Reducing paper work and billing conflict by digitizing merchant acquisition for a FinTech company
Product Name
Dexter
Client
bKash
My Role
Lead Product Designer
Team
Intelligent Machines
Platforms
Desktop & Android
Build
Dexter App on Play Store
🗺️ Project Overview
As the Lead Product Designer for Dexter, a revolutionary partner acquisition platform for bKash, Bangladesh's largest FinTech company, I spearheaded the transition from a manual, cumbersome onboarding process to a streamlined, digital solution. Dexter, formerly known as bKash Nimontron, significantly improved efficiency and operational transparency in onboarding agents and merchants.
📄 Problem Statement
bKash faced substantial hurdles with their manual partner onboarding process, which was fraught with inefficiencies, redundant efforts, and a lack of real-time oversight. The challenge was to digitize and optimize this process, enhancing the experience for all stakeholders involved.
🧑💻 Understanding The Users
The design process began with a deep dive into the users' world. I focused on three primary user groups:
Field Force: Young, tech-savvy individuals, fluent in tech and varying in English proficiency.
Distribution House Agents: With a broad educational range, these users had varying tech and English language skills.
bKash Partner Acquisition Management: Senior executives who demanded a sophisticated, data-driven tool for real-time insights.
Mamunur Rashid
Merchant Development Officer (Field Force)
Age
25
Gender
Male
Education Level
Bachelor’s Degree
Tech Literacy
Good
Occupation
MDO for bKash
Goals
Efficiently manage daily outlet visits
Reduce paperwork and scheduling conflicts
Improve accuracy in reporting visits
Frustrations
Overwhelming paperwork
Accidental double visits to outlets
Challenges in scheduling and outlet location
Nabila Haque
Distribution House Agent
Age
32
Gender
Female
Education Level
Master's Degree
Tech Literacy
Excellent
Occupation
DH Agent
Goals
Accurately track Field Force activities
Efficiently manage performance metrics
Identify and resolve fraudulent calls
Frustrations
Difficulty in tracking Field Force
Manual data entry errors
Billing disputes with clients
Nazmul Karim
Head of Merchant Acquisition, bKash
Age
45
Gender
Male
Education Level
MBA
Tech Literacy
Excellent
Occupation
Head of MA
Goals
Obtain real-time insights into the onboarding process
Streamline the partner acquisition procedure
Enhance decision-making with accurate data
Frustrations
Lengthy onboarding process
Lack of real-time process tracking
Inefficiencies in data management and reporting
Note: User Persona was recreated with the help of Figma, Framer, and DALL-E to update this case study.
🤝 Collaboration
Collaboration was key. I worked closely with product managers, engineers, and field operatives, leading the design vision while ensuring all stakeholder needs were met. This project underscores my ability to navigate complex team dynamics and drive a project from concept to launch.
🔬 Research
My approach was research-intensive. Insights from previous projects (Biponon) for the same client and a similar user groups with the same demography informed our user understanding. Through interviews with over 25 stakeholders, I identified core pain points, laying the foundation for a user-centric design solution.
User Interview & Pain Point Analysis
Field Force
20 People
Distribution House Agents
2 People
bKash Executives
3
Field Force Pain Points
There were too many papers to keep track of
Confusing one schedule/appointment with another
Sometimes they end up scheduling more appointments than they can handle on the same date
Sometimes it is difficult to find the location of an outlet
Sometimes they need to repeat calls due to lost or missing documents
DH Agents Pain Points
Difficult to keep track of the field force they have
Difficult to measure the performance of their field force
No way to identify fraudulent calls (an outlet visit submission is considered a Call)
Manually inputting data in an excel sheet takes a lot of time and sometimes it causes errors
The above issues create conflicts during billing the client (bKash)
bKash Executive Pain Points
No real-time information available of the process
Difficult to measure the performance of the field force, as well as the distribution houses
Very lengthy process onboarding process
Manually entering data in excel sheets, consuming a lot of time and causing errors
Preparing reports by inspecting multiple documents
Billing disputes
📐 Design Overview
After the initial discovery phase, I created a user flow for each user group and explored different ideas when creating the wireframes. I showed the wireframes to other stakeholders, gathered their feedback, discussed them, and started working on the UI. During the exploration and discussion, tried to address all of the issues we identified during the pain point analysis.
✅ Addressing The Pain Points: Field Force
Visit New Outlet – Visit an outlet that is not onboarded yet
Visit a Merchant - Already onboarded merchant, a routine visit
Appointments - Scheduled appointments for revisiting an outlet to pursue them into becoming a merchant
Saved Offline - Visits (calls) saved offline due to unavailability of a stable internet connection
Image: Home, Schedule Appointment, View Today's Appointments
Appointment flow is very similar to visiting a new outlet. The MDO goes to the outlet they previously visited but wasn’t able to close the lead. They choose outlets from the appointment list, visit the outlet, and then try to collect the documents to close the lead. If there is still something missing, they will reschedule for another date.
Image: A portion of the Android App, in Bangla Language
✅ Addressing The Pain Points: DH Agents & bKash Executives
Dexter dashboard is the analytics platform that is used by the bKash Executives and the Distribution Houses. The Distribution Houses have limited access to the dashboard data.
Image: Initial Version of The Dashboard for the bKash Executives. Designed in 2020.
The Dashboard provides access to important data and analytics, as well as filtered reports that can be downloaded to share with other stakeholders. Some of the key features of the dashboard are -
MDO stats - Number of registered MDOs, assigned for the day, number of active MDOs, etc
MDO Performance
Call stats with a breakdown
Image: Field Force Stats & the Geographic Distribution of Merchants
Image: Key statistics for the bKash Executives & DH Agents
🧪 User Acceptance Test
One of the key challenges of this project was to launch the app within a very short time. We didn’t have time to test any prototypes before production. So we built the app simultaneously, conducted a UAT after the app was built.
We learned that there would be roughly 80 users during launch and will increase up to 300 in a few phases
We started with 16 users (20%) for the UAT.
User list for UAT was provided by bKash, based on their geolocations
Image: First version of the Dexter Android App, during the UAT.
Key Findings From The UAT
The key findings from the UAT were -
The app was performing well without major bugs
4 users were frustrated with a long-form in the first step when visiting a new outlet
10 users complained about losing the information they entered in the form by accidentally tapping the back button
Most of the time Check-In (Geolocation) was taking a lot of time
Other than the issues above, all 16 users found the app easy to use
Addressing The Issues
After analyzing the findings, we tried to address the issues one by one.
📝 Issue: Too long form
The Spot Codes were mapped to District and Upazila (Administrative region under District)
We moved the Spot Code field above District and Upazila
When users selected a Spot Code, the District and Upazila fields were automatically populated
This way, users needed to fill 5 fields instead of 7. That’s 28.5% less input required to fill out the form.
Moving this dropdown field improved efficiency by 28.5% from the initial design and users had to fill out less fields manually, and Introducing a confirmation screen prevented accidental loss of user inputs
📝 Issue: Accidental Data Loss
We introduced a Confirmation Alert to prevent accidental data loss
Whenever the user tries to go back, it will tell the user the consequence of going back
📝 Issue: Check-In Takes Too Long
After further investigation, we found out -
The issue occurs on mostly Xiaomi and Huawei devices, especially in battery saving mode
It was happening due to the aggressive battery saving tendencies of these devices
This was an engineering challenge and unfortunately, we could not find a quick workaround of this issue.
📊 Impact
Dexter's launch led to marked improvements in several operational areas:
Accelerated the Merchant Acquisition and Onboarding process, bKash gained 15% growth in Merchant Acquisition
Reduced paperwork to only essential documents.
Eliminated redundant visits by Field Force to the same outlets, saving time and making them more efficient
Resolved billing conflicts between the Distribution Houses and bKash by providing concrete, real-time data.
Enabled the bKash management team to gain near real-time insights into the entire operation.
🆕 Dexter Has Evolved!
Dexter has evolved in the last 3 years! We have solved many issues, addressed new requirements, and developed a consistent design language across the platform.
Image: The latest version of Dexter Dashboard. Last updated August, 2023.
Image: The latest version of Dexter Android App. Last updated June, 2023.
