Vela Commerce — B2B Dashboard Overhaul

Overview

A logistics SaaS platform used by 3,000+ operations teams daily had a dashboard problem: too much data, too little direction. Operations managers were drowning in metrics but starving for answers. I restructured the entire dashboard experience — rethinking navigation, data hierarchy, and density — to turn a reporting screen into a decision-making tool that ops teams opened first thing every morning.

The Business Problem Operations managers spent an average of 18 minutes each morning piecing together their daily priorities from 5 different dashboard views. Critical shipment exceptions were buried in the same visual hierarchy as routine metrics, causing delayed responses that cost an average of $4,200 per missed SLA.

The Design Goal Reduce daily priority identification to under 5 minutes, surface shipment exceptions within the first screen, and consolidate the 5 existing dashboard views into a single adaptive workspace.

01. Research & Discovery

I started by spending a week in the operations centers of two mid-size logistics companies — not interviewing, just observing. The way ops managers actually used the dashboard was fundamentally different from how they described using it.

Methodology:

• 5 days of contextual observation across 2 operations centers

• 16 semi-structured interviews with ops managers, dispatchers, and account leads

• Heatmap and click-path analysis of 90 days of dashboard usage data

• Information architecture audit of the existing 5-view dashboard structure

Key Insights:

1. The Morning Ritual: Every ops manager had built a personal workaround — browser bookmarks, saved filters, even sticky notes on monitors — to assemble their morning view. The dashboard's default state was useful to nobody.

2. Exception Blindness: Shipment exceptions were displayed in the same card format as healthy shipments. In a grid of 200+ shipments, a delayed one looked identical to an on-time one until you clicked into it.

3. The Navigation Tax: Moving between the 5 dashboard views required full page loads averaging 3.2 seconds each. Over a day, ops managers spent 12+ minutes simply waiting for views to load.

02. Framing the Problem

I distilled the research into a problem frame that forced us to design for the first 60 seconds of an ops manager's day — the moment with the highest leverage.

How might we show an operations manager exactly what needs their attention right now — without requiring a single click, filter, or page navigation?

03. Ideation & Prototyping

I facilitated a 3-day design sprint with the product team, two ops managers (one from a small account, one from enterprise), and the engineering lead responsible for dashboard performance. We generated 40+ concepts and converged on three structural changes that redefined the experience.

Architectural Shifts:

• Exception-First Hierarchy: Inverting the dashboard's information architecture so that anomalies — delayed shipments, SLA risks, capacity warnings — surface at the top of every view. Healthy operations get less visual weight. The dashboard now answers "what's wrong" before "what's happening."

• Unified Workspace: Collapsing the 5 separate views into a single-page experience with contextual panels. Navigation becomes scroll and expand, not click and wait. Page load dropped from 3.2 seconds to zero because there's only one page.

• Glanceable Status Strips: Replacing detailed data cards with condensed status strips for healthy shipments — a single row showing route, ETA, and status color. Detail expands on interaction. This reduced the visual footprint by 60%, making exceptions impossible to miss.

04. High-Fidelity Execution

With the architecture validated through guerrilla testing with 8 ops managers, I built the production design system. The visual language was industrial and utilitarian — monospaced type for data, high-contrast status colors, and a deliberately unsexy aesthetic that prioritized scan speed over polish. This wasn't a marketing page; it was a cockpit.

05. Outcomes & Impact

The redesigned dashboard launched to all 3,000+ ops teams after a 3-week phased rollout. Impact was measured at 60 days post-launch.

The Results:

• -72% Morning priority identification time (18 min → 5 min average).

• -38% Missed SLA incidents attributed to late exception detection.

• 5 views → 1 — consolidated dashboard eliminated all page-to-page navigation.

• 4.6 / 5 Ops manager satisfaction score, with "speed" rated highest.

Reflection: The biggest lesson was about data density versus data clarity. My first iteration cut too much information in pursuit of simplicity, and ops managers pushed back hard — they needed density, just better-organized density. The final design actually shows more data than the original; it just structures it so the eye knows where to go first.