Partner Chargeback
Dispute Flow
2024
•
iOS & Android App
Summary
I redesigned and optimized the chargeback dispute process at Tinkoff Bank, shifting it from a manual support-heavy system to an automated self-service flow. This transformation significantly reduced operational costs and improved user autonomy.
30%
reduction in support requests
2–3 minutes
Users can now dispute cashback in 2–3 minutes without waiting for an operator.
USER & BUSINESS ISSUES
— The lack of confirmation messaging increased duplicate support tickets and slowed response times.
— Business costs rose as operations handled avoidable inquiries.
— Users weren’t clear if a chargeback was submitted or approved, leading to confusion and re-submissions.
SUCCESS METRICS
— At least 30% of support requests were shifted from frontline agents to the self-service flow.
— A potential monthly savings of nearly $500k on inquiry handling.
— Significant drop in support requests
GOALS
— Reduce support load by letting users handle partners cashback disputes independently.
— Cut costs by at least $500k a month
MY CONTRIBUTION
Discovery, Prioritization of hypotheses, Hi-Fi design, Design Review, edge cases, animations and prototyping
Discovery
Context
Many users expected chargebacks from partner promotions, but refunds often failed to appear. Without clear explanations in the app, they felt lost and turned to customer support for help. For Tinkoff, this created an unnecessary workload and frustration on both sides.
The goal was to build a transparent self-service flow where users could understand the status of their chargebacks, see what went wrong, and resolve issues without contacting support.
Gathering data
I began by collecting all relevant information to understand the full picture. After speaking with the project manager, I mapped the user journey for resolving partner cashback complaints. This helped reveal friction points, decision steps, and where users dropped off or contacted support.
KEY RESEARCH QUESTIONS
How do users currently resolve chargeback issues, and how much effort does that take?
What are the most common reasons for contacting support?
When a partner promotion is mentioned, how do agents locate it in the system?
What edge cases or unclear rules cause confusion?
How are agents handling chargeback-related questions?
KEY USER PAIN POINTS
Lack of transparency — users don’t understand why the chargeback wasn’t credited.
Long resolution times — disputes take several days, causing frustration and overloading support.
KEY SUPPORT PAIN POINTS
Many unnecessary requests — users skip reading promotion terms and dispute invalid transactions.
Manual verification — agents spend time reviewing simple cases that could be auto-declined using clear criteria.
Define
Key Challenges
After analyzing the data and understanding the process from both user and support perspectives, I identified two key challenges and developed hypotheses to address them.
First Challenge
Each promotion had unique eligibility criteria, making automation difficult. The system had to accurately determine which transactions qualified for a chargeback under each offer — even minor errors could lead to financial inconsistencies and user complaints.
First hypothesis
If we automate promotion and transaction eligibility based on MCC, we can reduce support workload, minimize errors, and speed up cashback processing, which should improve user satisfaction and loyalty.
Second hypothesis
If we add relevance-based search to the promotion list, users will be able to find valid offers faster, reducing frustration and increasing the completion rate for dispute submissions.
Second Challenge
Receipts were another issue: many were submitted as low-quality photos, slowing verification and reducing accuracy. To improve this, users needed a way to manually upload receipts and fill in missing details.
Hypothesis
If we implement smart receipt and QR code recognition in the TCRM system, we expect to reduce manual input errors, accelerate verification, and streamline the overall experience — which may result in fewer support tickets and increased task completion rates.
Third Challenge
The entry point for disputing partner chargebacks within the transaction section was unclear, leaving users uncertain about where to start.
Hypothesis
If we add an additional entry point to the cashback dispute flow from the wallet section, along with an intermediate context screen, users will be able to locate the relevant transaction more easily. This could improve process clarity, reduce support load, and create a more intuitive self-service experience.
Prioritise
Together with my manager, I evaluated the hypotheses using an effort–impact matrix to identify those offering the highest value relative to implementation cost.
We consulted with developers responsible for the TCRM player to estimate the effort for implementing search functionality, smart receipt upload, and QR recognition. I also proposed adding MCC-based auto-recognition to further streamline the process.
Based on these discussions, we prioritized relevance-based search and automated receipt verification for the initial phase, as both offered the greatest short-term impact. Additional improvements will be revisited after evaluating early results.
Develop
Deliver
Impact
2–3 minutes
Users can now dispute chargeback in 2–3 minutes without
waiting for an operator.
30% reduction in support requests
saving up to $500k per month
Takeaways
What else would I like to do
Adding a direct entry point to the dispute flow from the chargeback wallet, making the process more intuitive and discoverable.
Automating the matching of promotions and transactions via MCC codes to minimize errors and reduce manual verification.
Team 🫰🏻
Product designer (me)
Project manager
Product analysts
Developers
UX writer
Rent my mind and hands