Growberry
A workforce productivity system that helped a 1,800-worker farm identify underperformance and act on it fairly, at scale.
Year:
2025
Timeframe:
Tools:
Figma, Glide, Twilio, Looker
Category:
Product Design / Systems
Overview
Turning productivity data into fair, timely action
Farms run on large seasonal workforces and a short harvest window. Underperformance was measured but poorly managed. Feedback was inconsistent, manual, and often never delivered, so slow pickers stayed on the payroll season after season. As product owner, I designed a system that ingests weekly productivity data, surfaces who is underperforming, and turns disciplinary action from an ad-hoc scramble into a managed, accountable process.
Money lost to underperformance no one could act on
In one season the farm lost R1.2m on harvesters who consistently missed target, the gap between wages paid and the value of what was picked. The causes were operational, not just performance. Management couldn't reliably identify who was underperforming, and even when they could, they had no practical way to act: warnings were issued verbally, delivered by hand to a workforce of 1,800 spread across the farm, and recorded inconsistently. When records couldn't be trusted, fair disciplinary action stalled, and the same underperformers were often rehired season after season.
A system to identify, validate, and act, at scale
The system ingests weekly productivity data and surfaces who fell below target. HR reviews each case against mitigating circumstances, with every decision and its reason logged against the worker, so the record is accountable and defensible. Valid warnings are then delivered and tracked automatically over WhatsApp, the most accessible channel for this workforce, and previous warnings are referenced to determine the correct next step. The result is a managed process HR can trust, where fair, timely action replaces an ad-hoc manual scramble.
