Voice is a fundamental platform that all software developers must at least consider when building new apps. The growing presence of voice on mobile devices, desktops, home assistants, and appliances represents an accelerating shift away from graphical user interfaces for many key functions. Developers are being tasked with integrating new and creative uses of voice user interface and functionality into both new and existing apps. Simply put, software that can use voice, should.
Let’s take a closer look at how voice is fundamentally changing how many developers approach their apps and what the impact will be for the years to come.
Why Voice is a Must in the Modern UI
There were more than 1 billion devices with voice assistant integration through the end of 2018, and the majority of those are smartphones with heavy integration of both Google Assistant and Apple Siri. While home assistants get all the press, the voice assistant already in the pockets of hundreds of millions of users gives instant access to a new and intuitive tool for developers.
That ubiquity means several things. First, it means users are getting used to voice activation of their devices. They increasingly prefer to ask a question naturally than type in a combination of words attempting to get the answer or activate the functionality they need. comScore has predicted that half of all online searches will be conducted by voice by 2020. Users expect voice interaction and it applies well beyond search. Second, it means a high enough percentage of app users would be able to use voice functions for developers to consider their integration.
When implemented properly, voice serves many purposes that can directly impact developers:
- Increased convenience – Voice searches are fast and convenient, requiring no screen. Users can access and use applications from across the room and get answers faster than tapping through a multi-screen GUI.
- Improved accessibility – Those who cannot traditionally use screens or have difficulty manipulating a touch interface are able to use VUI with increased ease. Especially as the ASR and NLU programming behind both the core platforms (Siri, Google, Alexa) and the apps running on them evolve, technology becomes more accessible and impactful for all users regardless of age or physical ability.
- A more human experience – Technology is incredible, but it is by definition a cold, mechanical experience. Menus are opened, links are clicked, and options are selected. Through voice interface, developers can often a more human experience. Emotional Intelligence in voice is being paired with existing AI to provide a better, more empathetic experience for users. Therefore, introducing the development of Emotion AI.
- Easy integration into existing applications – As the APIs for the major voice assistants opened up in recent years, one of the most common requests for app developers is to upgrade existing applications with voice functionality. There are many tools now making this easier than ever to do.
Developers who integrate voice functionality are seeing higher adoption rates and better overall user performance in their apps than those who don’t. It’s no longer just a “nice to have” function. It’s expected and often requested from users.
The Creation of an Effective Voice User Interface
App developers must now not only consider the role of voice in the user experience, but integrate a functional VUI that supplements, and in some cases replaces, the GUI. While the fundamentals of backend development remain much the same, frontend implications are significant, and the underlying technology behind voice will determine just how much can be done without a screen.
ASR/NLU Programming
Automatic Speech Recognition (ASR) and Natural Language Understanding (NLU) are at the core of successful voice interfaces. The software must be able to quickly and accurately recognize when someone is talking regardless of age, gender, or accent, and then process the actual meaning behind the user’s query.
Early voice interfaces were notoriously bad at this. Voice-activated GPS systems have been in cars for more than a decade and frequently failed to accurately capture user input, while early voice assistants were known to struggle with certain vocal patterns, accents, and colloquialisms. These platforms have vastly improved but are still only as smart as the data they are fed. Complex frontend interactions need to be tested thoroughly to ensure user input is accurately captured, analyzed and executed.
GUI Expertise Doesn’t Always Translate
Many development teams have a similar structure and balance between backend and frontend (GUI) development. That balance won’t be sufficient as voice interface development becomes a new standard. Experts are comparing the transition to what developers needed to do as software transitioned to the cloud. New platforms, tools, and ideas will be required in the realms of machine learning and pattern recognition for voice-developers to be successful.
In a traditional interface, the pace is dictated by the user. As fast they can read and respond to system prompts, the system will respond. In a voice interface, pacing must be more deliberate and capture the cadence of human speech. At the same time, there are no established rules for VUI in the way there are for GUI. We know collectively where things should be on a screen, but we don’t yet know what questions we should ask.
Combining GUI and VUI
Rarely will an app operate in a vacuum with no integration between traditional graphical and newer voice interfaces. This requires developers to think in terms of how those interfaces will interact with one another. On smart devices especially where touch and voice are both available, when will users opt for voice input and when will they use the screen? How will the app respond in each case and how seamless will the transition between the two be? These are questions developers must now consider.
The Future of Voice in Development
Voice user interfaces are rapidly becoming a user-expectation, and users will want to see those applications across their full ecosystem of applications. An iPhone app with Siri support won’t be enough. Users will ask for integration on Alexa, OSX, Windows, and everything in between. With the future of personal technology becoming increasingly screen independent, it’s important for developers to have a roadmap in place for how best to develop, test, implement, and measure voice in their applications.
