Data Driven
UXUI
QA
Wallet Trading Conversion Rate Improvement
16 Weeks
Member
3 Developers
1 Designers
5 Stakeholders
Tools
Figma, Redash
Responsibility
Lead Product Designer
UIUX Design
Prototyping
User Story Planning
QA
Time frame
16 weeks
Summary
I led the integration of Swap and Exchange trading features for a cryptocurrency wallet, serving as both lead designer and product manager throughout the entire development cycle.
Data analysis revealed app wallet users had 25% lower conversion rates and a 92% drop-off rate due to complex 6-step flows.
Through strategic UX optimization, I delivered a 21% increase in swap conversions, 9% increase in exchange conversions, and dramatically reduced initial drop-off from 92% to 9%.
Background
Background
Background
Unifying Swap and Exchange
for Conversion Rate
Unifying Swap and Exchange for Conversion Rate
Unifying Swap and Exchange for Conversion Rate
Our service offered two separate trading features—Swap and Exchange—based on the same functionality but different transaction logic.
Our service offered two separate trading features—Swap and Exchange—based on the same functionality but different transaction logic.
Our service offered two separate trading features—Swap and Exchange—based on the same functionality but different transaction logic.
Both features were essentially the same in terms of "trading," but they had different advantages/disadvantages as shown below and covered different trading pairs.
Both features were essentially the same in terms of "trading," but they had different advantages/disadvantages as shown below and covered different trading pairs.
Both features were essentially the same in terms of "trading," but they had different advantages/disadvantages as shown below and covered different trading pairs.
To solve this problem, the team decided to integrate the two features, and I joined the project after this decision was made.
To solve this problem, the team decided to integrate the two features, and I joined the project after this decision was made.
To solve this problem, the team decided to integrate the two features, and I joined the project after this decision was made.
Swap
Swap
Swap
Decentralized Provider
Decentralized Provider
Decentralized Provider
Broad Coverage
Broad Coverage
Broad Coverage
Low Minimum Amount
Low Minimum Amount
Low Minimum Amount
Exchange
Exchange
Exchange
Centralized Provider
Centralized Provider
Centralized Provider
Limited Coverage
Limited Coverage
Limited Coverage
Higher Minimum Amount
Higher Minimum Amount
Higher Minimum Amount
Data Analysis
Data Analysis
Data Analysis
I first analyzed existing data to understand the current service landscape
I first analyzed existing data to understand the current service landscape
I first analyzed existing data to understand the current service landscape
High App Wallet Usage Rate
High App Wallet Usage Rate
High App Wallet Usage Rate
Despite internal expectations that most users would be hardware wallet users, app and hardware wallet usage turned out to be nearly equal.
Despite internal expectations that most users would be hardware wallet users, app and hardware wallet usage turned out to be nearly equal.
Despite internal expectations that most users would be hardware wallet users, app and hardware wallet usage turned out to be nearly equal.
App Wallet User Ratio in Swap
App Wallet User Ratio in Swap
App Wallet User Ratio in Swap
58%
58%
58%
Low Swap Conversion Rate Among App Wallet Users
Low Swap Conversion Rate Among App Wallet Users
Low Swap Conversion Rate Among App Wallet Users
However, app wallet users' swap conversion rate was only 72% of that of hardware wallet users.
However, app wallet users' swap conversion rate was only 72% of that of hardware wallet users.
However, app wallet users' swap conversion rate was only 72% of that of hardware wallet users.
App Wallet CVR Compared to Biometric Wallets
App CVR Compared to Biometric
App CVR Compared to Biometric
72%
72%
72%
Approach
Approach
Approach
Let's pinpoint the differences and check the dropout rate
Let's pinpoint the differences and check the dropout rate
Let's pinpoint the differences and check the dropout rate
+3
I deposited assets directly into our app to experience the trading flow and identify issues. The results showed that because our app is optimized for Biometric Wallet, the user experience flows significantly differently from other wallet apps I had used before. Based on these findings, I sought to analyze specific differences from competitors and verify whether these variations caused meaningful user drop-offs.
I deposited assets directly into our app to experience the trading flow and identify issues. The results showed that because our app is optimized for Biometric Wallet, the user experience flows significantly differently from other wallet apps I had used before. Based on these findings, I sought to analyze specific differences from competitors and verify whether these variations caused meaningful user drop-offs.
I deposited assets directly into our app to experience the trading flow and identify issues. The results showed that because our app is optimized for Biometric Wallet, the user experience flows significantly differently from other wallet apps I had used before. Based on these findings, I sought to analyze specific differences from competitors and verify whether these variations caused meaningful user drop-offs.
Competitor Analysis
Competitor Analysis
Competitor Analysis
3 Problems
3 Problems
3 Problems
The comparison revealed that three major drop-offs occurred in areas where our user flow was distinctly different from competitors.
The comparison revealed that three major drop-offs occurred in areas where our user flow was distinctly different from competitors.
The comparison revealed that three major drop-offs occurred in areas where our user flow was distinctly different from competitors.
Swap
Swap
Swap
92% drop-off during 6-step asset selection process
92% drop-off during 6-step asset selection process
92% drop-off during 6-step asset selection process
Unlike typical wallet apps that create only one account per coin, our biometric wallet allows multiple accounts for the same coin. Consequently, users must select not only coin, network information, but also address when trading, requiring 6 steps to choose the correct trading account—compared to just 2-4 steps for other companies.
Unlike typical wallet apps that create only one account per coin, our biometric wallet allows multiple accounts for the same coin. Consequently, users must select not only coin, network information, but also address when trading, requiring 6 steps to choose the correct trading account—compared to just 2-4 steps for other companies.
Unlike typical wallet apps that create only one account per coin, our biometric wallet allows multiple accounts for the same coin. Consequently, users must select not only coin, network information, but also address when trading, requiring 6 steps to choose the correct trading account—compared to just 2-4 steps for other companies.
Competitor
Competitor
Competitor
D'CENT
D'CENT
D'CENT
Exchange
Exchange
Exchange
17% drop-off during provider option selection
17% drop-off during provider option selection
17% drop-off during provider option selection
Generally, other platforms either provide fixed options or display pre-set optimal options, while our platform lists all options and emphasizes allowing users to set their preferred ones.
Generally, other platforms either provide fixed options or display pre-set optimal options, while our platform lists all options and emphasizes allowing users to set their preferred ones.
Generally, other platforms either provide fixed options or display pre-set optimal options, while our platform lists all options and emphasizes allowing users to set their preferred ones.
Competitor
Competitor
Competitor
D'CENT
D'CENT
D'CENT
Exchange
Exchange
Exchange
33% drop-off when entering asset transfer flow after transaction confirmation
33% drop-off when entering asset transfer flow after transaction confirmation
33% drop-off when entering asset transfer flow after transaction confirmation
Unlike typical flows where users proceed directly to signing after transaction confirmation, our app redirects to an "asset transfer" flow, requiring another "transfer transaction confirmation." Testing the service myself revealed that this flow feels unnecessary and could be mistaken for an error when users encounter the "transfer" flow during swaps.
Unlike typical flows where users proceed directly to signing after transaction confirmation, our app redirects to an "asset transfer" flow, requiring another "transfer transaction confirmation." Testing the service myself revealed that this flow feels unnecessary and could be mistaken for an error when users encounter the "transfer" flow during swaps.
Unlike typical flows where users proceed directly to signing after transaction confirmation, our app redirects to an "asset transfer" flow, requiring another "transfer transaction confirmation." Testing the service myself revealed that this flow feels unnecessary and could be mistaken for an error when users encounter the "transfer" flow during swaps.
Competitor
Competitor
Competitor
D'CENT
D'CENT
D'CENT
Solution Seeking
Solution Seeking
Solution Seeking
3 Solutions
3 Solutions
3 Solutions
Swap
Swap
Swap
Allow users to select accounts at once through account nicknames
Allow users to select accounts at once through account nicknames
Allow users to select accounts at once through account nicknames
Nickname and address, account 1:1 matching
Nickname and address, account 1:1 matching
Nickname and address, account 1:1 matching
Based on database analysis, the unique information each account has was "address" and "nickname."
Based on database analysis, the unique information each account has was "address" and "nickname."
Based on database analysis, the unique information each account has was "address" and "nickname."
Since addresses are random strings that users cannot realistically memorize, we decided to leverage the existing nickname system.
Since addresses are random strings that users cannot realistically memorize, we decided to leverage the existing nickname system.
Since addresses are random strings that users cannot realistically memorize, we decided to leverage the existing nickname system.
However, since "nicknames" were not actively exposed previously, we decided to verify later whether users could distinguish accounts by their nicknames.
However, since "nicknames" were not actively exposed previously, we decided to verify later whether users could distinguish accounts by their nicknames.
However, since "nicknames" were not actively exposed previously, we decided to verify later whether users could distinguish accounts by their nicknames.
Exchnage
Exchnage
Exchnage
Show the best option and hide the rest
Show the best option and hide the rest
Show the best option and hide the rest
Provider selection rate 0%
Provider selection rate 0%
Provider selection rate 0%
Based on actual data analysis, the provider change rate was 0%. Therefore, we updated the system to set the option with the best trading rate as the default, while placing other options at a deeper level so they can only be accessed when desired.
Based on actual data analysis, the provider change rate was 0%. Therefore, we updated the system to set the option with the best trading rate as the default, while placing other options at a deeper level so they can only be accessed when desired.
Based on actual data analysis, the provider change rate was 0%. Therefore, we updated the system to set the option with the best trading rate as the default, while placing other options at a deeper level so they can only be accessed when desired.
Exchnage
Exchnage
Exchnage
If entry to Transfer flow is unavoidable, at least provide notification
If entry to Transfer flow is unavoidable, at least provide notification
If entry to Transfer flow is unavoidable, at least provide notification
Direct entry to signing after confirmation, high resource consumption expected
Direct entry to signing after confirmation, high resource consumption expected
Direct entry to signing after confirmation, high resource consumption expected
After communicating with the development team, we found that significant technical improvements were needed to allow trades to proceed immediately after swap confirmation.
After communicating with the development team, we found that significant technical improvements were needed to allow trades to proceed immediately after swap confirmation.
After communicating with the development team, we found that significant technical improvements were needed to allow trades to proceed immediately after swap confirmation.
Therefore, within this project, rather than improving the flow, we aimed to minimize friction in the experience by providing at least a cue for the next action.
Therefore, within this project, rather than improving the flow, we aimed to minimize friction in the experience by providing at least a cue for the next action.
Therefore, within this project, rather than improving the flow, we aimed to minimize friction in the experience by providing at least a cue for the next action.
Total Flow
Only 6 Steps
For SWAP
Only 6 Steps
For SWAP
Total Flow
Only 6 Steps
For SWAP
Prototyping
Prototyping
Prototyping
I communicated with the development team and shared updates with stakeholders whenever the direction was finalized. After the overall picture was established, I prototyped the final flow and obtained approval via Slack.
I communicated with the development team and shared updates with stakeholders whenever the direction was finalized. After the overall picture was established, I prototyped the final flow and obtained approval via Slack.
I communicated with the development team and shared updates with stakeholders whenever the direction was finalized. After the overall picture was established, I prototyped the final flow and obtained approval via Slack.
Test
Test
Test
Internal UT for basic usability and solution 01 verification
Internal UT for basic usability and solution 01 verification
Internal UT for basic usability and solution 01 verification
We performed prototyping-based UT targeting internal team D'CENT users. The evaluation of nickname-based account identification showed that accounts could be distinguished with 85% accuracy
We performed prototyping-based UT targeting internal team D'CENT users. The evaluation of nickname-based account identification showed that accounts could be distinguished with 85% accuracy
We performed prototyping-based UT targeting internal team D'CENT users. The evaluation of nickname-based account identification showed that accounts could be distinguished with 85% accuracy
High Fidelity Design & Documentation
High Fidelity Design & Documentation
High Fidelity Design & Documentation
Design Hand Off & QA
Design Hand Off & QA
Design Hand Off & QA
After receiving confirmation, I organized the design and components, then created the PRD based on the finalized design.
After receiving confirmation, I organized the design and components, then created the PRD based on the finalized design.
After receiving confirmation, I organized the design and components, then created the PRD based on the finalized design.
Results
Results
Results
CVR increased by ~21%
Exceeding the KPI of 64.7%
CVR increased by ~21%
Exceeding the KPI of 64.7%
CVR increased by ~21%
Exceeding the KPI of 64.7%
After one month, I reviewed the project metrics and achieved the following results
After one month, I reviewed the project metrics and achieved the following results
After one month, I reviewed the project metrics and achieved the following results
Exchange CVR increase
Exchange CVR increase
21.2%
21.2%
21.2%
Exchange CVR increase
Exchange CVR increase
9.1%
9.1%
9.1%
Solution 01 Area Dropout Rate
Solution 01 Area Dropout Rate
Solution 01 Area Dropout Rate
9.4%
9.4%
Previous : 92%
Previous : 92%
Solution 02 Area Dropout Rate
Solution 02 Area Dropout Rate
Solution 02 Area Dropout Rate
0.0%
0.0%
Previous : 17%
Previous : 17%
Solution 03 Area Dropout Rate
Solution 03 Area Dropout Rate
Solution 03 Area Dropout Rate
8.3%
8.3%
Previous : 33%
Previous : 33%
Retrospective
Retrospective
Retrospective
Project Wrap-Up: Retrospective
Project Wrap-Up: Retrospective
Project Wrap-Up: Retrospective
To ensure continuous improvement, I tracked key service metrics for two weeks post-launch and facilitated a retrospective with engineers and stakeholders to review the impact and takeaways.
To ensure continuous improvement, I tracked key service metrics for two weeks post-launch and facilitated a retrospective with engineers and stakeholders to review the impact and takeaways.
To ensure continuous improvement, I tracked key service metrics for two weeks post-launch and facilitated a retrospective with engineers and stakeholders to review the impact and takeaways.
