In numbers
The problem
Farm managers handed out jobs by phone and WhatsApp. Workers lost time hunting for the exact spot on big properties, and no one had a record of what was done.
- 3 of the last 5 onboarding calls asked: “Can I assign jobs?”
“Which trough? There are four up that end.”
How might we let a farm manager hand over a job at a place, and know when it's done?
How I took it from idea to release-ready
The rhythm I used for every Blockbase feature.
Align seven roles around the problem and scope
Features usually came from the PM or CEO. I ran 90 minutes on the problem, who it's for, an assumption map, dependencies and risks, scope and T-shirt estimates.
WithPMCEOPOBAFEBE
Farm Tasks discovery workshop, 16 Aug. I facilitated, and every action left the room with an owner. Consolidate and delegate
I wrote up decisions and actions with owners, and handed scope to the BA for tickets.
WithBAPO
Workshop decisions and requirements, logged in the PRD.
Learn the farm, then map the flow
Our in-house SME gave me the context of use: patchy signal, 5 am starts, gloves and sun glare, and owners who think in paddocks, not coordinates. Then a hand-drawn flow, tidied into a user flow from dropping a pin to the assignee's email.
WithSME
Rough flow, before team walkthroughs. The same flow, tidied into a user flow. Walk it through with the team
Their feedback went into journey map v2, which made the lowest point clear: finding the spot.
WithPOBAFEBE
Journey map v2, after the SME chat and walkthroughs.
Hand sketches, reviewed early
Sketches of the create-task panel and task list, reviewed with the design team and my manager before any pixels.
WithDesign teamManager
Create task: two required fields. Task panel: select a task, zoom to its pin. Mid-fi, demoed and tested
Demoed to whoever needed it, in a meeting or a Loom. When feedback raised a real question, like “Is a pin enough to find the spot?”, I tested the mid-fi with pilot farms before moving to UI.
WithFEBEBAPOPilot farms
UI, test, document, hand over
I designed the UI, ran user testing and documented feedback. Small changes I made myself; bigger ones went back to the team.
WithPOFEBE
Key decisions
Only title and assignee are required
A long form is a chore between jobs, so due date defaults to today and priority to Medium.
Select a task, zoom to its pin
Fixes the lowest point in the journey: finding the exact spot on a 2,000 ha property. Pins carry paddock names, because owners think in paddocks, not coordinates.
Email now, push later
Email reused an existing service; push and offline waited for the mobile app.
Working with stakeholders
Blockbase went through restructures while I was there, so these constraints shaped the work as much as the brief.
Who I worked with, and on what
Blockers, and how I moved them
- Back-end capacity was thin for the next sprintReused the existing email service, spiked by BE before we committed.
- The farm map wasn't tuned for tablets yetDesigned for desktop in the farm office first, with tablet logged as a known risk.
- “Is a pin enough to find the spot?”Tested on mid-fi with pilot farms, with paddock names as the fallback.
What we set out to hit
v1 was planned for release behind a flag, to three pilot farms first, with a six-week pilot.
Pilot success metrics I defined with the PO. I left before the results came in, but setting them up front shaped every design decision.
Reflection
Facilitating the workshop myself meant I knew the reasoning behind each decision. When managers changed, four times, I could explain a call instead of reopening it, and the decisions log kept Farm Tasks on track.