Problem
Years of fixes layered on fixes. Six core usability failures compounding across hundreds of daily interactions.
Solution
Two-panel configuration. Collapsible filtering. Inline status editing. Unified component system.
Impact
Order status updates 5 steps → 2. Approved as development basis without revision rounds.
During my internship at LuminateOne, Automote became my core project. It's a stock and inventory platform used daily by dealership staff to track orders, manage vehicle builds, and monitor inventory.
The existing interface had accumulated years of fixes. My job: redesign the core workflows in Figma, producing a fully interactive prototype for development.
Dealership staff work under time pressure with real financial transactions. Confusion means mistakes with money and customer relationships.
Steps to update an order status after redesign
Reported by LuminateOne stakeholders · Validated in walkthroughs
Based on stakeholder feedback and internal LuminateOne audit notes, failures clustered across three themes.
The original interface. Text truncated at edges, no visual hierarchy, actions requiring multiple navigations away and back.
Too many steps to reach common actions
Stakeholders reported dealers memorising locations rather than discovering them. Flat hierarchy with unlabelled sections.
No visual hierarchy between actions
Primary and secondary CTAs looked identical. Dealers hesitated or clicked wrong buttons, requiring undo steps.
Layout broke across different screens
Built for one fixed resolution. Components collapsed or overlapped unpredictably on different displays.
Actions completed silently
No confirmation, no errors, no loading indicators. Dealers couldn't tell if something had worked.
The most frequent dealer action. Redesigned for inline editing with immediate confirmation.
Stakeholder walkthroughs confirmed dealers found this significantly faster.
Sketched out the core problems before touching Figma. Getting the logic right on paper is faster than getting it wrong in high-fidelity.
Identifying the navigation and workflow failures before any design decisions
Mapping which interactions caused the most friction and where steps could be removed
First-pass high-fidelity screens to test layout logic and information hierarchy. The goal was validating the approach, not visual polish. These informed what got built and what got cut.
Filter panel layout: testing how much of the primary view to preserve when filters are open
Second pass: refining how filters group and whether they collapse cleanly
Third pass: settled on the collapsible sidebar approach
High-fidelity draft: filter panel grouped by category, visual hierarchy tested
Configuration flow: testing step structure before the two-panel layout was locked
Editing flow draft: inline save and cancel hierarchy tested here before the final approach
Step progress draft: numbered flow tested here, simplified in final
Where dealers spend most of their time. Filterable, scannable, and clear about which orders need action.
Order management: filterable by ID, customer, model, date, and production status. Visual weight distinguishes action-required statuses.
.png)
Filter panel open. Collapsible sidebar keeps the primary content visible while filtering
Inline order editing. Status updated without leaving the screen
Inventory overview: consistent row structure and quick-action controls
The two-panel layout keeps order details and configuration steps visible at the same time. No navigating back and forth.
Two-panel configuration: order details persistent on the left, steps on the right
Trim level selection with visual preview
In stakeholder reviews, dealers confirmed they could find orders faster and update statuses without leaving the screen. Approved as the development basis without further revision.
LuminateOne stakeholder feedback · 2024
Two-panel configuration
Details and steps in persistent split view. Dealers never lose context mid-task.
Tradeoff: more complex layout. Worth it. Eliminated mid-task navigation entirely.
Collapsible side filter
Visible when needed, hidden when not. Filters by ID, customer, model, date.
Tradeoff: additional UI element. Worth it. Dealers found orders significantly faster.
Status as text labels
Production status as readable text, not icons alone. No legend required.
Tradeoff: less visual minimalism. Worth it. Zero misinterpretation in testing.
Structured sidebar nav
Tools grouped into logical categories with persistent context across all screens.
Tradeoff: requires initial learning. Worth it. Dealers always knew where they were.
Primary/secondary hierarchy
Primary actions use filled buttons. Destructive actions deliberately subdued.
Tradeoff: less visual uniformity. Worth it. Reduced accidental destructive actions.
Unified component system
Spacing, type, buttons, forms applied consistently across every screen.
Tradeoff: initial setup time. Worth it. Users relied on patterns without hesitation.
Status column not scannable
Dealers scanning a long list couldn't quickly identify which orders needed action.
Visual weight for action-required statuses
Subtle colour coding for statuses requiring action. Text labels retained. Colour additive, not replacing.
Bulk actions missing
Sales managers needed to update multiple orders at once. v1 required individual updates.
Bulk selection and action bar
Row-level checkboxes and a contextual action bar that appears when items are selected.
Constraints
Pain points confirmed solved
LuminateOne confirmed the redesign addressed navigation and clarity issues. Approved without further revision.
Reduced cognitive load noted
Two-panel configure screen and collapsible filter specifically cited as solving dealer problems.
Development ambiguity reduced
Interactive prototype with annotated specs and documented decisions enabled clearer planning.
Not having direct access to end users was the biggest constraint. Even one session watching a dealer use the existing system would have been worth more than hours of assumption. I'd push back harder on this in future.
The most valuable pattern that emerged: in testing, users relied on repeated patterns to navigate without hesitation. The interface should disappear. Consistency reduced cognitive load more than any clever interaction.
If I were to continue this work: bulk actions would be phase two, and I'd explore saved filter views for high-volume dealers.
Next project
Splashy →