Built-in Voice Assistant

Aligning the experience of a third party with our established experience


screenshot of the voice search

Problem Space

DISH decided to partner with Google in order to improve the power and capability of ASR technology. This meant that we would have to represent both brands while keeping a familiar functionality for our users. Voice interactions and search are a complex problem space. Different users may use the same phrase for vastly different intentions. Our goal is to have a rich, intuitive voice experience for our customers, allowing them to get all the functionality of a voice assistant, without the need for more hardware in their home.


image of a user testing session

Discovery

In our discovery phase of this project, we tested our existing voice experience as well as  different voice assistant products on the market. Our data showed our users prefer to use voice to go to a particular channel (i.e. Go to ESPN). Our qualitative data showed this was due to the small amount of cognitive effort along with reliable functionality. In our analysis of competitor products along with a diary study of new users, we found, similarly, users would try a wide range of commands, but over time would only use a small portion of the assistants’ features that were reliable and intuitive.


screenshot of the early voice search

Build and Measure

We paid particular attention to the search functionality in our upgraded voice experience. We found that users desired this feature but were also dissatisfied with how and why certain results were displayed. Our voice UI differed greatly with our third party partner, so we needed to meet in the middle of the experiences.


screenshot of the final voice search

Learn and Iterate

We wanted to find the right balance of information to display on the search results. The user’s goal in voice search could range from finding one particular asset to simply browsing all available comedy movies. Through prototyping and testing, we found the key amount of information that would allow users to make a decision without being overwhelmed while trying to browse content.



Copyright © Katie G Tanaka 2024 All Rights Reserved.