Client
Weather
Role
Product Designer
Timeline
1 year and 4 months
Responsabilities
UX Writing
Research process
Data analysis
Wireframes
Agile development
Team collaboration
Weather* is a leader in 24-hour assistance solutions. Their main clients are insurance companies, vehicle manufacturers, banks, financial institutions, and retailers in Brazil. The challenge was to create easy contact with users through Chatbot solutions for various clients.
More info
Weather needed to help its business clients save money on customer service by replacing call center companies with an easy-to-use communication and service solution that allowed users to self-serve. The goal was to establish efficient user support, ensuring responsiveness and communication free from dissatisfaction and friction.
More info
Weather already had a Chatbot and recognized the need to enhance its conversation flows to provide better service to its users. We initiated the project with structured research, such as user journey and initial conversational flows, to create new messages with improved interactions based on heuristics.
To understand Weather's client users, we conducted several interviews with stakeholders and launched a Minimum Viable Product (MVP) to collect data on how people would navigate the first version of the chatbot.
More info
All these challenges faced by users resulted in a significant increase in the utilization of other customer service channels, especially Interactive Voice Response (IVR) or telephone contact to request assistance.
"no more human because a situation of help and despair is horrible to talk to something that sometimes can be simple, and the tool doesn't understand, and another 0800 and another number doesn't always work."
Amanda Silveira
Client
"They could enable assistance to be done through this channel."
Diana Rusback
Client
"You ask for the location and don't accept it... it's difficult."
Elizabete Flores
Client
"I didn't solve my problem, and they gave me the phone number for capitals and metropolitan regions."
Marcos Alexandre
Client
The user feedback was extracted from our analyses, so they are genuine. We've translated them into English to aid understanding.
The creation of the persona helped us understand the needs and expectations of both the user and the business, serving as a strategic archetype for the majority of the surveyed users.
Resolution
It was extremely frustrating for Renato not to have his problems resolved quickly. Seeking help in a scenario of risk/fear, he felt that the chatbot was not useful.
Data Driven
Solving the identified problems was important, but the project should be directed towards error prevention through analyzed conversation patterns
Understanding
One of the premises for Weather hiring us was to find ways to make the chatbot more assertive in its responses. Therefore, it was necessary to improve the understanding of user topics
Whitelabel
We would need to resolve most of the problems that Renato had so that this solution could reach other Weather clients. Without risk, we could then reach other people
More info
With the goal of defining and aligning the problem that needed to be addressed, we formulated user need statements, which helped organize our insights about the 'problem space'. Thus, faced with many problems and solutions that Weather needed, we decided to initially build an 'example' chatbot to become a whitelabel product. As the next steps, we planned to iterate on user experience and discover new features for continuous improvement.
More info
Looking for a complete chatbot
Iteration was crucial for gathering meaningful metrics
After dedicating initial efforts to identify various issues, we prioritized solutions through research. We implemented several improvements, launching them quickly and obtaining valuable feedback.
175%
growth in service openings compared to the start of the project, around 22,000 new openings by 2022.
with the help of the ‘cascata de validação’, we were able to improve customer retention, increase engagement, and enhance the chatbot's understanding, addressing Renato's issues more effectively.
36%
of users no longer need to contact a human attendant in order to solve their problems via Chatbot, which reduced investment in call centers.
I had the opportunity to collaborate with highly skilled professionals in the areas of design, data, development, and artificial intelligence. This team was crucial in achieving excellent results of which I am immensely proud.
One of the key lessons I learned was the importance of keeping the user always in focus to genuinely understand their needs. Additionally, I used data analysis to validate our assumptions and comprehend the demands of all stakeholders involved.
This challenging project was a valuable opportunity to recognize my limitations and overcome them through an iterative approach, where failures and restructurings were essential. In the end, we achieved outstanding results by continually enhancing our solutions.