If you live in a multiuser household with a smart speaker system, you may already be hearing "Alexa" or "OK, Google" all the time. The workplace has been relatively free of these hot words, but not for long.Pew Research reports that nearly half of Americans already use a voice assistant. Like it or not, voice-based bots -- voicebots -- are coming to the business world the same way they've penetrated the home. Amazon, Google, and Microsoft have different approaches for penetrating the business environment, and unified communications vendors are attempting to participate in various ways.
Voicebots in the Office
Amazon launched Alexa for Business late last year as a way of giving IT managers control over devices with company-specific voice apps distributed throughout an enterprise. And just earlier this month, it released a skill that lets employees reserve conference rooms quickly via voice without having to pull up a calendar app. Amazon isn't alone here; in late September, Microsoft released its own enterprise skills kit for Cortana.
UC vendors are looking to leverage voice assistants with their offers. Scan through the Alexa Skills directory, and you'll find many skills from familiar names -- RingCentral, Phone.com, GoToMeeting, Polycom, Cisco, and Zoom. These let you dial contacts, hear messages, and schedule conference calls. Google Home and Amazon Echo devices already let U.S.-based users dial out to a contact or any U.S. number, so these UC vendor skills aren't novel. That may be one reason they don't appear to have many takers, based on the number of user reviews for each.
Employees using voice assistant skills need to be careful they don't become annoying -- to coworkers and themselves. Clicking on a hyperlink to listen in to a conference call is quick and quiet. However, while launching a call with a, "Hey Alexa, ask ABCphoneApp to call 617-380-7152," isn't difficult -- if it works (and these systems are getting good for most speakers) -- it doesn't really add a whole lot of convenience unless you're driving or in some other circumstance where blurting out a command is OK.
Voicebots in the Meeting Room
Inside the conference room is another matter. Here, voicebots are a natural fit. Conference rooms often have minimal user interfaces, like a conference phone dial pad, and are meant for group discussion. Since people speak out loud in these rooms anyway, no one is going to get too annoyed by others blurting out voice commands.
The hardware is a good fit here, too. Conference room phones already have far-field microphones meant to pick up audio in all directions with good audio playback, just like smart speaker devices. Cisco, Microsoft, and Google already have showcased these capabilities, and our research discussions indicate other conference room manufacturers aren't far behind.
Voicebots in the Contact Center
But perhaps even more significant than using voicebots to help users dial will be using voicebots to answer the phone. The same technologies used to power voice assistants are relatively easy to adapt to IVR systems to help improve customer service. Most IVR systems today, even the speech-enabled ones, are driven by hierarchical menu systems that infuriate users as they listen to a bunch of options, choose one, and then listen to the next set of menus. In a DTMF keypad-driven world, this is the only way they can work.
Voicebots aren't subject to these constraints, allowing users to just say what they want in a more conversational manner. This results in a flatter menu system that is faster and more natural for the caller to navigate. It also allows for more options than is practical in a traditional IVR, helping to provide better routing and increased agent efficiency.
Conversational IVRs from the likes of Nuance, Genesys, Aspect, and others have been around for years. Now they're seeing competition from Amazon, Google, and Microsoft as these companies target their technology. IBM introduced its Watson Voice Gateway in May 2017 to connect callers to the Jeopardy-winning AI. Soon after, Amazon added Lex integration (its bot engine) to its AWS Connect service. In the spring, Microsoft quietly included a Skype for Business Bot Connector.
Finally, Google went much more aggressively in this direction this summer with its Contact Center AI initiative, which includes machine learning tools for the contact center and partnerships with nine contact center telephony providers. Google even added a Phone Gateway service that lets its developers add a phone number to their bots in minutes. One of Google's customers, Marks & Spencer, claimed it freed up the equivalent of 100 full-time employees by shifting to a voicebot IVR for 12 million of its calls. With prices ranging from 10 cents a minute to less than three cents a minute, these systems can be even more cost effective than traditional IVR systems, too.
This technology isn't just for inbound calling, either. Google just started publicly touting its ability to place human-like outbound calls to small businesses with its Duplex technology.
When Will Smart Speakers Be Standard Issue?
These are definitely early days for smart speaker voicebot technology in the enterprise, but things are moving fast. Amazon and Google are in heavy competition, with Microsoft trying to stay in the game, so expect frequent announcements. UC vendors are in a somewhat awkward position and have been slow here. They want to use the latest AI technology, and leveraging existing voicebot platforms is the fastest path to do that. However, no major UC vendor wants to help improve the growing telephony portfolios of Amazon and Google. Users are demanding more voice interfaces, so UC vendors must get over these fears before they get left behind.
Looking to learn more about voicebots and their place in real-time communications? Interested in how machine learning, computer vision, and speech analytics fit into your strategy? Check out our report on AI in real-time communications.
This is the third piece in an ongoing series. Check out the previous posts:
And, come back for more on the impact of machine learning and AI on the communications community.