Project V2 | Swati's UX Portfolio

Re-designing for quick, valuable financial insights for users

UX Research · UX Design · Desktop · Mobile

Overview

Smartkarma is a subscription based FinTech platform for investment recommendations from highly qualified analysts. It also serves as a network which connects investors with analysts.

The project goal was to make it quick and easy for users to takeaway valuable insights.

My Role &
Duration

I was the sole UX Designer on this project and managed the process from Research & scoping to tech handover. I collaborated with my manager who was in charge of the Design System.

The project took approximately 4 months to complete.

The Challenge

Incoming Feedback:

We were getting consistent feedback that extracting quick takeaways from the platform is difficult. This feedback had remained unaddressed in the backlog.

Change in usage patterns that made us prioritise this feedback:

Users with lower technical knowledge started growing in numbers and were less motivated to read through an entire article.
More users started logging in through smaller devices and spending less time on the platform.

We realised that to retain users going forward, it is important to deliver valuable information in a simple, quick and accessible manner.

Revamped Entity Page

New Features

Stock worthiness chart

Related Entities Widget

Price Chart on Hover

Qualitative
Research

Due to time constraints, we decided to conduct secondary research. I conducted interviews with the internal sales team that engages extensively with enterprise clients & channel partners that engage with retail clients.

The main goal was to understand:

What are the most important pieces of information that users are looking for?
What hurdles are they facing?

Quantitative
Research

I collaborated with the data analysts to gain a deeper understanding of user behaviour so that I could target crucial touch-points. We looked into common user journeys and 2 prominent patterns emerged i.e. the journey of users who are "exploring" and are open to a wide variety of topics and users who are "searching" for specific content.

This helped me identify the following touch-points as important for quick takeaways:

When users are scanning a list of articles to decide which one to read (low engagement in each topic presented)
When users are reading an article (High engagement in topic)
When users are enquiring about a specific stock (Highest engagement in topic)

An Important Opportunity

A crucial realisation we had when going through the data was that:

Entity pages are the 2nd most visited page category for registered users and subscribers
The most searched category on the global search is specific stocks
However, out of the 20000+ entities, less than 20% of entities offer Smartkarma's USP i.e. insights by Smartkarma Analysts.

This means what users see on that page (news, filings, financial charts etc) are pieces of information they can easily access on Google, Yahoo Finance and other free resources.

This led us to believe that to deliver value to users, it is important for us to provide:

Valuable insight on specific stocks
An insight that is unique to Smartkarma
A scalable solution that can be offered for the 2000+ stocks in our database

Business
Considerations

Important business consideration laid out by stakeholders was that they would like to maximise value from current investments in:

Analysts (75% of revenue is allocated to them)
Hefty annual fees the company pays to access financial data on stocks

Synthesis

I synthesised all the findings into HMW questions which helped me ideate potential solutions. Below is a screenshot of the table which highlights the Insight, User Needs & HMW questions.

Ideation

Stakeholders, 2 Data Analysts, the Head of Tech and the Design team had a brainstorming session to ideate new features that address the identified issues. After the session, I mapped the ideas on an Effort/Impact graph. I color coded them based on the risks/feasibility issues we discussed. This helped streamline the ideas further.

Final Solution

Below are some principles that guided my final design:

Create assets that are easy to scan and visual in nature. Users are feeling overwhelmed with walls of text.
Focus on streamlined high-value information.
Create assets that are easily accessible/highlighted.

01

Quick Win: Make the analyst’s sentiment easy to comprehend

Before: Users were shown a green line which meant "bullish" and a red line which meant "bearish". However, this was far from obvious to the users. They had different interpretations of what the line meant or were confused about what it indicated.

After: I mocked up multiple other ways of indicating the sentiment and decided to settle for the most obvious and non-controversial option i.e. to use the words "Bullish" and "Bearish" and color code them to make them stand out while scanning a list.

Sentiment Indicator Before: Colored line

Options Considered

Final Design: Sentiment written in colored text

02

Give users quick access to financial charts

We made the primary stock's chart sticky on the right panel and for all other stocks that are mentioned in the article, hovering on the name triggers their price chart to load. This provides users valuable context while reading without breaking their flow.

03

Enable users to easily access a summary of analyst's recommendations and information about stocks mentioned

Before: Users would have to read through the entire article to grasp what the analyst was recommending about each entity. Moreover, the extent of the recommendation was limited to just the sentiment.

After: We gave analysts the option to specify not just sentiment but also the conviction level and stop loss level in case the price moves in the opposite direction due to unforeseen circumstances. We compiled this information in a widget which also mentions all the other entities and relevant price information about them.

This provides more actionable information to users.

Video showcasing the Entities Widget

Input form which enables Analysts to enter the information required for the widget.

04

Enable users to screen stock worthiness: A new feature

Before: Users did not have access to any unique information on 80% of the stocks in the database. Moreover, articles written > 3 months ago often cease to be relevant due to the fast-paced changing nature of the markets. Hence, users often end up seeing information they can access on free resources or outdated information.

After: With the help of the Data Science team, we designed a new feature which rates stocks on various criteria against a base average (i.e. country, sector or country & sector). The Data Science team developed the criteria, the calculation formula and the tooltip text, whereas I worked on the visual representation & interactions.

Options Considered

Final Smart Score Widget

05

Watchlist Centric View

Before: Users had access to a single list of articles related to all entities in their watchlist.

After: Users can conveniently access articles grouped based on the stocks they are interested in.

Mobile Experience

Below is the mobile experience for the aforementioned changes and features. Tweaks were made to the design to suit mobile width and interactions (no hover). After user testing, I made some adjustments to the tap area of the prototype.

User Testing

Smartscore: I conducted 1 round of testing with internal users and 2 rounds of testing with external users to test the comprehensibility of the graph as well as labels and tooltips, intuitiveness of the interactions and perceived value of the feature. This helped us select the most simple representation and refine some labels.

Challenges &
Learnings

Some of the challenges that I learnt from during this project were:

While it remains crucial to prioritise high-impact features, it is important to exercise extreme caution when considering high-impact solutions in matters that are highly sensitive to users, such as money and investments.
I learnt to be ok with a certain amount of risk. I was tempted to conduct early-stage validation for the Smartscore but given tight deadlines and resources in a startup, this was not feasible. The key in such scenarios is taking continuous feedback and iterating accordingly.
Even the most seemingly obvious parts of the platform can seem ambiguous to new users. In this project, the sentiment indicator was a legacy element which had never been challenged. Learning that users struggle to understand the meaning was crucial to delivering the platform's USP more quickly.

Outcomes

1. Improved Sentiment Comprehensibility
100% success in comprehending the analyst’s sentiment towards a particular stock vs. 20% in prior user test.

2. Increase session time
Increase in session time by 20% for entity pages without recent analysis.

3. Better utilisation of business investments
Information from analysts and data analytics are used to deliver more value to users.

Next Steps

Monitor metrics on engagement with the new features and conduct qualitative interviews with users to understand how they are utilising the Smartscore.
Create a feature which allows users to compare criteria between specific stocks. I believe that would make the feature more actionable.
Make improvements to the Related Entities widget which highlights entities that have additional actionable information from analysts.