# Warehouse workforce management: shift planning case study

> Team experience at a large marketplace’s warehouse: shift planning, actual timesheets and agency hour reconciliation in one system, at 800 person-shifts a day.

Page: https://opsalute.ai/en/cases/hrms
Русская версия: https://opsalute.ai/llms/ru/cases/hrms.md

Case study · Team experience · Yandex Market

## Warehouse workforce management at 800 shifts a day. Every hour has to add up.

A fulfillment center runs on permanent staff, agency workers and self-employed contractors. The team linked shift planning, agency requests, check-ins, timesheets and agency invoices in one system, so disputes about hours are settled with data rather than message threads.

**800** person-shifts a day at one fulfillment center

**400 × 2 shifts** day and night, every day

**Team experience** HRMS development at Yandex Market

### A timesheet everyone kept their own way.

Timesheets were kept by hand or half by hand. Warehouses know the “buddy punch”: one worker checks in a friend who never showed up. Every such shift gets paid.

Staffing was planned by eye, so quiet hours had too many people and peaks had too few. The agency sent its invoice, the site kept its own timesheet, and differences were chased in message threads. Pay depended on hours, not on output.

### One chain: shift planning, check-in, timesheet, agency payment.

Before the shift, staffing needs come from forecast volume and productivity norms. Without a valid medical check, safety briefing or equipment permit, the system will not let a person start. Requests to agencies live in the same place.

During the shift, workers check in biometrically. The timesheet follows actual attendance, and overtime is tracked. Completed operations add up to output, which drives piece-rate pay.

After the shift, the agency sees requests, no-shows, hours and invoices in its own portal. A timesheet correction needs a reason and an approval, and it stays in the history.

### Reconciling agency hours now starts from one table.

The site manager sees the plan, check-ins and actual hours side by side. When numbers do not match, it is clear where: who did not show up, who stayed late, which correction was made and who approved it. The agency sees the basis for payment instead of screenshots from a messenger.

The detail that changes the conversation: nobody can edit a timesheet silently. Every change has a reason, an approval and an author.

### How this relates to Hrusty.

This is our team’s experience at Yandex Market. Hrusty is a separate product we are building for frontline teams in warehouses and stores. The project shows the team’s operational experience; it does not mean Yandex Market uses Hrusty.

### See what this could look like for your team.

Blog

### Further reading

- [Warehouse incentive pay and piece rates, without the disputes](https://opsalute.ai/en/blog/piece-rate-pay) — Warehouse incentive pay people can trust. Why hourly pay caps output and how piece rates work with per-task standards, scan data and a balance workers see.
- [Where warehouse labor costs leak, and how to find it in a week](https://opsalute.ai/en/blog/frontline-payroll-leaks) — Five places warehouse labor costs leak, from guesswork timesheets and buddy punching to agency no-shows, unchecked overtime and workers waiting for tasks.
- [Warehouse task assignment without the queue for the supervisor](https://opsalute.ai/en/cases/taskrouter) — A worker finishes a task and waits. How sending warehouse tasks straight to terminals cut waiting by 30%, and why supervisors were slow to let go of top people.