Chib-bot is a new way for users to send money through WhatsApp
Chib-bot is a WhatsApp based bot designed to conduct money transfers within Chib, an online banking service. All transfers are conducted securely through text.
Metrics
As a result of my design iterations and implementations, we noticed the following performance increases:
43%
in user activity
39%
in overall transfers
32%
in register invites
73%
in ticket optimization
Problem
Too many transfers for one team to handle
Chib is a money transferring platform that assists users in sending remittance to family via our web, app, and messaging portals. As Chib's user base began to grow, our communications team had increasing difficulties being able to attend to every user's need via messaging.
It was clear that it was time for us to expand Chib to include a new messaging system that could attend to common user requests and queries.
"How can we simplify money transfers within WhatsApp to alleviate our communication team's workload?"
Solution
A Chat-bot for money transference needs
I designed for Phase I and II of Chib's DTC Chat systems, using research to explore and address common user frustrations, and using API bot integration through WhatsApp to design an internal messaging system.
Chib-bot as designed in Whatsapp has the ability to
Conduct monetary transfers
Answer customer queries
Provide transference history
Deliver notifications
View your balance, make transfers, and invite friends to register
All without leaving WhatsApp.
Main Goals
To understand user perceptions and motivations for utilizing customer support by evaluating user and business needs
User needs
Discover the main reasons why users choose
to use our over-the-phone assistance line.
Scale down our findings to only the
user essentials for our MVP.
Understand the typical user walkthrough
for conducting money transfers with Chib
Business needs
Differentiate between issues that are
best resolved via bot and which are best suited to the communications team.
Keep our bot's branding consistent with
the language, design, and functionality of Chib.
Discover limitations and advantages of
designing within WhatsApp's API
What is the ideal interaction model that would minimize user's struggles and prevent an overwhelmed communications team?
Hypotheses
Key questions for our research
What are the most prevalent reasons that users are utilizing our over-the-phone assistance line?
Which issues are best resolved via bot and which are more suited to the communications team?
How can we resolve these struggles given the limitations and advantages of designing within WhatsApp's API?
White paper research
I wrote a report detailing information on WhatsApp's API in relation to functionalities that we need to know before designing
01
Learning how to prevent common WA based errors
02
Knowledge of API constraints,
ex. bot communication is limited to action/trigger words and numbers
03
Understanding of common conversational flows
Changing direction
A Pivot
User Interviews ā Ticket Synthesis
User Interviews were initially selected as the best method for determining user motivations, where we could speak with users and our communications team directly.
However, we eventually decided on conducting analysis using support tickets, or records from the communications team describing incidents reported to them from users.
Why did we choose this method?
Synthesizing ticket reports from the communications team was time effective, at no excess cost, and most importantly, fallowed us to view a history of documented user problems (rather than risking error by asking users to recall them by memory).
Primary research
Support Ticket Synthesis (n = 808)
1. Gathering and organizing all our ticket data in one place
In total, we pooled reports together to find that 1,582 independent users were listed as having used our WhatsApp service
2. Omitting irrelevant data
Such as tickets that got mixed with that of outside teams, (engineering team, graphic design team, etc.), accidental repeats of tickets, QA test users, and reports with key missing information (n = 808).
3. Converting data into descriptive statistics
Data sets were then converted into nominal data, with consideration taken to calculate(1) Frequency distribution and (2) Measure of central tendency (mode).
4. Data Visualization
Providing graphics depicting our findings
Key research insights
Users mainly utilized Chib's WhatsApp communications team for the following reasons
Agent Assistance
To speak directly to agents over account issues because they were already familiar with WhatsApp (35.5%)
Transference
Ability to conduct processes relating to transfers via WhatsApp was done because it was easy to share with friends & family (48%)
Balance Overview
To view their balance because it did not require having to log in using the web or app (20.9%)
Digital architecture
User Flowchart
Mapping out our conversational design into something more tangible and cohesive for engineers.
Tracking output dialogue was the most important part of documentation, with inputs limited to specific numbers and variations of key words.
Focus was placed on three main inputs for our user to choose from that divulge into different actions.
Prototyping
Flows and UI Visuals
Mock up samples were designed in order to present to our clients what our future rollout would look like.
Prototypes were categorized as either part of the general flow or as error screens, using WhatsApp's system design for the UI design.
One of the most important aspects of this rollout was the organization of Figma prototypes to include descriptors, states, experiences, flows, and engineering notes. This allowed engineers assigned to other projects to jump in with minimal context if necessary.
System design
UI Abiding guidelines
Because this was an API heavy project, visual design was predetermined by WhatsApp's existing system design. Care was taken to rebuild the WhatsApp library in Figma to include guidelines for margins, typography, colors, spacing, layouts, and components.
Abiding and localizing existing UI in Figma was essential for not only creating prototypes, but also campaign designs.
How it works
Simplifying money transference with Chib-bot
Balance
Users can view their balance, taking into account different currencies and best exchange rates
Transfer
Make an instant transfer to people within your circle, all without leaving WhatsApp
Customer Service
Contact an agent directly to receive assistance for any qualms or queries
Recommend
Recommending Chib to friends is easy - just send our bot their way to help them register
Bonus
Considerations for a final handoff
Design Team
Before and after autolayout
Organizing & updating all designs and templates
Research Team
Sample of final findings
Summary of report & data sorting for future researchers to be able to replicate and analyze our study
Takeaways
Impact
43%
in user activity
39%
in overall transfers
32%
in register invites
73%
in ticket optimization
Personal
Zerin believes in keeping tech design accessible and transparent, so if you have any questions regarding any designs, just send an email and I'll be sure to get back to you!
Want to see more work?
Check out my other case studies and projects
Back to Portfolio