Technician Dashboard
Evaluating and Designing Prepango’s Internal Tool
The Tool That Didn't Fit the Job
The Technician Role
Prepango's field technicians managed vending machine restocking, maintenance, and reporting across multiple locations per shift at airports or other vending locations.
The Reality
They worked on their feet in high-traffic areas with unreliable wifi, moving between machines with heavy carts and tight windows for each stop.
The Gap
The portal they used was built for administrators at desks. It had no offline support, no status hierarchy, and no awareness of how the job actually moved.
The existing portal experience diminished efficiency to a point where technicians were serving the software, instead of the software serving them.
What the Portal Covered
Before redesigning anything, I mapped the existing portal as a product surface, not in terms of what it was supposed to do, but in terms of what technicians actually encountered when they used it.
Planogram details tableA data-heavy table nested inside the restock form showing item name, price, capacity, and minimum unit thresholds. No sorting, no filtering, no urgency signaling. Every row read the same regardless of stock level.
Restock entry form The primary action surface for logging a completed restock. Machine selector dropdown, manual date and time inputs, and a Create Report button. No prefill, no smart defaults, no context about what actually needed restocking.
Restock report edit screenPost-submission editing view with fields for technician name, timestamps, machine location, and a free-text notes area. Action buttons stacked below the form: Edit Packing Slip Photo, Add Before Photo, Add Opening Counts, Items Added. Each action opened its own isolated page with no surrounding restock context.
Photo upload flowsDedicated full-page screens for attaching photos. File chooser only. No preview, no confirmation beyond a success banner after the fact.
List ViewsTabular lists for browsing restock history: List Restocks, Incomplete Restocks. No filtering, sorting, or urgency differentiation.
Specialized InputSeparate form flows for edge cases: Input Cash Only Report, Input Service Report, Tech Invoice Recon.
Global NavigationNot role-differentiated. Customer Service, Operations, Reports, Setup, Sprinkles, Inventory Counts; all visible to everyone regardless of what their job actually required.
Going On-Site
Before defining anything, I went into the field. I shadowed two technicians in person: one at the Las Vegas airport, and another in San Diego, specifically to restock the Sprinkles Cupcakes machines on her route.
The Las Vegas visit was September 12, 2024. I spent a full day with field technicians Daniel and Miguel, following them through a complete restocking shift across airport terminals, during which I observed details that no remote interview would have captured:
False restock alerts were sending technicians across terminals for nothing.
The picklist was requesting more units than the machines could physically hold.
Technicians had stopped trusting the system and were relying on paper counts.
When wifi dropped, which it did constantly, there was no offline fallback.
The San Diego visit was a different context entirely. One product line, a fixed route, and a job built around speed. Watching that shift made clear that the problems in Las Vegas weren't specific to airports or scale. They were about a tool that had been built around the data model, not the person using it.
From there I ran structured phone interviews with technicians across all four archetypes to validate what I'd seen on the ground and fill in the gaps across different roles, routes, and locations. That combination of observation and interviews is what made the archetypes concrete. Not assumed profiles, but patterns documented across real people doing real work.
Players, Parts, and Key Interaction Points
FOUR USER ARCHETYPESThe research made one thing clear: the technician population was more segmented than anyone had mapped. Four archetypes emerged, each with different responsibilities, different mental models, and entirely different needs from a dashboard.
❋ The illy SpecialistWorks exclusively with illy coffee products, managing perishable inventory and freshness windows. They need precise unit counts, not percentages. Their ask: show me only what I need, nothing extra.
❋ Sprinkles OnlyHandles one product line across multiple locations and needs to move fast. Minimal friction, fast in and out. Their ask: my job is simple, the tool should be too.
❋ Expert FacilitationJuggles multiple task types across a shift, balancing restocks, service calls, and reporting while avoiding illy and Sprinkles tasks. They need the system to do the prioritizing. Their ask: tell me what to do first and where to go.
❋ The 100%The full coverage technician. Handles everything except illy, works across multiple machine types, manages restocks, service, and reporting all at once. Their ask: give me the full picture.
Designing for The Floor Not The Desk
Re-framingThe existing portal was form-based with no status hierarchy or offline support. It treated every user the same and assumed a reliable connection throughout. Rebuilding it meant starting from a different set of questions.
PrioritizationThe wireframe phase organized information around how technicians actually move through a shift, not around how the database was structured.
From the Field Up
The final concept was grounded entirely in what we found on-site. Offline first data entry with automatic sync when connectivity is restored. Alerts sorted by severity. Counts in units, not percentages. A dashboard that adapts to whoever is logged in, pulling the right view for their role rather than showing everyone the same thing.
The project reached the concept phase. Based on the research, we projected a 25% to 35% reduction in average task time and up to 50% fewer errors through real-time updates, offline syncing, and standardized workflows.
The archetypes were defined, the wireframes annotated, and the concept screens built as a grounded starting point for whatever came next.
All Prepango Work