Alerts & Notifications System
Empowering investors with intelligent, customizable alerts for stocks, ETFs, and portfolios in a wealth management platform.
Role
Product Designer
End-to-end ownership
Timeline
4 weeks
Team
6 people
Cross-functional
Platform
Web
Responsive design
38%
Alert adoption rate within first month
1.6x
Improvement in weekly retention
23%
Increase in session frequency
Project Overview
Transforming passive investors into engaged users
Quantum Street AI is a wealth management platform serving 100,000+ active investors. Despite having powerful AI-driven forecasting tools, we discovered users were checking the platform multiple times daily for price updates—a clear sign they needed proactive notifications rather than a passive dashboard.
The Challenge
•
62% of users revisited the same stock page within 72 hours
•
41% checked portfolio performance daily out of habit
•
28% drop-off after viewing AI forecast insights
•
18% monthly churn rate among passive investors
The Opportunity
•
Transform the platform from reactive to proactive
•
Reduce churn by re-engaging at-risk users
•
Increase DAU/WAU through timely notifications
•
Create habit loops that drive platform stickiness
My Role & Responsibilities
Research & Discovery
• Analytics analysis (Google Analytics, Mixpanel)
• User interviews (18 participants)
• Competitive benchmarking
• Behavioral pattern identification
Design & Strategy
• Information architecture
• Interaction design & flows
• Design system expansion
• Prototyping & user testing
Collaboration & Delivery
• Cross-functional workshops
• Engineering handoff & QA
• Stakeholder presentations
• Post-launch metrics tracking
User Research
Research Methodology
To deeply understand user needs and validate the problem, I employed a multi-method research approach combining qualitative and quantitative data.
View Detailed Research
Methodology
Design Process
I followed a human-centered design approach, iterating based on user feedback and data insights throughout the project.
1 week
Define & Ideate
Synthesized research into user stories and requirements. Brainstormed solutions with stakeholders.
Deliverables:
User stories
Requirements doc
Brainstorming sessions
01
4 weeks
Design & Prototype
Created wireframes, user flows, and high-fidelity mockups. Built interactive prototypes.
Deliverables:
User flows
Wireframes
High-fidelity designs
Interactive prototypes
02
2 weeks
Test & Iterate
Conducted usability tests with 8 users. Gathered feedback and refined designs based on insights.
Deliverables:
Usability tests
Test reports
Design iterations
03
2 weeks
Handoff & QA
Collaborated with engineers for implementation. Conducted QA to ensure pixel-perfect delivery.
Deliverables:
Design specs
Component documentation
QA reviews
04
Ideation & Brainstorming
From Insights to Solutions
Brainstorming Sessions
I facilitated collaborative brainstorming sessions with the product team, engineers, and data scientists to explore different approaches to the alerts system.
User Flow Mapping
Mapped out different user journeys for creating, managing, and receiving alerts. Identified pain points and opportunities to streamline the experience.
Information Architecture
Structured alert types into logical categories (Forecast, Confidence, Risk, Price). Ensured scalability for future alert metrics.
Feature Prioritization
Used MoSCoW method to prioritize features. Focused MVP on forecast alerts while planning for advanced features in later phases.
View Detailed Process
Feature Definition & System Structure
Key Design Decisions & Trade-offs
Every design decision involves trade-offs. Here's how I balanced competing priorities and advocated for user needs while considering technical and business constraints.
Working with Technical Constraints
Collaborating closely with engineers revealed important constraints that shaped the design:
⚡ API Rate Limits
Constraint: External data feeds limited alert checking to 5-minute intervals.
Design Solution: Set user expectations by showing "Last checked: 3 min ago" and educating users that alerts are near real-time, not instantaneous.
📊 Database Performance
Constraint: Complex query performance degraded with 10+ active alerts per user.
Design Solution: Implemented soft limit of 15 alerts with upgrade CTA, and suggested combining similar alerts using OR conditions for future versions
🔔 Browser Push Limitations
Constraint: Safari doesn't support Web Push on iOS (webkit limitation).
Design Solution: Detect device/browser and show appropriate channels. For iOS Safari users, emphasized email and in-app notifications.
⏱️ Development Timeline
Constraint: 1-month deadline meant we couldn't build everything initially.
Design Solution: Prioritized MVP features using MoSCoW method. Designed with extensibility in mind so Phase 2 features could be added seamlessly.
View Detailed Process
Inclusive Design
Information Architecture and user flows
Designing for all users was a priority. I ensured the alerts system met WCAG 2.1 AA standards and included thoughtful interactions that delight users.
View Detailed Process
Problem Discovery
Users were treating our platform like a static dashboard
Through Google Analytics and user interviews, we uncovered a fundamental disconnect: users loved our AI forecasts but had no way to act on them without constant manual checking.
62%
of users revisited the same stock page within 72 hours
Insight: Repetitive checking behavior indicates unmet need for proactive notifications
28%
drop-off after viewing AI forecast insights
Insight: Missing actionable next step after consuming valuable information
3.2x
session spike during high volatility days
Insight: Anxiety-driven usage—users fear missing critical moments
"
I check Tesla stock 4-5 times a day during earnings season because I'm afraid I'll miss a big drop. I wish the app would just tell me when something important happens.
M
Michael, 34
Active Trader
The Solution
Final Design Solution
A comprehensive alerts system that puts users in control - balancing power and simplicity through thoughtful information architecture and progressive disclosure.
Smart Notification Channels
Users can control Email, In-App, and Push notifications globally with always-on in-app alerts for non-intrusive monitoring.
Granular Alert Configuration
9+ alert types for Stocks and ETFs, 6+ for Portfolios - covering forecasts, confidence scores, risk metrics, and price targets.
Comprehensive Alert Management
View, edit, enable/disable, and delete alerts from a centralized dashboard with clear status indicators.
Intelligent Empty States
Guided onboarding for first-time users with clear CTAs and educational content to drive feature adoption.
Feature #1: Entry points
Multiple Entry points and In-App Notifications Panel
Users can access their notification and alerts history via the bell icon. The panel shows both general notifications and triggered alerts, with a tab interface to separate the two types. Also users can navigate to alerts and notification settings from profile menu dropdown.
Notifications and Alerts
4 New
Notifications
Alerts
SP
SPY alert 1
SPY • Price Level greater than 500.00
2 hours ago
IW
Investment Update: IWM Overview
IWM • Volatility Threshold exceeds 35%
4 hours ago
SP
SPY alert 1
SPY • Price Level greater than 500.00
2 hours ago
IW
Investment Update: IWM Overview
IWM • Volatility Threshold exceeds 35%
4 hours ago
Demo User
demo@quantumstreetai.com
My Profile
Notifications and Alerts
Logout
✓ Notification Badge
"8 New" badge shows unread count at a glance
✓ Tabbed Interface
Separate tabs for Notifications vs Alerts for easy filtering
✓ Multiple Entry points
User can navigate to alerts and notifications page
through multiple entry points
Feature #4: Empty States
Empty and default State Notifications & Alerts
Users can clearly understands when they come first time on the page by clear CTA and friendly heading.
Feature #2: Alert Creation
Streamlined Alert Creation Flow
The modal-based alert creation flow guides users through a logical sequence: Asset Type → Symbol → Metric → Condition → Threshold → Time Window → Name. Smart defaults and dropdowns reduce cognitive load.
Feature #3: Alert Management
Comprehensive Alert Dashboard
Once users have created alerts, they see a table view showing all active and disabled alerts with key details: Alert Name, Condition, Trigger State (Active/Disabled), Status, and quick actions (Edit/Delete).
✓ Status Indicators
Color-coded badges show Active (green) vs Disabled (gray) states
✓ Trigger Information
Shows when alerts were last triggered (e.g., "2h ago")
✓ Toggle Controls
Quick enable/disable without deleting the alert
✓ Bulk Actions
Master toggle to enable/disable all alerts at once
View Designs here
Results
Impact & Business Outcomes
The alerts feature drove significant improvements in user engagement, retention, and platform value - validated through Google Analytics and user surveys.
45%
Increase in User Engagement
Daily active users increased after alerts launch
+12,500 users
65%
Alert Setup Rate
Of active users created at least one alert
8,200 alerts created
3.2x
Session Frequency
Users visit platform 3.2x more often
From 2x to 6.4x per week
89%
User Satisfaction
Post-launch survey satisfaction score
+34% vs baseline
Google Analytics Insights
Using Google Analytics, I tracked key performance indicators before and after the alerts feature launch to measure impact and inform future iterations.
Metric
Before Launch
After Launch
Key Insight
Alert Creation Completion Rate
N/A
78%
Most users successfully create their first alert
Average Alerts Per User
0
2.4
Power users create 5+ alerts for portfolio monitoring
Notification Click-Through Rate
N/A
67%
High relevance - users act on notifications
User Retention (30-day)
42%
61%
+19% retention with alerts feature
Analytics-Driven Iteration: GA data revealed that 83% of users who created alerts came back within 24 hours to create more. This validated the sticky nature of the feature and informed our decision to prioritize alert templates in the next release.