🧭 Project Overview
Project: CityUnderConstruction (NYC Construction Visibility App)
Role: Product/UX Design (end-to-end)
Platform: Mobile + Desktop (responsive map-first experience)
Timeline: MVP-first, iterative validation
🎥 Product in Action
A map-first experience that turns fragmented civic data into a clear, interactive system.
What it is
CityUnderConstruction makes NYC construction activity legible and explorable for non-experts by turning fragmented public data into a map-first product with address lookup, parcel detail cards, and permit history.
Who it’s for (MVP focus)
Real Estate Watchers and Curious Urbanists (primary for v1)
Construction Professionals (secondary early audience; more relevant in later tiers)
Why now
NYC construction is everywhere, but understanding it still requires navigating fragmented, technical systems (DOB tools, ACRIS, zoning maps, news blogs). The gap is a modern, mobile-friendly way to see: what’s being built, where, and what it means.
🧩 The Challenge
Problem
NYC’s construction ecosystem is served by tools built for professionals and bureaucratic workflows. For everyone else, the experience is:
fragmented research workflows
inconsistent formats
unclear real-world meaning
no map-based “single source of truth”
People end up relying on:
manual searches across multiple city systems
neighborhood walks + rumor threads
niche blogs and social posts that cover only parts of the story
Design challenge
How might we transform high-friction civic data into a product that feels:
fast to understand
easy to browse
credible enough to trust
simple enough to return to weekly
💡 The Opportunity
Product thesis
If we make construction activity visual + searchable + contextual, we can turn passive curiosity (“what’s that scaffolding?”) into repeat behavior.
Business outcomes (AARRR)
Success for the MVP is traction and behavior, not monetization:
🌐 Acquisition: map discovery + first-time searches
🚀 Activation: users open a parcel card and get an “aha” moment
🔁 Retention: users come back within 7 days to re-check areas/projects
🔗 Referral (lightweight): link sharing / copy link behaviors
💰 Revenue: intentionally excluded from MVP
User outcomes & benefits (HEART)
The MVP aims to deliver:
❤️ Happiness: clarity + delight in exploring the city visually
⚡ Engagement: map interactions, multiple parcel deep-dives
🌱 Adoption: prosumers use it as a repeat tool
🔁 Retention: weekly/daily check-ins
🎯 Task success: quickly answer “what’s being built here?”
🔍 Research & Insights
What I did (MVP-appropriate research)
Because the pain is already well-known and the constraint is data complexity, early research focused on:
understanding how existing systems fragment workflows
learning what data is usable and meaningful for non-professionals
identifying the minimum set of fields needed to build trust quickly
Key insight
People don’t want “all the data.” They want legibility.
They want:
a simple “What’s being built here?” answer
status signals (“Active” vs not)
a trustworthy trail (permit history)
parcel context (BBL, district info)
Implication for design
The product must prioritize:
map-first discovery
one-tap parcel comprehension
structured info hierarchy
progressive disclosure (summary → expand details → history)
✏️ Design Process
Approach
I used a Lean UX Canvas to keep scope disciplined and tie every feature to:
a business outcome (AARRR)
a persona outcome (HEART)
a measurable hypothesis
MVP features chosen (Ship & Measure)
These were selected because they are high value + low risk:
Map view (permit points / parcel summaries)
List view (permits grouped by parcel)
Simple filters (permit + parcel filters)
Address lookup + instant “what’s being built” summary
Parcel detail page
Permit history per parcel
Save list (bookmarks)
30-second onboarding tour
Risky experiments deferred (Test)
High value but higher uncertainty / complexity:
Neighborhood time-lapse
Parcel watch + notifications
Competitor activity heatmap (kept name as-is)
Parcel timeline visualization
⚙️ Collaboration & Leadership
What I owned
End-to-end product definition from problem → MVP scope
Information architecture and interaction model (map → parcel card → history)
Design decisions grounded in measurable hypotheses
Data UX strategy: how to present civic data without overwhelming users
Product judgment (scope discipline)
A key leadership decision was to exclude monetization and advanced pro tooling from the MVP. The goal was to first prove:
desirability (people return)
usability (people can understand quickly)
credibility (people trust the data)
✅ The Solution (at a glance)
Experience summary
CityUnderConstruction is a map-first interface where users can:
explore construction activity visually
search for an address
open a parcel card
see “What’s being built here,” and expand for details
review permit history and parcel context
UI structure (current design)
The core UI hierarchy is:
Search bar (entry point)
Parcel detail sheet (core “aha” container)
Three sections:
What’s Being Built Here
Permit History
About This Parcel
Progressive disclosure: “Expand Details” reveals more fields
Empty-state handling
When no active permits exist, the system clearly communicates it (“No active construction permits at this address”) while keeping the parcel context accessible.
📊 Impact
What success looks like (MVP metrics)
🌐 Acquisition
growth in first-time visitors
% of visitors who perform a search
🚀 Activation
% opening at least one parcel detail card
time on map during first session
early filter usage
🔁 Retention
% returning within 7 days
repeat searches for known parcels/neighborhoods
revisits to saved parcels (if implemented)
🔗 Referral (lightweight)
link copy/share behavior
traffic from shared links
💰 Revenue
intentionally excluded from MVP definition
🧠 Reflection
What I’m validating next
The MVP proves core value if:
users repeatedly return to check known areas
the parcel card becomes a “destination”
permit history builds trust and legitimacy
Biggest risks to test after MVP
Do users truly want historical visualization (time-lapse), or is raw history enough?
Can alerts drive retention without becoming noisy?
Do professionals interpret heatmaps as actionable intelligence?
What I’d improve with more time
deeper filtering and “pattern exploration” for urbanists
clearer “status language” and data transparency
stronger shareable outputs (cards, snippets) for social virality