© FOLIO '24
RACH SARRA
My local time –
GMT (-3:00)
© FOLIO '24
RACH SARRA
My local time –
GMT (-3:00)
© FOLIO '24
RACH SARRA
My local time –
GMT (-3:00)
Role
Product designer
Product Manager
Made for
z1
Tags
Banking app
ios & android
chatbot
Customer support
© 2022
Challenges in customer support amidst rapid growth.
Z1 is a Gen Z-oriented fintech that provides a tailored banking experience for Brazilian teens and kids. Its mission is to offer financial education and inclusion to a demographic that often lacks access to traditional banking products. Z1 enables users to manage their finances through a digital account equipped with Pix*, a prepaid card, gift vouchers, and other tools for spending, budgeting, and saving.
In the first half of 2022, Z1's customer support team experienced a notable increase in contact volume due to the rapid growth of its user base. Without any technical automation, customer experience (CX) analysts were required to manually address common queries, resulting in service inefficiencies and bottlenecks. This lack of automation caused delays, overwhelmed the CX team, and adversely affected the user experience during peak times.
* Pix is a real-time payment system in Brazil, enabling instant, 24/7 free money transfers using simple identifiers like phone numbers or email.
Challenges in customer support amidst rapid growth.
Z1 is a Gen Z-oriented fintech that provides a tailored banking experience for Brazilian teens and kids. Its mission is to offer financial education and inclusion to a demographic that often lacks access to traditional banking products. Z1 enables users to manage their finances through a digital account equipped with Pix*, a prepaid card, gift vouchers, and other tools for spending, budgeting, and saving.
In the first half of 2022, Z1's customer support team experienced a notable increase in contact volume due to the rapid growth of its user base. Without any technical automation, customer experience (CX) analysts were required to manually address common queries, resulting in service inefficiencies and bottlenecks. This lack of automation caused delays, overwhelmed the CX team, and adversely affected the user experience during peak times.
* Pix is a real-time payment system in Brazil, enabling instant, 24/7 free money transfers using simple identifiers like phone numbers or email.
Challenges in customer support amidst rapid growth.
Z1 is a Gen Z-oriented fintech that provides a tailored banking experience for Brazilian teens and kids. Its mission is to offer financial education and inclusion to a demographic that often lacks access to traditional banking products. Z1 enables users to manage their finances through a digital account equipped with Pix*, a prepaid card, gift vouchers, and other tools for spending, budgeting, and saving.
In the first half of 2022, Z1's customer support team experienced a notable increase in contact volume due to the rapid growth of its user base. Without any technical automation, customer experience (CX) analysts were required to manually address common queries, resulting in service inefficiencies and bottlenecks. This lack of automation caused delays, overwhelmed the CX team, and adversely affected the user experience during peak times.
* Pix is a real-time payment system in Brazil, enabling instant, 24/7 free money transfers using simple identifiers like phone numbers or email.
Project overview: defining the problem and user challenges.
Z1’s absence of automated support tools led to long wait times and repetitive tasks for the CX team. Users struggled to resolve urgent issues quickly, and even with a FAQ section available in the app, the team remained burdened by simple, recurring questions that could easily be addressed through self-service solutions, as demonstrated by market competitors. This inefficiency resulted in user frustration and placed undue pressure on the support team.
Challenge
Design an automated solution that would reduce the workload of the CX team while improving the overall user experience.
User pain points
✸ Long Wait Times: Users faced delays in resolving urgent issues like security locks or password resets due to high contact volumes and slow responses.
✸ Lack of Self-Service Options: Despite having a FAQ section, users were still inclined to ask simple questions about logistics, payments, or card operation, which at the time could not be easily resolved without support, causing frustration.
✸ Inconsistent Experience: Users were frustrated with the time it took to receive help, resulting in an inconsistent support experience.
✸ Overloaded CX Team: The support team was overburdened with repetitive tasks, limiting their capacity to focus on more complex issues that required personalized assistance.
Project overview: defining the problem and user challenges.
Upon analyzing the data from Z1's CX team, we discovered that 46% of all chats were already being handled using predefined scripts, which were manually copied and pasted by the support team. Of this, 30% of inquiries were related to data changes, which required follow-up support after the initial scripted response, while 16% were purely resolved with scripted answers, indicating a clear opportunity for automation.
Benchmarking the market
A review of industry standards showed that leading companies in customer support had already implemented robust chatbots as part of their service strategy. These businesses followed the customer service funnel, which consists of three layers: prevention, self-service, and support. Z1 had solidified both the prevention and support layers, but lacked a sufficient self-service solution that reasoned with users.
Opportunities for improvement
The absence of a comprehensive self-service layer at Z1 presented a clear opportunity. By implementing chatbot technology—already widely adopted by major services—Z1 could automate responses to simple, repetitive queries. This would streamline the user experience, reduce wait times, and allow the CX team to focus on more complex cases.
Project overview: defining the problem and user challenges.
Z1’s absence of automated support tools led to long wait times and repetitive tasks for the CX team. Users struggled to resolve urgent issues quickly, and even with a FAQ section available in the app, the team remained burdened by simple, recurring questions that could easily be addressed through self-service solutions, as demonstrated by market competitors. This inefficiency resulted in user frustration and placed undue pressure on the support team.
Challenge
Design an automated solution that would reduce the workload of the CX team while improving the overall user experience.
User pain points
✸ Long Wait Times: Users faced delays in resolving urgent issues like security locks or password resets due to high contact volumes and slow responses.
✸ Lack of Self-Service Options: Despite having a FAQ section, users were still inclined to ask simple questions about logistics, payments, or card operation, which at the time could not be easily resolved without support, causing frustration.
✸ Inconsistent Experience: Users were frustrated with the time it took to receive help, resulting in an inconsistent support experience.
✸ Overloaded CX Team: The support team was overburdened with repetitive tasks, limiting their capacity to focus on more complex issues that required personalized assistance.
Identifying automation opportunities.
Upon analysing the data from Z1's CX team, we discovered that 46% of all chats were already being handled using predefined scripts, which were manually copied and pasted by the support team. Of this, 30% of inquiries were related to data changes, which required follow-up support after the initial scripted response, while 16% were purely resolved with scripted answers, indicating a clear opportunity for automation.
Benchmarking the market
A review of industry standards showed that leading companies in customer support had already implemented robust chatbots as part of their service strategy. These businesses followed the customer service funnel, which consists of three layers: prevention, self-service, and support. Z1 had solidified both the prevention and support layers, but lacked a sufficient self-service solution that reasoned with users.
Opportunities for improvement
The absence of a comprehensive self-service layer at Z1 presented a clear opportunity. By implementing chatbot technology—already widely adopted by major services—Z1 could automate responses to simple, repetitive queries. This would streamline the user experience, reduce wait times, and allow the CX team to focus on more complex cases.
Identifying automation opportunities.
Our initial solution involved a "rule-based" chatbot, also known as a "click bot." This system utilized pre-defined conversation flows, guiding users through a series of selected responses based on our existing macros.
To validate its effectiveness, the initial experiment focused on the "Top 10 FAQs" received by CX agents at Z1. This approach aimed to assess the success and usefulness of the self-service layer. Users were presented with predefined options, allowing the conversation to progress based on their selections. If none of the available macros addressed a specific case, the user was redirected to a CX agent for further assistance.
Identifying automation opportunities.
Upon analysing the data from Z1's CX team, we discovered that 46% of all chats were already being handled using predefined scripts, which were manually copied and pasted by the support team. Of this, 30% of inquiries were related to data changes, which required follow-up support after the initial scripted response, while 16% were purely resolved with scripted answers, indicating a clear opportunity for automation.
Benchmarking the market
A review of industry standards showed that leading companies in customer support had already implemented robust chatbots as part of their service strategy. These businesses followed the customer service funnel, which consists of three layers: prevention, self-service, and support. Z1 had solidified both the prevention and support layers, but lacked a sufficient self-service solution that reasoned with users.
Opportunities for improvement
The absence of a comprehensive self-service layer at Z1 presented a clear opportunity. By implementing chatbot technology—already widely adopted by major services—Z1 could automate responses to simple, repetitive queries. This would streamline the user experience, reduce wait times, and allow the CX team to focus on more complex cases.
Introducing Zelfy: a rule-based chatbot solution.
Our initial solution involved a “rule-based” chatbot, also known as a “click bot.” This system utilized pre-defined conversation flows, guiding users through selected responses based on our existing macros.
To validate its effectiveness, the initial experiment focused on the “Top 10 FAQs” received by CX agents at Z1. This approach aimed to assess the success and usefulness of the self-service layer. Users were presented with predefined options, allowing the conversation to progress based on their selections. If none of the available macros addressed a specific case, the user was redirected to a CX agent for further assistance.
Introducing Zelfy: a rule-based chatbot solution.
Our initial solution involved a “rule-based” chatbot, also known as a “click bot.” This system utilized pre-defined conversation flows, guiding users through selected responses based on our existing macros.
To validate its effectiveness, the initial experiment focused on the “Top 10 FAQs” received by CX agents at Z1. This approach aimed to assess the success and usefulness of the self-service layer. Users were presented with predefined options, allowing the conversation to progress based on their selections. If none of the available macros addressed a specific case, the user was redirected to a CX agent for further assistance.
Introducing Zelfy: a rule-based chatbot solution.
Our initial solution involved a “rule-based” chatbot, also known as a “click bot.” This system utilized pre-defined conversation flows, guiding users through selected responses based on our existing macros.
To validate its effectiveness, the initial experiment focused on the “Top 10 FAQs” received by CX agents at Z1. This approach aimed to assess the success and usefulness of the self-service layer. Users were presented with predefined options, allowing the conversation to progress based on their selections. If none of the available macros addressed a specific case, the user was redirected to a CX agent for further assistance.
Evaluating Zelfy's impact.
After two months of making Zelfy available to 100% of the user base, the data showed promising signs of improvement:
✸ Retention Rate: Zelfy achieved an average retention rate of 24%, indicating that a significant portion of users engaged with the chatbot throughout their support journey without needing further assistance.
✸ Resolution Rate: The chatbot had an average resolution rate of 4.85%, with some weeks reaching nearly 10%. Since the issues users faced varied from week to week and Zelfy initially only addressed the top 10 frequently asked questions, it is understandable that the resolution rate was not very high at the start. These results reflect the early stages of automation and provide a foundation for future development.
✸ Customer Satisfaction (CSAT): A CSAT survey conducted regarding the chatbot's performance yielded an average score of 80.16%, showing that the majority of users were satisfied with the support provided by Zelfy.
This experiment served as a proof of concept for Zelfy. The chatbot continued to be expanded with more topics to tackle a wider range of issues. Recognizing Zelfy's potential, Z1 later invested in a generative chatbot model to further enhance support for both users and analysts, taking customer service to the next level.
Evaluating Zelfy's impact.
After two months of making Zelfy available to 100% of the user base, the data showed promising signs of improvement:
✸ Retention Rate: Zelfy achieved an average retention rate of 24%, indicating that a significant portion of users engaged with the chatbot throughout their support journey without needing further assistance.
✸ Resolution Rate: The chatbot had an average resolution rate of 4.85%, with some weeks reaching nearly 10%. Since the issues users faced varied from week to week and Zelfy initially only addressed the top 10 frequently asked questions, it is understandable that the resolution rate was not very high at the start. These results reflect the early stages of automation and provide a foundation for future development.
✸ Customer Satisfaction (CSAT): A CSAT survey conducted regarding the chatbot's performance yielded an average score of 80.16%, showing that the majority of users were satisfied with the support provided by Zelfy.
This experiment served as a proof of concept for Zelfy. The chatbot continued to be expanded with more topics to tackle a wider range of issues. Recognizing Zelfy's potential, Z1 later invested in a generative chatbot model to further enhance support for both users and analysts, taking customer service to the next level.
Evaluating Zelfy's impact.
After two months of making Zelfy available to 100% of the user base, the data showed promising signs of improvement:
✸ Retention Rate: Zelfy achieved an average retention rate of 24%, indicating that a significant portion of users engaged with the chatbot throughout their support journey without needing further assistance.
✸ Resolution Rate: The chatbot had an average resolution rate of 4.85%, with some weeks reaching nearly 10%. Since the issues users faced varied from week to week and Zelfy initially only addressed the top 10 frequently asked questions, it is understandable that the resolution rate was not very high at the start. These results reflect the early stages of automation and provide a foundation for future development.
✸ Customer Satisfaction (CSAT): A CSAT survey conducted regarding the chatbot's performance yielded an average score of 80.16%, showing that the majority of users were satisfied with the support provided by Zelfy.
This experiment served as a proof of concept for Zelfy. The chatbot continued to be expanded with more topics to tackle a wider range of issues. Recognizing Zelfy's potential, Z1 later invested in a generative chatbot model to further enhance support for both users and analysts, taking customer service to the next level.