Ann Thymé-Gobbel - Mastering Voice Interface

Ann Thym-Gobbel and Charles Jankowski

1st ed.

ISBN 978-1-4842-7004-2 e-ISBN 978-1-4842-7005-9

https://doi.org/10.1007/978-1-4842-7005-9

Ann Thym-Gobbel, Charles Jankowski 2021

Apress Standard

The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Introduction

Figure 1

An actual conversation between a user and a voice assistant controlling household lights

We watch our friends ask voice devices for information and entertainment, try various approaches, and too often not get the result they hoped for. They settle on a few use cases that work while they also get frustrated. Yet other friends with voice development experience worry about a Voice Winter, where an emerging technology cant keep up with the hype once its in users hands. Large companies are betting big on voice, but only a few voice apps are delivering the ROI many hope for. Why? Theres no lack of opportunity, but the streets arent lined with gold either. Gartners 2019 hype cycle analysis concludes that voice assistants are at risk of falling into the dreaded valley of despair because of fundamental barriers and challenges.

Having been around voice for quite a while, weve seen what voice solutions can do, so we started wondering, Why isnt it better? And we started analyzing, I see what happened there It wasnt tuned enough or There was an incorrect assumption about context or That wording is too vague or It's not taking meaning into account and so on, which lead to We want to use voice devices that work well!

We looked at the available resources and found that even when the material is good, the typical focus is only on one portion of the full voice creation process, either design or development. Missing was something that brought it all together to explain the why and how of voice as an interface modality without the limitations of a specific platform. Something explaining why the needs of the users, system, and business all must be met and work together. Something incorporating voice-appropriate discovery and testing methods for complex real-world applications . Creating voice applications without understanding how people process and generate speech is a bit like judging the readiness of pasta by throwing it at the wall to see if it sticks. You can do it, but its unnecessarily messy and probably doesnt have the result you hoped for.

This is where our motivation for this book began to take shape. How can we share our experiences? Canwehelp others succeed and avoid unnecessary mistakes? What do we wish someone could teach us today based on what we now know about voice development? Over the years, weve learned that success in this space starts with understanding whats special about spoken natural language . Weve worked at big companies and small startups, in different organizations doing product development, client services engagements, R&D, design, development, testing, research, teaching, and even marketing research, strategy, and planning. All of it has convinced us that the way to create successfulvoice interactionsis to work with a team that understands real users needs to apply the available technology and resources appropriately.

Heres the good news: fully functional voice systems have been around for quite a while and analysis of spoken language much longer. Voice systems did not suddenly appear out of nowhere a few years ago, nor will they disappear anytime soon. If we thought voice had nothing to offer, we wouldnt stay in this field our whole careers! But we do think about how to avoid a Voice Frost. The core of Alexa and Google Assistant voice design is based on best practices built on years of well-tested observations. And theres a reason for that: