Overview

This case study is about my UX work on a Human Capital system, focused mainly on payroll-related features: compensation factors, formula configuration, manual value entry, payroll calculation, and issuing accounting documents.

The product is built for expert, enterprise-level users. It’s powerful and highly configurable—which is great—but that flexibility also makes it fragile. When people are working with large volumes of sensitive financial data, even a tiny mistake can turn into a serious problem.

So instead of zooming in on one screen or a single feature, this case looks at how I designed UX guardrails across multiple parts of the system. The goal was simple: reduce financial risk without limiting expert users or turning the product into a rule-heavy monster.

The Problem

Users could:

• Configure complex payroll formulas
• Adjust coefficients and calculation parameters
• Manually enter large volumes of payroll data
• Run payroll calculations
• Issue accounting documents

All of these actions were tightly connected. I remember thinking how scary it was that a single wrong number—maybe entered late in the day, maybe copied one row too far—could ripple through the system and affect payroll results or accounting outputs.

At the time:

• Most feedback appeared after actions were completed (for example, after issuing documents)
• Payroll calculations already had anomaly detection, but from a UX perspective, the feedback wasn’t easy to act on
• High-volume manual data entry made fatigue-related and repetitive errors very likely

The core challenge was:
How do you support expert users in a highly configurable, data-heavy financial system—while actively reducing costly mistakes—without adding friction or unnecessary development complexity?

My Role

• UX Designer working closely with the Product Owner and the development team
• Responsible for interaction design and UX decisions
• Focused on reducing risk through design—not defining payroll rules or business logic

All business rules and calculation logic were owned by the PO. My job was to translate those rules into clear, safe, and efficient user experiences. That translation layer was where most of the real UX challenges lived.

Constraints

• Enterprise product with expert users
• High-impact financial data
• Large-scale manual data entry
• Limited development capacity

Every solution had to be realistic, lightweight, and scalable. Big ideas were great, but if they couldn’t survive real-world constraints, they weren’t useful.

Key UX Decisions

1. Guardrails for Safe Configuration

On pages related to payroll formulas and compensation factors, the biggest risk wasn’t misuse—it was accidental or unclear changes.

To reduce that risk, the team and I focused on:

• Clear separation between formula structure and editable values
• Controlled input patterns to prevent accidental edits
• Meaningful naming and grouping to lower cognitive load

The intention was never to block users. These safety nets were about helping people stay oriented and confident about what exactly they were changing, especially in complex configurations.

2. Guardrails for High-Volume Data Entry

On the input values page, users enter monthly payroll data for many employees. It’s repetitive work, and honestly, that’s where mistakes love to hide.

The main risks here were data loss, slow workflows, and subtle copy-paste errors.

Split View for Focus and Data Safety

• Users work with one employee at a time
• Data is automatically saved after each employee is completed
• A clear “Save & Next” action creates a natural checkpoint

This approach helped users focus and made sure work wasn’t lost halfway through a long session. One thing that really mattered here was giving users peace of mind while moving fast.

Vertical Data Entry Pattern

• Users can move vertically between employees for the same payroll factor
• Navigation is supported via keyboard shortcuts
• The interaction mirrors familiar spreadsheet behavior

The cool part was how quickly users adapted to this. It matched their mental model just like Excel, improved speed, and noticeably reduced repetitive mistakes.

Contextual Value Visibility

• For fixed factors, previous values are shown as reference
• This makes unexpected changes easier to spot
• Supports consistency and accuracy over time

Early Anomaly Signals (Exploration)

As a potential next step, we explored adding lightweight anomaly signals during data entry.

The idea wasn’t to block users or enforce validation rules, but to gently say: “Hey, this value looks unusual.” These smart safeguards were meant to raise awareness early, so fewer issues would reach the payroll calculation stage.

3. Designing Around Anomaly Detection

Payroll calculation already included anomaly detection. From a UX point of view, detection wasn’t the real problem—actionability was.

So the focus shifted to helping users understand what went wrong and fix it quickly.

The design aimed to:

• Take users directly to the source of the issue
• Minimize the effort needed to investigate anomalies
• Support both correction and intentional overrides

Key interaction decisions included:

• Automatically navigating users to the exact page where the anomaly originated
• Highlighting the specific row or input field involved
• Allowing users to either correct the value or explicitly ignore the anomaly when the outcome was intentional

This turned anomaly detection from a passive warning into an active, usable part of the workflow.

It’s also worth calling out that calculation-level anomaly detection works independently from the early anomaly signals explored during data entry. They operate at different moments in the journey and are designed to support each other—not conflict.

Why This Matters

Payroll systems are high-risk by nature.

In environments like this, UX isn’t just about usability or speed. It directly affects trust, accuracy, and financial safety.

This case shows how thoughtful protective patterns can:

• Increase confidence in complex financial workflows
• Reduce errors without restricting expert users
• Protect critical data through interaction design

Outcome

• Safer configuration and data-entry experiences
• Lower risk of incorrect or lost data
• Faster, more focused workflows for expert users
• A solid foundation for future UX guardrails

Key Takeaway

In high-risk enterprise systems, guardrails don’t always look like warnings or hard restrictions.Most of the time, they’re quiet, well-considered interaction decisions that guide users toward safer outcomes—without taking away the flexibility and control they depend on.