Voice Assistant App Development and Approval

Voice assistant adoption is experiencing an extraordinary growth. In just eight months (from June 2017 to February 2018) the number of households owning a smart speaker more than doubled globally. IDC Canada projects that smart speakers will be in one million Canadian homes by the end of 2018. IDC also projects that nearly one-third of Canadian businesses will launch an AI-powered assistant in the coming year.

This kind of growth trajectory presents enormous opportunities for businesses. Voice assistants can help make organizational processes (such as redirecting incoming calls to self service) more efficient, while also serving as another channel for customer interaction.

However, for businesses to get the most value out of their voice assistant application, it's essential to have well-designed applications, starting with concept to design, development and deployment.

Keys for the development of successful voice assistant apps

The best voice applications have all three user experience criterias. Namely, simplicity, usability and repeatability. Let’s take a closer look at each of these requirements.

  • Simplicity. Great apps help simplify the lives of their users. Needless complexity will cause user engagement to drop precipitously, ultimately resulting in abandonment - the last thing any business wants.
  • Usability. High-level usability is especially important for voice apps as people do not have visual cues they can use for navigational context clues. Voice apps need to work intuitively during the first interaction, as there are no visual supports nor opportunities to re-read or read later.
  • Repeatability. Most popular voice apps, the ones that keeps being used over time, deliver service to recurring problems. For example, a voice app can tell the user which disposal bins are scheduled for garbage pickup on their street for the current week, solving a small but important life problem.

Master these three elements, and you'll have an app that's compelling and easy enough to use to encourage long-term use.

Smart voice user interface design

A truly stellar application generates user satisfaction and loyalty by delivering an attractive, enjoyable and easy to use experience. In terms of voice applications, user experience should be the paramount concern in the design phase.

Users interact quite differently with voice interfaces compared to graphical interfaces. They also have different expectations, assuming that a voice app will interact more like a human conversation, rather than a touch-based graphical interface.

To satisfy these expectations, you'll need a voice interaction design that anticipates how people talk to their voice device. One example: An app can provide a new, unfamiliar user with suggestions for possible commands such as "would you like to check the weather forecast?"

Information should also remain streamlined. With graphical interfaces, users can consume large chunks of visual data and simply scroll up if anything is forgotten. Not so with voice apps, where the user must remember spoken information. By keeping information concise, it reduces the likelihood of confusion or data overload.


Development and deployment of voice applications

Amazon Alexa and Google Home make it easy for businesses to deliver new skills and actions (aka your new voice application). Amazon and Google allow users to build new skills and actions in the cloud with developer tool kits. No coding is required, though more sophisticated apps do require technical knowledge.

Though Alexa already has a vast library of skills, Amazon's API makes it easy to build a personalized app that's optimized for your specific needs. Google, meanwhile, runs the Google Assistant Developer Community Program, which helps new developers get started and offers benefits and incentives for voice app creation..

Additionally, there are a growing number of vendors who specialize in the design of state of the art voice assistant applications. Once the build is complete, skills or actions can be hosted online via Amazon or Google.

In order to speed up deployment, Amazon and/or Google approval guidelines should be reviewed as soon as an app idea is identified. This helps direct the conception of the application, avoiding along the way delays in the approval process, and ultimately minimizing time to market.

Here are a few additional considerations for conversational app deployment and approval:

  • Amazon Alexa and Google Home are closed systems. This means that approval must be granted by Amazon or Google before an app can go live.
  • Apps should be designed in a way to minimize touch points with Amazon and Google. If an app needs to add additional features for subsequent versions, it’s possible that the entire approval process will need to be repeated. This scenario can be avoided by keeping new feature functionality on a backend layer that's controlled by the app owner, as opposed to hard-coding into the language model. One example of this is the text for asking questions. You may hard code the sentence “What is your name?” into the language model, but this will trigger the certification process if that sentence changes. If this had instead been generated from a backend service, the app owner can freely make changes without going to the approval process again.

The takeaway

Great design and smart development and deployment are the keys to generating the most value from investments into voice-enabled applications. To accomplish this, it's important to do the following:

  • Focus on key design criteria such as simplicity, usability and repeatability.
  • Understand the differences between designing for graphical and voice interfaces.
  • Review Amazon/Google app approval guidelines as soon as an app idea is identified.
  • Design apps in a way that minimizes the odds of requiring secondary certification/approval from Amazon or Google.

By taking these steps, application providers will help ensure that their voice apps are useful, practical, popular and readily deployed.