From Chaos to Clarity: Redesigning Device Error Communication

01 · Context

The Challenge

Grainger’s KeepStock tools spanned mobile, tablet, kiosk, and vending-device touchpoints, but surfaced more than 100 cryptic and inconsistent error messages. When something went wrong, field technicians and customers often couldn’t determine what had happened, whether they could resolve it themselves, or what to do next. The result was unnecessary support calls and avoidable escalation for problems that could often be handled in the field.

As the sole UX designer on the initiative, I owned the experience from discovery through final design—conducting technician ride-alongs, auditing the existing error ecosystem, defining the new taxonomy and interaction model, and partnering with support and engineering stakeholders through implementation and sign-off.

What I Owned

Audited and categorized 100+ error messages across device types and touchpoints

Conducted ride-alongs with field technicians to observe how errors were diagnosed and resolved under real-world conditions

Partnered with support leadership to identify recurring troubleshooting patterns and escalation points

Developed a behavioral taxonomy that consolidated fragmented error states into three actionable categories

Designed end-to-end error flows that gave users a clear next action based on the type of problem encountered

Translated the framework into reusable interaction patterns that could scale across the KeepStock ecosystem

How the Direction Changed

System failure → user action

The existing experience described what had failed technically. Field research showed technicians primarily needed to know what they could do next.

100+ messages → behavioral taxonomy

Rather than redesigning individual errors, I grouped conditions by recovery behavior and reduced the system to three actionable categories.

Message → recovery path

Each category defined not only what the system communicated, but the appropriate guidance, user action, and escalation path.

One-off fixes → reusable system

The resulting interaction framework could accommodate new error conditions without creating another collection of disconnected messages.

The resulting model shifted the experience from describing system failures to supporting recovery: audit the error landscape, organize conditions by user action, define consistent recovery paths, and translate the model into reusable interaction patterns with engineering.

What I Shipped

40%reduction in support-call volume at launch

Error-State Audit — 100+ messages analyzed across device types

Behavioral Error Taxonomy — three action-oriented categories

Recovery Interaction Model — consistent guidance, action, and escalation patterns

Reusable Error Components — scalable patterns across the KeepStock ecosystem

Field-Validated Flows — informed by technician ride-alongs and troubleshooting behavior

Self-Service Troubleshooting Experience — reduced unnecessary escalation

Why this pattern matters

Design recovery around what the user can do, not what the system knows. Complex systems often organize errors around technical architecture because that is how engineering understands the problem. This work demonstrated that a user-facing recovery model can abstract that complexity into a small number of actionable paths without eliminating the technical specificity needed behind the scenes.