How to make voice technology a business tool

Voice technology brings a number of opportunities, but there are also challenges to overcome –- security being the big one.

alexa virtual assistant echo amazon alexa voice control
Getty Images / Amazon

Voice presents a both unique challenge and a unique opportunity. It is the most personal interface we have to interact with technology in a robust way. It is more efficient than typing (and pointing and clicking and tapping and swiping).

Since Apple introduced Siri in 2011 and Amazon announced Alexa three years later, voice technologies have been adopted by an amazing swath of humanity. The overarching concept of interacting with a computer or other device by voice is no longer foreign to most people and the technology has generally proven itself in the consumer space (even natural language processing -- a cornerstone of voice -- has come far enough to bring what was once sci-fi into our daily lives).

Related: Why CIOs are moving to voice-first strategies

The ubiquity of voice and the range of several different voice assistants mean virtually every device  ̶  PC, smartphone, watch, earbud or TV  ̶  offers one voice platform or another, with dedicated devices voice enabling rooms, entire homes and offices.

The two big questions for IT are these: How do I secure voice? How do I take advantage of it?

Voice represents a paradigm shift much bigger than mobile because it is ambient computing at its fullest, meaning that it is always available across many devices and it requires no additional keyboard (real or virtual) or other input/interface. This throws the vast majority of rules for system administration and software development out the window. Let’s break it down for each of these disciplines.

Voice policy administration

The biggest problem with managing voice is that there’s no clear demarcation as to what technology to manage. Voice crosses the lines separating PCs, mobile, wearable and internet of things (IoT). 

Despite the promises of Unified Endpoint Management (UEM), there is no product or service that can manage voice across all those technologies. Making things more difficult, some of these device categories don’t even support any degree of real manageability. Mobility management is the logical point from which to manage wearOS and watchOS yet the available options are limited in regards to allowing or disallowing a device. Some wearable devices don’t even offer that option, leaving disabling Bluetooth as the only recourse.

IoT devices are no better. IoT management in the enterprise is still a nascent technology and is, quite rightly, focused on devices like thermostats, alarms and other infrastructure. Even if IoT management was more mature, voice likely wouldn’t top the list of capabilities -- despite Amazon and other companies beginning to press smart speakers and similar products into the business market.

Even in areas with mature management solutions -- mobile and desktop -- there are key challenges. One is whether or not operating systems support restricting or shaping voice capabilities and to what extent. Related is how difficult it is to implement controls across platforms in a coherent way (searching for that almost mythical single pane of glass administrators long for).

Another is the diversity of voice assistants out there. No one assistant has emerged as the assistant. The popular options, predominantly Alexa, Google Assistant and Siri, battle for dominance with credible challengers from Microsoft, Samsung and others.

This isn’t a big problem in a home, individuals can choose a single platform or use more than one (each for different tasks or contexts) easily enough. Standardizing for the workplace is another matter, particularly in regulated industries like healthcare and finance.

All of this means that there’s no technical solution to implement voice policies. Much as in the early days after Apple released the iPhone, policy needs to be a collaborative effort among stakeholders -- IT, Human Resources, executive staff/management and employees.

It won’t be a perfect process but voice offers incentives around deployment that mobile didn’t.

Chief among these is deployment itself. With the exception of mobile devices, voice requires IT investment. Placing smart speakers throughout a building is an undertaking even if only in conference and management spaces. The alternate hardware option, using PCs themselves (and Microsoft’s Cortana) actually plays right into ITs wheelhouse.

In short, if voice is going to go beyond mobile and wearables, there’s a hardware requirement that must be met. This means employees and executives have an incentive to adhere to policies if they want to be able to use a voice assistant. It also streamlined the choice of assistant to an extent based on the hardware selected, though there is nothing to prevent access to alternative choices via personal device or app. Using an assistant integrated into enterprise systems, if possible, can help nudge users to accepting policies.

In some ways the challenge may not be in getting policy adherence or a preferred assistant chosen. The bigger challenge might be in getting one used in the first place. Although voice is a rapidly grow interface, it isn’t universal and some people will take time to trust and acclimate. Some holdouts may simply never get over an aversion to voice.

As with most technology, example is going to be the best teacher. The more employees see their coworkers using voice, the more its potential will unfold. This is also where getting buy-in for your voice policies and preferred assistant(s) is key. If voice is introduced along with these, preferred behavior will be introduced as well.

Building apps for voice

The other challenge to voice in business lies in creating tools that leverage voice to provide workflows that are capable, efficient and easy to use. This is a bigger challenge than it might seem. Almost all software has been designed with a screen in mind. Developing for voice changes that and requires a real rethink of how human beings interact with technology.

There are, of course, development tools out there for various voice assistants ranging from full SDKs for Alexa to iOS Siri shortcuts – so, there isn’t a lack of capability. What’s needed is imagination, creative risk taking and a willingness to start over from scratch.

One good starting point is to explore business and enterprise apps that already offer voice integration. See what works and doesn’t before investing in voice. Once you do start your own development, initially choose simple problems/needs that impact a broad range of people. This low hanging fruit will let your teams gain experience and confidence while also showing off voice to a wide audience. Make something that works well even if it’s limited at first. Don’t try a moonshot project out of the gate.

Related: 3 ways AI assistants improve enterprise productivity

Realize the goal posts will keep moving

What’s capable with voice is constantly evolving. The comfort level of users keeps changing. Even the mix of assistants is dynamic. It’s crucial to realize that your approach to voice cannot be static or one-and-done. Be prepared for projects that evolve iteratively and don’t be afraid to throw out today’s success when it becomes tomorrow’s boondoggle (and that will happen- more than once).

When you are your teams are in the weeds with voice, remember that you are literally building the future of workplace computing and that is no small thing.