Gokul's Portfolio

Designing debt management eco-system

Designing debt management
eco-system

Role

Lead Designer

Design Team Size

Functions Involved

Design, Tech, Product, CEO

Background

Navi's lending business has grown rapidly, but this expansion has been accompanied by an increase in loan defaulters.

An early portal was developed to support collection agents, but it lacked efficiency, crucial information, and user-friendly features. This forced agents to rely heavily on manual processes to recover payments, significantly increasing their workload and reducing overall collection efficiency.

My Role

As the design manager in the stream, I had to closely work with CEO (Sachin Bansal - Co-Founder of Flipkart), leading the team (of 2 designers) to do research, come up with strategical solutions to tackle the pain points and convince the product and engineering leaders and CEO himself to get the solutions implemented.

Who are the users

The solutions are catered for 'Collection Agents' (Debt management executives), segmented into In-house Tele & Field and External Field agents.

Defining Success metrics

Increase collections efficiency

Collection efficiency was at 98.2, the primary target was to improve the collection efficiency to 99.5 or above

Increase agent productivity

Agents were spending significant time in doing manual work, automate workflows and making the tool information rich

Existing problems

Insights from User Research + UX Audit

Difficult to navigate:

Users struggle to find key features and perform essential tasks due to unclear menus, poor information hierarchy, or unintuitive workflows.

Difficult to consume information:

Important data and insights are either cluttered, overly complex, or not well-structured, making it hard for users to quickly grasp key takeaways.

Not information rich:

The portal lacks sufficient data or contextual details that users need to make informed decisions, leading to inefficiencies and reliance on external tools.

Scroll leading to loss of information below the fold:

Critical content gets hidden as users scroll, causing frustration and requiring extra effort to locate important details.

Agents spending time on excel sheets to track performance and calculate incentives:

Due to the lack of built-in performance tracking, agents resort to manually maintaining spreadsheets, leading to inefficiencies and potential errors

Agents losing trust in system due to no visibility over incentives and their calculation:

Without a transparent breakdown of their earnings and incentives, agents feel uncertain about fairness, leading to dissatisfaction and disengagement.

TLs manually tracking agent performance, availability on sheets:

Team leaders have to rely on manual data entry in spreadsheets to monitor agent performance, which is time-consuming and prone to inaccuracies.

No application for field agents, had to rely upon printed sheets

Field agents lack a digital solution and must depend on printed documents, which are cumbersome, outdated quickly, and limit real-time updates.

How might we improve the existing system

By adopting new strategies focused on agent behaviour, we aim to empower agent with all the required tools and data visibility. Which is expected to improve agent productivity & efficiency, leading to more collection efficiency. Also depreciating old ways of working over time.

Strategies should be defined for 3 buckets

DPD 0 to DPD 10

DPD 10 to DPD 30

DPD 30 to DPD 90

DPD = Days passed due

Build reminder calling platform for easy customers

Call duration should we within 1 minute to cover more customers

Single screen platform

DPD 0 to DPD 10

Build agent productivity tools

Making platform information rich

Optimising horizontal scrolls across

Build performance indicators and dashboards

DPD 10 to DPD 30

Build field agent app

Making app information rich

Build performance indicators and dashboards

Easy navigation

DPD 30 to DPD 90

Biggest challenge for me as design manager?

As a design manager, one of the major challenges was to simultaneously ideate and drive solutions across three distinct product areas, each with its own set of user needs, technical constraints, and business objectives. Balancing these parallel workstreams required a structured approach to problem-solving, ensuring that each product received adequate focus while maintaining alignment with the broader design strategy.

Key difficulties included:

Context Switching: Shifting between different product domains while maintaining deep understanding and strategic oversight for each.
Resource Allocation: Distributing design efforts effectively across teams, ensuring efficiency without compromising quality.
Cohesive Experience: Ensuring consistency across all three products while addressing their unique challenges, user journeys, and technical landscapes.

Key difficulties included:

Context Switching: Shifting between different product domains while maintaining deep understanding and strategic oversight for each.
Resource Allocation: Distributing design efforts effectively across teams, ensuring efficiency without compromising quality.
Cohesive Experience: Ensuring consistency across all three products while addressing their unique challenges, user journeys, and technical landscapes.

This required strong collaboration with cross-functional teams, prioritizing initiatives based on impact, and continuously refining design processes to drive efficiency and innovation.

DPD 0 to DPD10

Human remainder calling

Using this portal, the customers who fall into bucket DPD form 0 to DPD 10 will be reminded to pay the EMI amount, continuous calls will be automatically getting connected to the agents as well as customers. Separate team of 20 agents started using this portal.

Considerations made while designing:

Least minimal data required to remind the customer

Easy to skim through - Grasp customer information within 5secs of call getting connected

Quick filling of feedback for the call

Key Screens

Call connection

Call connected - Customer details

Feedback submission alert after call disconnected

Result & Impact

97% of the outstanding EMI was collected just using reminder calling done by this platform which reduced significant load of cases from normal tele agents and allowed them to focus on hard cases.

DPD 10 to DPD30

Tele collections

Using this portal, the customers who fall into DPD bucket form 10 to DPD 30 will be called by agents, leveraging the data provided in the portal to convince the customers to pay the EMI outstanding as soon as possible. Around 400 agents will be using the platform.

Considerations made while designing:

Information rich

Easy navigation

Ease of consumption of data

Minimise manual efforts outside the platform

Agent performance modules

Increase agent trust on the platform

Improve contact-ability

Key Screens

All cases screen

Overview

EMI Schedule

Addresses

Phone number

Documents

Interactions

Skip Tracing

Performance indicator

Festival theming for agent delight

Result & Impact

⬆️ Increase in agent productivity

⬆️ Contactability increased by ~15%

⬆️ Tele collection efficiency increased by ~200bps

DPD 30 to DPD90

Field Collections - Cosmos app

Using this app, the customers who fall into bucket DPD form 30 to DPD 90 will be , home visited by the agents to collect money from them, this will be used by ~600 agents across india

Considerations made while designing:

Data rich app

Ease of use despite of data richness

Should work in offline conditions

Accuracy of information provided

Shouldn’t drain battery

Minimal usage of graphic elements

Should be a very functional app

Less clicks to perform an action

Key Screens

Result & Impact

⬆️ Field contactability increased by ~40%

⬆️ field collection efficiency increased by ~70bps

Overall Impact

HRC + Tele Collections + Field Collections

99.5% (+1.3%)

Increase in Collection Efficiency from 98.2

+30%

Increase in contactability

Back