Workrate

Closing the loop between work delivered and work invoiced.

Every month, a service company sends people out to do the work. Shifts get planned, hours get logged, jobs get done. But a job isn’t really finished when the work ends. Someone still has to make sure it was billed, and billed correctly.

Every job goes through three teams before it’s paid. Sales agrees the work with the client, Invoicing bills it, and Finance does the final check: Does what we worked match what we billed?

That’s harder than it sounds. The hours worked and the sales orders sit in different places, so a mismatch is easy to miss. Until finance closes that gap, the month isn’t really closed either.

My task was to design the Finance check page, where finance reviews, approves, and flags every job each month.

year

2024

timeframe

8 months

tools

Figma · ChatGPT

Problem

At the end of each month, finance had to answer a simple question: are we done? The system made that hard to answer.

Jobs and invoices weren’t shown together in one overview. To find out whether a month was closed, finance had to check jobs one at a time and piece the status together. Open work and finished work looked the same, so nobody could say at a glance how much of the month was left.

On the VPC side, the biggest gap was the link between a job and its sales order. A job can’t be invoiced without one. But a missing or wrong link only showed up late, often when an invoice failed or the numbers didn’t match in SAP. By then, fixing it meant going back through several screens and asking other teams.

Talking with the finance team, we learned one more thing that changed the direction of the project approving jobs early, even before they’re invoiced, makes monthly reporting faster and more reliable. The page didn’t just need to show information. It needed to help finance work through the month and actually close it.

Solution

The goal wasn’t just a cleaner screen. Finance needed to close the month faster, catch billing errors before they reach the client, and make sure no worked hours go unbilled. Every design decision supported one of those three goals.

The page is organized around the reporting month, because that’s how the business measures itself. Jobs are grouped into Open, Approved, and Completed invoices, each with a live count, so finance can see at a glance how much work is left before the month is done.

Because jobs can be approved before they’re invoiced, finance doesn’t have to wait for billing to catch up. Reporting moves forward while invoicing follows, which means faster and more reliable month-end numbers for management.

Finance check overview

The month at a glance. Open, Approved, and Completed jobs each have their own tab with a live count, so finance always knows how much work is left. Colored sales order chips and Action required statuses show where to look first, and expanding a row reveals the PO number, comments, and any open alert without leaving the page.

Open jobs

Linked vs. not linked

Everything about one job on a single page: sales order, period, approval, SAP invoices, and a supplement table comparing sales order hours with invoiced hours for each rate. If the job isn’t linked to a sales order, the Invoice button stays disabled. No link means no invoice, and that rule holds on every screen.

Not linked job

Linked job

Outcome

Finance Check turned the last step of invoicing into something the finance team can actually manage. Instead of piecing the month together from different screens and emails, finance now has one place that shows where every job stands, what’s been approved, and what’s still blocking an invoice.

For the business, the design supports what matters most at end of the month, work gets billed, mistakes get caught before they reach clients, and the month can be closed with confidence. Because jobs can be approved before invoicing, reporting no longer waits for billing to catch up.

For me, this project was about designing for a process, not just a screen. Working as both designer and product owner, I learned the finance team’s language, turned their rules into clear interface decisions, and took open questions back to the team until the logic held up.