AI-Powered Prediction Platform

Refined the fintech dashboard and subscription process, ensuring accessibility for beginners and depth for professionals.

Overview

Indicator Lab is an AI-powered fintech product for analysts to make predictions. To expand the market, the startup developed new features targeting amateur investors. However, the new pages retained patterns for professional users, causing challenges.


Through in-depth research and multiple iterations with cross-functional teams, I successfully delivered the desired outcomes, securing funding and attracting many retail investors.

Responsibility

Product design Designer - Low to high-fi prototype, Competitor analysis, Usability test, Data visualization

Timeline

July - Oct, 2022 4 months

Collaboration

Designers, 1 PM, 2 Engineers, Finance

Impact

+300%

Achieved an increase in user growth

+ $$$

Succeed funding

-70%

Reduce user first time exploration time

Overview

Business focus shift

Indicator Lab, originally for professional analysts, now caters to amateur investors as well, offering profit insights to analysts and stock predictions to amateurs.

Problem

Over 70% of users leave after their first visit, reflecting trust and comprehension issues that impact engagement and commercial success.

Goal and Metric

Our goal is to reduce cognitive load, ensuring clarity and trust for both analysts and investors. Success will be measured by retention, engagement, and funding milestones.

design Highlight

Before

The platform prioritized creators, forcing all users to navigate the "create" section to access models, adding unnecessary complexity for investors.

Final design

The redesign separated model creation for analysts and provided investors with a simple, intuitive workflow for exploring, subscribing to, and using models.

Agile Research

Two Persona

We conducted 12 interviews to identify 2 user groups and their challenges, amateur investors (Kevin) and professional investors (Alex).

Define problems

Why over 70% of users abandoned the platform after their first visit?

Through tests, we identified three key pain points:

How might we

Create clear user flows for both user groups while building trust in the product and ensuring a clear understanding of the forecast results?

Competitive Analysis for Model Exploration & Usability

I studied competitors to uncover strategies for navigation, trust-building, and actionable insights. These findings guided design enhancements to improve user flows, establish credibility, and link data to decision-making.

Drawing from these insights, I proposed solutions:

Two rounds of tests

Solution

Solution 1:

Simplify Navigation and Clarify Product Structure for All Users

By separating the creator and investor views, I create a seamless subscription-to-use flow. This reduced information overload and streamlined user navigation, improving accessibility and clarity.

Challenge 1: Balancing Layout for Diverse Users

When designing model operations, I faced challenges in balancing information density and workflow efficiency. To address these, I adopted a Dynamic Layout solution, meeting the needs of amateur users for simplicity, advanced users for complex comparisons, and business goals.

Idea 1: Make Charts and Model Panel Collapsible

Direct suggestions is easy to access

Limited space for large charts.

Idea 2:Fold Two Charts into a Single Tab

Maximizes space for charts.

Limited advanced use.

Idea 2: Fold Two Charts into a Single Tab

Maximizes space for charts.

Limited advanced use.

Idea 3: Three-Section Dynamic Layout

Chosen

Enhances comparisons for advanced users.

Maintained clear space for results.

Challenge 2: Reduce Information Density to Enhance Space Efficiency

User testing showed users prioritized trends over detailed values. I redesigned charts by reducing detailed information and emphasizing trend shapes, improving readability and reducing visual clutter.

Final flow

Smooth Transition: From Model Exploration to Subscription

I streamlined the navigation flow from exploring to subscribing and using models, eliminating unnecessary steps. Users now have direct access to pre-assembled models and enhanced filters for easier model discovery.

Final flow

Smooth Transition: From Model Exploration to Subscription

Direct access to pre-assembled models.

Enhanced filters for model discovery.

Simplified navigation flow, reducing unnecessary steps.

Easy Model Switching and Adapt Views for Different User Needs

I enabled seamless model toggling on the interface, with collapsible panels reducing clutter for beginners and expanded panels supporting in-depth comparisons for advanced users.

Easy Model Switching and Adapt Views for Different User Needs

Toggle between models directly on the interface.

Collapsible panels reduce information overload. Expanded panels enable in-depth comparisons

Solution 2:

Build Credibility Through Clear and Trusted Data for All Users

User interviews showed that credibility is critical. I redesigned the model cards to highlight usage, accuracy, and historical performance, building user trust and confidence.

Final flow

Model Context Display

Key metrics like usage, accuracy, and adoption rates are displayed clearly in the subscription flow for better decision-making.

Solution 3:

Enhance Data Interaction and Connect Data to Action

To improve usability, I introduced features like dynamic feedback, synchronized colors, and tooltips to clarify chart relationships and provide actionable insights.

Challenge: Strengthening Chart Connections

“I didn’t get any hints that the results are changed by my input”

Users struggled to link inputs with chart results. By adding dynamic feedback and tooltips, I ensured users could easily interpret changes and understand data relationships.

Before

After

Final flow

AI summary for Amateur User

Provides direct summaries for quick understanding of key data points.

Cross-Reference, Tooltips, and Color-coding for Deeper Understanding

Dynamic visuals like color-coded charts, tooltips, and cross-references help users connect related data for deeper insights.

User feedback

User quote

The information displayed on the cards feels friendly and easily accessible. It helps me know which model I need.

Jason Chen, amateur investor

User quote

Your newly designed charts and legends are incredibly helpful—I can actually try to understand them now.

Kai Tan, stock trader

Reflection

01. Leveraging Research Data and User Insights

Research findings and user quotes are compelling tools to align teams and build consensus, transforming evidence into impactful design direction.

02. Enhancing Data Visualization for Accessibility

Clear, accessible visuals balance information density with thoughtful design—leveraging color-coding, tooltips, and inclusive practices for maximum impact.

Thanks for visiting!

Let’s build something together!

Thanks for visiting!

Let’s build something together!

Thanks for visiting!

Let’s build something together!

Thanks for visiting!

Let’s build something together!

Thanks for visiting!

Let’s build something together!