Semi-automating repayment of loans
Background
Early loan repayment was one of the customer support department's most time-consuming tasks, taking 23% of total time. Most time went to repeated conversations about which documents were needed and where to find them. The process was phone-only.
Research
Johanna shaped the flow using in-depth interviews and created a customer journey map that pinpointed trust-breaking moments and documentation confusion. The design leaned on guides written by customer service representatives and tone-of-voice patterns from their successful messages. Multiple feedback rounds with the support team refined the MVP.
Money is a sensitive topic. Customers could feel questioned about their reasons for repayment, especially when switching banks, and skepticism increased if the flow felt like interrogation or upsell. Respectful framing and transparent explanations of regulatory requirements were needed to maintain trust.
Solution
Regulatory questions could feel abrupt, increasing skepticism about why information was requested.
Anti-money laundering context appeared early with progressive disclosure, including a tap-to-reveal option for customers who wanted more detail.
Customers relied on repeated phone conversations to understand which documents were needed and where to find them.
A structured bot flow guided customers to locate the right documentation step-by-step, using language aligned to how the support team already explained the process.
After the call, customers often remained unsure about what to do next, leading to additional contact.
End-of-conversation guidance clarified next steps and what to submit, improving follow-through quality.
User interviews and feedback loops with the support team were often skipped in bot delivery.
Johanna introduced several research rounds and brought a UX writer into the bot team for the first time to strengthen tone, clarity, and regulatory framing.
Results
• 155 early loan repayment inquiries handled in the first month, exceeding expectations • 87.5% of customers reported it helped answer their questions, outperforming the support baseline • Most customers submitted documentation correctly, reducing rework for the support team
• Progressive disclosure improved trust in sensitive financial flows • Reusing language proven by customer service representatives increased comprehension • UX writing and feedback loops with the support team elevated both quality and stakeholder confidence
Client at big retail customer
@Deloitte Digital
“The team has been truly impressed by Johanna's ability to maintain a perfect balance between standing her ground, confidently expressing her opinions, and at the same time showing humility and genuinely listening to others.”
Semi-automating repayment of loans
Background
Early loan repayment was one of the customer support department's most time-consuming tasks, taking 23% of total time. Most time went to repeated conversations about which documents were needed and where to find them. The process was phone-only.
Research
Johanna shaped the flow using in-depth interviews and created a customer journey map that pinpointed trust-breaking moments and documentation confusion. The design leaned on guides written by customer service representatives and tone-of-voice patterns from their successful messages. Multiple feedback rounds with the support team refined the MVP.
Money is a sensitive topic. Customers could feel questioned about their reasons for repayment, especially when switching banks, and skepticism increased if the flow felt like interrogation or upsell. Respectful framing and transparent explanations of regulatory requirements were needed to maintain trust.
Solution
Regulatory questions could feel abrupt, increasing skepticism about why information was requested.
Anti-money laundering context appeared early with progressive disclosure, including a tap-to-reveal option for customers who wanted more detail.
Customers relied on repeated phone conversations to understand which documents were needed and where to find them.
A structured bot flow guided customers to locate the right documentation step-by-step, using language aligned to how the support team already explained the process.
After the call, customers often remained unsure about what to do next, leading to additional contact.
End-of-conversation guidance clarified next steps and what to submit, improving follow-through quality.
User interviews and feedback loops with the support team were often skipped in bot delivery.
Johanna introduced several research rounds and brought a UX writer into the bot team for the first time to strengthen tone, clarity, and regulatory framing.
Results
• 155 early loan repayment inquiries handled in the first month, exceeding expectations • 87.5% of customers reported it helped answer their questions, outperforming the support baseline • Most customers submitted documentation correctly, reducing rework for the support team
• Progressive disclosure improved trust in sensitive financial flows • Reusing language proven by customer service representatives increased comprehension • UX writing and feedback loops with the support team elevated both quality and stakeholder confidence
Client at big retail customer
@Deloitte Digital
“The team has been truly impressed by Johanna's ability to maintain a perfect balance between standing her ground, confidently expressing her opinions, and at the same time showing humility and genuinely listening to others.”