Eyefly Analytics
Real Estate Intelligence Dashboard — Zero to Launch
Designed a B2B analytics dashboard turning 47 raw data points into actionable sales intelligence — cutting weekly reporting from 6+ hours to under 1 hour and lifting lead conversion 40%.
Role
Lead Product Designer, UX Researcher, Data Visualization Specialist
Timeline
12 weeks — Discovery to MVP Launch
Year
2023
Team
- 1 Product Manager
- 2 Frontend Engineers
- 1 Backend Engineer
- 1 Data Analyst
The Problem
Why this mattered.
Eyefly's virtual tour platform gave buyers immersive 3D property experiences — but developers had zero visibility into which units attracted interest or which floor plans drove intent. Multiple enterprise clients threatened churn without analytics.
Users who favorited units converted at 8× the rate of casual browsers — this signal existed in the data but was invisible to sales teams. Sales directors spent Sundays doing 6+ hours of manual reporting.
Starting Baseline
6+ hrs weekly manual reporting · 0% data-driven lead targeting · multiple enterprise accounts at churn risk
Discovery
What the research revealed.
8 developer interviews across 3 market segments, 5 sales team workflow sessions, 3 executive interviews, 2 sales presentation observations, 6 competitive analyses, data audit of 47 available data points
"I need to know which units to push before my Monday meeting" — the Sunday deadline drove the entire information architecture
Users who favorited units converted at 8× the rate of casual browsers — favorites became the primary dashboard signal
Developments with 5+ towers required filtering to prevent overload — progressive disclosure was non-negotiable
"I need insights, not just data. Tell me what to do." — recommendation-first layout, not data dump
Solution
The architecture.
Three-level progressive disclosure: glanceable overview (KPIs with trend indicators), filterable analysis (tower-level breakdowns), actionable detail (unit-level availability matrix with visual status encoding).
Level 1 — Glanceable KPIs: visits, favorites, leads, unique users, avg session duration with trend deltas
Level 2 — Tower filtering: pill navigation for 5+ building developments, persistent across all sections
Level 3 — Unit matrix: floor-by-floor grid, bold = available / light = sold, instant visual scan
Insight-first: dashboard leads with the recommendation, data supports beneath
Design Process
Data-dense but scannable. Information hierarchy driven by decision urgency — what do you need to know by Sunday evening?
V1: Data table
All 47 data points in tabular format. Rejected — users felt overwhelmed, not informed.
V2: Chart-heavy dashboard
Better for trends but buried actionable priorities. Sales directors couldn't identify which units to push without digging.
V3: Insight-first layout (Final)
FinalLed with the unit priority recommendation, supported by data beneath. Key metrics identifiable within 5 seconds in testing. Shipped.
Specifications
24 reusable dashboard components. Color-blind accessible palette. 5 usability testing sessions targeting 90%+ task completion. Real production data in prototypes surfaced scaling issues before build.
Constraints & Solutions
Real-time data requirements conflicted with static architecture. Solved with 60-second polling — fast enough to feel live, light enough not to overload the backend at peak Monday usage.
Outcomes
What moved.
| Metric | Before | After | Delta |
|---|---|---|---|
| Weekly Reporting Time | 6+ hrs | Under 1 hr | ↓85% |
| Lead Conversion Rate | Baseline | +40% | ↑40% |
| User Satisfaction (SUS) | 68 (benchmark) | 94 | +26 pts |
| Enterprise Retention | At-risk | 100% | 0 churn |
| Daily Active Usage | 0% | 78% | ↑78pp |
| New Enterprise Deals | Pipeline | 3 closed | Analytics cited |
Impact
"The favorites metric alone helped us close 15% more deals." Analytics became Eyefly's primary enterprise upsell lever — zero client churn post-launch.