How voice technology will re-shape business
A huge gulf exists between what businesses know about voice-enabled technologies, and what they do about it.
To illustrate that fact, Globant surveyed more than 600 senior decision makers and found that while huge majority of respondents see big benefits to voice technology, only 31% use it daily at work.
It’s no secret that voice technology is valuable. Just look at the consumer smart speaker market. More than 43 million Americans now own one. Many of the people who will eventually use voice technology at work are already using it every day at home.
Google says some 20% of all searches today are voice searches.
Within two years, 30% of all web browsing will be done without a screen, according to Gartner, along with half of all searches, according to ComScore.
And yet only a tiny fraction of this is being done at work for professional purposes.
Smart speakers – such as the Amazon Echo line and related devices, Google Home devices and the Apple Home Pod – are directly aimed at consumers. But all the major companies are working on developing platforms for businesses and enterprises.
Pindrop says that three-quarters of all businesses are planning to invest in customer-facing voice systems based on Cortana, Google Assistant and Alexa. Smaller but significant percentages are investing in similar systems based on IBM’s Watson, Apple’s Siri and even Samsung’s Bixby.
IBM’s Watson Assistant is a white label tool for enterprises to build voice-activated virtual assistants using their own datasets. It’s basically Watson Conversation and Watson Virtual Agent cobbled together, and uses IBM‘s analytics APIs.
IBM says the tool is preferable to using Alexa for Business because it doesn’t give the data and control to Amazon – or IBM, for that matter.
Cisco’s Spark Assistant is focused entirely on meetings inside enterprises. The company intends to build it into all its conference room devices and make it available on any client device.
The Spark Assistant is especially interesting, because it’s based on the work of a company called MindMeld.
Five years ago, MindMeld was an experimental app I used; it would listen to a conversation and kind of free-associate with search results based on what was said. Further development of this concept, especially if in-house data were employed, would make an amazing meeting tool, where a screen would be constantly refreshed with facts and context about whatever’s being discussed at the moment.
In a few years, voice technology will enable intelligent products like Spark Assistant to actually participate in meetings, interjecting facts, making suggestions and answering questions.
The assistant can also record meetings with a command, and the product will be able to take action items, send meeting summaries and help plan future meetings, according to the company.
Even Mozilla is building a voice-activated web browser called Scout, according to a leaked agenda item for a company all-hands meeting in San Francisco recently.
Everybody knows these products are here or coming soon. And the consensus is they’ll make a big impact. But what’s poorly understood is the glaring problem we’re facing with one application – voice search – and the true benefit of voice-enabled technology.
Let’s start with the Big Problem with voice technology: search.
The unresolved problem with voice-activated search, which is a major part of how people do and will use virtual assistants, is that the result is a single answer, not an endless list of ranked options.
That sounds like a small thing; in fact, it’s a huge problem that nobody has a solution for.
