ASIA THE PERSON
SECTION
OVERVIEW
THE CHALLENGE
How can we support our advisors and our customer service team (CSU) ?
Ensuring advisors have a strong understanding of products, easy access to resources, and prompt support is essential for Empire Life. However, long call times and delays in addressing routine inquiries have created challenges for both advisors and the customer support unit. The increasing volume of calls and emails makes it difficult to maintain satisfaction and deliver timely assistance to Empire Life’s advisor pool.
THE SOLUTION
Develop a chatbot to help CSU address complex inquiries more efficiently.
An analysis of thousands of emails and calls revealed that most support requests revolve around a few recurring topics. By introducing a chatbot to handle these common inquiries directly, we can free up time and resources for the CSU to focus on more complex and urgent matters. This approach not only improves operational efficiency but also enhances the overall experience for advisors and support teams.
THE CHALLENGE
Let’s CHAT about how I helped to turn this
from an idea into a reality.
I transformed thousands of call transcripts and emails into a dynamic dashboard that tracked support trends. I designed and built 500+ conversational pathways in both French and English, ensuring the chatbot was engaging and informative. Additionally, I contributed to creating impactful presentations and fostering regular communication between internal teams and advisors, helping the product evolve into a valuable support tool.
SECTION
RESULTS
140%
increase in chat usage from 2019
340%
growth in successful chats
SECTION
IMPACT
Introducing a new channel for support and self service.
-
Improved the support journey by introducing an on-demand support intermediary for advisors.
-
Revamped the chatbot window within the portal, enabling users to navigate seamlessly or receive real-time updates on their requests.
-
Empowered our Customer Support Unit (CSU) to shift focus from routine inquiries, leveraging their expertise for more complex and personalized assistance, adding a valuable human touch to interactions.
SECTION
APPROACH
How did we create
and deliver Ely?
Our objective was to create a personable, 24/7 support solution for advisors. To prioritize our MVP, we analyzed historical advisor requests to identify high-impact, high-value topics. By studying trends and behavior patterns, we defined the chatbot's capabilities, determining what it could address, how it should respond, and the appropriate moments to escalate inquiries to a human for further support.​
SECTION
RESEARCH
To understand how to support advisors, we conducted an audit of our existing chat experience and customer support inquiries.
PROBLEM DISCOVERY
In our initial phase, I worked with the product manager to collect all the data from Five9 (phone calls), Zendesk (portal chat window), and email interactions we received from advisors in 2017-2020. We created a dashboard and management system that allows us to review trends within these interactions in real time.
Through reviewing all this data, we wanted to answer the following questions.
What are the most common topics or questions that our CSU receives?
What makes our customers happy? What makes them upset?
When should we start to escalate support (switch to a human or manager)?
Why do our current customers prefer to go straight to CSU, rather than our existing chatbot?
What did advisors ask for?
30.04%
asked for forms or documents that they could use to submit to Empire Life. They were typically denoted by their form number, name, or function.
25.70%
asked how-to questions for different processes that they wanted to complete. It could range from “how do I change my address?” to “how to facilitate a redemption from one account to another?”
16.99%
asked product-related questions for getting access to product information, marketing materials, and other resources for themselves or their clients.
15.11%
wanted tech support to navigate through the site or other portals and to get assistance with issues they encountered while on the advisor site.
4.35%
asked for contact information (emails, phone numers, addresses) to specific offices, sales teams, or individual contributors within Empire Life.
SECTION
PRODUCT GOALS
Our audit guided our strategy and form three main values for
designing the new chatbot.
Become the front-line defense.
The chatbot should be able to answer most questions that advisors would ask. To do this, we need to focus on the top inquiries and filter out any topics with added complexities (variations in answers, long form content, topics that require ID verification, etc.)
Be adaptable and aware.
Apart from being able to answer questions, great chatbots are able to adapt to their audience. The chatbot should be proactive in noticing subtle cues from the advisor that can signal frustration - so that we can immediately escalate them to a live agent, before they reach a boiling point.
​Monitor and enhance as we go.
Tools that stand the test of time need to remain current. Apart from building a solution, we want to know how it performs so we should build a tool that allows us to see how our chatbot performs in detail. It should provide sentiment scores, transcripts, and feedback from users.
SECTION
/ FINAL PRODUCT
SECTION
/ KEY TAKEAWAYS
Communication is more than just talk
Through this journey, I learned about the different ways we can enhance a conversation outside of words. While building the chatbot, we had to consider numerous variables outside of content. We created new branding, additional functional features, and constructed a personality that matched Empire Life’s brand.
What made Ely notable to advisors, was not only his succinct answers - but the way he delivered it. People loved the design and his friendly look that matched his witty and amicable dialogue.
Becoming a jack of all trades
In creating a new intermediary for customer support, our team was exposed to a variety of other departments within Empire Life. It was a great learning experience into the financial industry - as we got to learn more about our products and also the operational teams and processes that support it.
The wide exposure to different teams at Empire also helped me in future projects, as I already had some contextual knowledge to help guide my design process.
Efficient data collection is key
The chatbot management system addressed an issue that impacted our team in the earlier stages - which was the lack of an efficient data collection method. We learned that previous attempts to create a chatbot were often blocked by how difficult it was to synthesize data and we wanted to avoid that pitfall in the future.
By introducing a content management system and analytics dashboard our team were better equipped to make rapid content updates based on user feedback.
