slowed by cluttered tables filled with irrelevant account items, and difficulties accessing historical reference data.

  • "It's a nightmare... 6 tables... each with hundreds of items. Half the time, I'm scrolling through stuff that doesn't even apply to us." — P16
  • "I spend days digging through old files trying to find the right reference..." — P7

• Insight 03

  Budget managers struggled to track approval status and manage document versions, leading to confusion and inefficiency in the review process.

  • "Once it's submitted, I have no idea where it is or who's looking at it. Sometimes I wonder if anyone's even seen it." — P5
  • "The file names... 0820_v1... 0820_final... but to be honest, I don't know which is the final version." — P8

Mapping the process before the screens

Before any interface, I mapped two things. First, a full process map of the current budget submission, review, and approval workflow across every tier — clarifying the whole chain and exposing where it could improve. Then the information architecture for the system: a clear, scalable structure organized around user tasks, built to adapt across departments and subsidiaries.

The data table: one table, every user's own slice

The hardest screen was the budget data table. Financial budget data is enormous — hundreds of account items — and most of any given table does not apply to the person looking at it. Different departments report on entirely different financial items.

My first design laid the data out and let users work the whole sheet. I took it into user testing with budget managers — and the testing was blunt about three things:

1. 01Users wanted to review one line of data at a time, not the whole sheet at once.
2. 02Users wanted a clearer, more direct view of when a specific data line was last changed.
3. 03Users found selecting benchmarking options through a dropdown took too much effort.

The redesign answered all three findings directly. The default view is account-based and collapsible. Every line shows its last-updated date, so everyone editing stays aligned to the same version, with a review-history control to see how a line changed over time. And the benchmarking dropdown became a tab control — the comparison set is now one click, not a buried menu.

Designing for the working area

Research showed users spent almost all their time in one place — the budget data sheet itself. But the navigation bar took a wide, fixed strip of the screen, squeezing the area that mattered most.

Speaking the engineering team's language

I handed off a set of product requirement documents — and the engineers came back with questions the PRDs had already answered. That was the signal something was wrong.

The PRDs were complete. But I had written them in business terminology — and the development team didn't share that vocabulary. Worse, some data terms genuinely mean different things in business language than in their system. The document was right and still unreadable to the people who had to build from it.

The Iteration

Learn their language. I studied the development team's system directly and re-annotated every PRD in their terms — not my own.

Communication sharpened immediately, and unnecessary rework dropped. A spec only works in the vocabulary of the person reading it.

The dashboard: tested, then reshaped

The budget dashboard was the live view China Post had never had — so it had to be right. I designed a first version, then put it through UAT with 100+ budget managers from across the provinces.

That testing reshaped it. The dashboard gained direct navigation from each summary card to its detail page; a toggle to read departmental expenses as either an amount or a percentage; a flexible timeframe control on the expense-trend chart; and — because corporate-level stakeholders needed to see any region — a province selector and an interactive map of expense by province.

A SaaS budget management platform that turned China Post's manual, paper-and-email process into one scalable digital system — the complete Cost Budget Planning experience, the dashboard, and the documentation handed off to the client's development team.

• 01 — The budget dashboard

  A real-time, customizable view of budget performance — expense trends, departmental breakdowns, and an interactive province map — that adapts to whether a provincial manager or a corporate stakeholder is looking.

  

• 02 — The account-based data table

  The core working surface. Each user sees only the budget items tied to their account, in a collapsible view, with version history on every line and one-click benchmarking — reshaped directly by user testing.

  

• 56
  Interviews
• 8
  Product Requirement Documents
• 20
  Interfaces
• 4
  Key Artifacts

The system was implemented across China Post's nationwide network — one shared platform replacing a manual, email-driven workflow for 900,000 employees. It removed the need for external software, cut operational costs, and took a month off the budget review cycle. For an organization that had never run its budgeting on a digital system, the real result was quieter and larger: a process that used to live in scattered files and inboxes now had one place to live.

How success should be measured

A system this large isn't proven on launch day — it's proven over the months that follow. Part of the handoff was defining how China Post should measure whether the system was actually working, in two groups.

System-related Metrics

• Time Saved

  Compare the time taken for budget preparation before and after system implementation.

• Reporting Efficiency

  Measure the time taken to generate standard budget reports against the current baseline.

• Error Reduction

  Monitor the number of budget revisions or corrections required pre- and post-implementation.

• User Satisfaction

  Conduct user surveys after implementation and track the Net Promoter Score (NPS) over time.

Business-related Metrics

• Performance Improvement

  Monitor key financial ratios such as operational profitability against their level before implementation.

• Financial Decision-Making Efficiency

  Measure the time from budget request submission to final approval, compared with historical data.

• Cost Control Effectiveness

  Track the percentage of departments staying within their allocated budgets against pre-implementation figures.

• Return on Investment (ROI)

  Calculate the financial benefits of the new system — cost savings and efficiency gains — against its implementation cost.

• Designing for scale means designing for variation.

  A system for 900,000 employees across 31 provinces and many subsidiaries isn't one design — it's a flexible structure that bends to each unit. The phased rollout and the customizable views weren't nice-to-haves; they were the only way a system this large could realistically succeed. Mapping the org chart and the process first — before any interface — is what made that possible.

• Engineering constraints are design problems, not design limits.

  The data-table workaround and the PRD-terminology fix both began as friction with the development team. Treating that friction as a shared design problem — rather than a blocker handed over a wall — is what kept the project on its strict five-month timeline without cutting real functionality.

• A digital transformation is carried by structure, not screens.

  These users had never worked in a digital system before — they came from hand-filled tables and email. At that point, a clean interface isn't enough. What carries 900,000 people across a change that large is a structure clear enough that every unit can find its own place in it.