Ask Aetna Alexa Skill

Role: Lead conversational designer

Project: Create an Alexa skill to answer our customers' questions about their insurance plan.

Conversational UX is the way a user experiences an interaction with a business or a service through an automated conversational interface that learns along the way.
There are a variety of interaction models that designers follow when creating a Voice User Interface (VUI)
There is a useful framework for VUI called the Natural Conversation Framework (NCF) This model attempts to emulate actual human conversational behavior.
Our goal on the Innovation team was to work with the Product team to create Aetna's first Alexa skill. This was all new to Aetna UX and Engineering.

The team decided on creating an Alexa skill that would not require a user to sign-in with a user name and password so that we could avoid the issues that center on the use of personal data and security.
Our goals focused upon providing answers to our customers' most common insurance questions. At the same time, we hoped that this would lessen calls made to our phone representatives. Below you can see a list of objectives.

We consulted our Aetna personas.
UX and Engineering downloaded Alexa skills from our competitors and studied them to understand best practices and how we could improve on their designs.
We consulted Alexa's resources, and educated ourselves on best practices in building and designing skills.
In addition, we had to learn how the Natural Language Framework operated. All of this was new to our team.
Finally, in order to generate topics, we researched hundreds of caller recordings provided by Member Services.

In order to work with content experts and product teams, we had to sample transcripts such as the one below.
We could then take these transcripts to sessions where we could edit them and create new ones based on the requirements.
Questions were gathered from phone representative call logs. We chose the most popular questions first, moving on to less critical later.
In addition, we also had to design the text for the Alexa Echo Show device at the same time.

Below you can see the format we used for every question.
Some more complex topics had more information that the customer could access.
In addition, our team had to provide the multiple ways in which this question could be asked as well as synonyms for all terms, such as "copay"

Below is an example of a more complex dialog.
This one is focused on understanding an Explanation of Benefits (EOB) document that the customer received in the mail.

The entire question and answer flow was diagramed in a similar way as the one below.
This example is from another Alexa skill for API finder created by our team as Ask Aetna flows are proprietary and could not be displayed here.

As our team completed dialogs and flows, they were handed off to Development.
Once development was complete, each section was tested by QA, UX, Content, and the Product team.
Negative testing was also completed to hear how Alexa handled unexpected input.
Once every utterance had been tested, the project was complete and put into use.