Turning a Dashboard into a Decision Surface
Overview
Rather than simply displaying data, the new Overview experience prioritizes relevance, context, and usability. The goal was to help diverse stakeholders quickly understand what matters most, reduce cognitive load, and make faster, more confident decisions.
Challenge
Last Mile Solutions’ platform serves a complex ecosystem of users:
Charge Point Operators, eMSPs, financial analysts, customer success teams, and end users — each with very different goals, metrics, and daily decisions.
Yet all of them were landing on the same Overview page.
Originally, this page displayed operational data (charger status, energy usage, transactions), mainly for CPOs. But research and usage data revealed a deeper problem:
Most users didn't find the overview useful. It failed to deliver real value—even to CPOs.
Business Objectives
We aligned with leadership and stakeholders around measurable impact:
Reduce Time-to-Insight
Users answer their most important question within 15 seconds.Increase Meaningful Engagement
Users take actions directly from the overview.Improve Retention & Adoption
More repeat usage across roles, fewer support requests for data discovery.Increase Decision Confidence
Measured through surveys after using the new overview.
User Research
Phase 1 — Understanding the Real Problem
We began by interviewing users across all core roles and internal stakeholders—to ground the redesign in real needs rather than assumptions.
Our research goals were clear:
Understand pain points and frustrations with the current overview page
Identify the most valuable information and features each user group expects to see
Commonalities and differences between user needs
Explore user preferences for how the overview should behave
(widgets, templates, multiple views, role-based dashboards, etc.)
These conversations quickly revealed an important shift in perspective.
Key Insight — Users Care More About “What” Than “How”
When we asked users how they wanted the overview to work, many initially discussed interface preferences.
But as interviews progressed, a deeper truth emerged:
What mattered most to users was not how the dashboard behaves, but whether it gives them the right data to make decisions.
This reframed the entire project.
Instead of leading with layout patterns or interaction models, we realized the core design challenge was deciding what information truly belongs on the overview.
This was further reinforced by practical constraints: backend capacity was limited, which meant we had to be extremely deliberate about which features and metrics we prioritized.
From Features to Decisions
To identify what really mattered, we shifted our interviews from feature requests to decision-driven conversations.
For every data point users asked for, we followed up with:
“What decision will this help you make?”
“How often do you make that decision?”
“How critical is it to your daily or weekly workflow?”
This approach helped us separate:
Critical decision-enabling data from
Nice-to-have metrics that created noise.
Research Outcome — Defining What Truly Matters
We synthesized all interview insights, data requests, and constraints into a clear outcome:
We could not build everything at once — so we needed to ship the most decision-critical metrics first.
To do that, we mapped every requested metric against:
Decision impact (How much it influences real decisions)
Frequency of use (How often users need it)
Role criticality (How essential it is for each role)
Feasibility (What we could realistically deliver in phase one)
This allowed us to define a first-release insight set for each role.
Defining the Behavior — How the Overview Should Work
With clarity on what information mattered most, we moved to the next critical question: How should the overview behave for different users?
We intentionally explored multiple structural models before committing:
Role-Based Default Dashboards
Each role sees a tailored overview by default.Fully Customizable Widget-Based Overview
Users assemble their own dashboard from components.Hybrid Model: Role Defaults + Light Customization
Role-optimized foundation with limited personalization.Multiple Switchable Dashboards
Users manually switch between predefined views.Separate Overviews by Domain
Individual overviews for Cards, Stations, Customers, etc.
At the same time, we needed to respect real technical constraints and avoid over-engineering.
Comparative Concept Testing
We conducted comparative concept testing using fast interactive prototypes built with Cursor, each representing a different dashboard model. We focused on qualitative feedback, asking users to walk through real scenarios and share their challenges, expectations, and preferences while using each concept.
The results clearly favored the hybrid model: role-based default dashboards with light customization. Users consistently described it as the most intuitive, least overwhelming, and best aligned with how they think about their work, while the other models introduced unnecessary complexity or fragmented their workflows.
Product & User Impact
While the solution has not yet been fully implemented, these outcomes were consistently observed and validated through usability testing and stakeholder evaluations.
Time-to-Insight ↓
Participants reached key answers significantly faster when using the new overview concepts.Task Success Rate ↑
Users completed core decision-making tasks with higher accuracy and fewer errors.Role-Based Adoption Potential ↑
Testing showed strong engagement and preference from roles beyond CPOs, including analysts and eMSPs.
Potential Organizational Impact
Sales demos could become faster and more compelling
The redesigned overview is expected to communicate the platform’s value within the first minutes of a demo.Customer success onboarding time could decrease
The clearer, role-based experience should enable new users to ramp up faster with less guidance and fewer support touchpoints.
Next projects
Carbon Footprint
Designing an app to reduce the carbon footprint of phone usage for Fairphone.
