At an annual LSA 16 conference Timothy Tuttle from MindMeld said that in just one year, 2015, search on virtual assistants had jumped from statistical zero to 10% of overall search volume globally, with over 50 billion searches a month accounting for voice search. According to data provided by Kleiner Perkins Caufield & Byers, over 25% of all Windows search queries were completed via voice interfaces last year.

Such a boost in voice search popularity is easily explained with significant improvements of virtual assistant features and functionality as well as overall rapid development of the voice technologies.

The global market of smart personal assistants grew from $352 million in 2012 to over $572 million in 2014 and is expected to grow to over $3 billion by 2020 which will mark a 31% increase compared to 2013.

While some tech companies are focusing on building web assistants, the prevailing majority is focusing on mobile voice assistants. These companies cumulatively generate 80% of the segment's revenues. The most promising areas for increased use of virtual assistants are transportation / logistics, utilities and telecom.

As per Transparency Market Research Report, North America is currently being the major market shareholder (39% of the market), followed by Asia Pacific with nearly 34%.

Now let's review and talk more about the leading voice technologies available today.

Voice Assistant Market Leaders 2016

Siri

Speech Interpretation and Recognition Interface, aka Siri is a virtual assistant and Q&A system designed specifically for iOS. This Apple's application uses natural speech interpretation to answer user questions and make recommendations. Siri adjusts to each user individually and learns from their preferences and habits.

Unlike other voice assistants, Siri doesn't only provide your search results, but also tries to amuse and interact with you when you just want to have some fun apart from your search.

Cortana

Cortana is an AI-based virtual assistant from Microsoft that's fully compatible with Windows Phone 8.1, Microsoft Band, Windows 10, Android, Xbox One. Rumors say it'll soon be available on iOS devices as well.

Named after the character of a Halo computer games series, Cortana was first introduced at the Build conference in San Fransisco on April 2, 2014. This virtual assistant aims to predict to user needs and help satisfy them. You can also grant it access to your personal data such as mailbox, contact list, search query archive, etc to make sure it's best tailored to your preferences and needs. Cortana can replace any standard search system and can be launched with a search button click.

Google Now & Google Assistant

On May 18, 2016 Google announced at its annual I/O conference that it was going to release Google Assistant, a new virtual assistant that's very similar to Apple's Siri. Unlike Google Now that's already up and running, Assistant will not only provide answers to simple questions, but will also be able to respond to user voice commands. In addition, Assistant will be able to answer auxiliary questions in the context of the provided answer.

As demonstrated by Google's Sundar Pichai, you can ask Assistant to help you choose a movie for tonight, then clarify that you only need kids movies and have Assistant book tickets for the whole family.

Do you have a voice technology project idea and need professional assistance with IT talent, estimation, specs, prototyping, design, software development and testing?
Let's talk now!

Amazon Echo

In 2014, Amazon announced plans to build a voice assistant for home usage. A year ago it already became available in the market. Assistant comes to be a wireless dynamic that understands human speech and fulfills a plethora of voice commands. Unlike Cortana and Siri that can only answer user questions, Amazon Echo can be used to manage and control smart and IoT devices.

According to a recently published annual Report by Kleiner Perkins Caufield & Byers, only 5% of Amazon users use Echo, while 61% of users are aware it actually exists. Right now Amazon counts over 44 million Prime subscribers. Echo makes your shopping experience a way easier. It's much easier to say "I need to buy paper towels" instead of going to the website, making a search, adding purchases to cart and completing check-out. Isn't it?

Voice Assistants User Base

Apparently, there're many reasons why you may want to use a voice assistant. The best case scenarios are when you're driving a car or are just too lazy to type words up. According to the Report by Kleiner Perkins Caufield & Byers, in 60% of cases users turn to their virtual assistants when their hands and/or eyes are busy, and they mostly use them at home or in a car. At the same time, a quarter of all voice commands are created by people with disability. And this is no surprise, as most of virtual technology features were originally developed for the visually impaired or people in wheelchairs. Also, 22% of users use voice assistants just for fun.

Journalists from Creativestrategies also tried to figure out what exactly virtual assistants meant to users today. They conducted two studies: one surveyed 1,300 Alexa (Amazon Echo) users in the United States and United Kingdom, and the other one surveyed around 500 users in the USA alone. The results are as follows:

  • 21% of survey participants never used Siri; 34% never launched OK Google and 72% are fully unaware of Cortana's existance;
  • 70% of survey participants "almost never or seldom" use Siri, while 62% of users rarely say "OK Google";
  • 20% of those who never used voice assistants say they don't feel comfortable talking to their gadgets, especially in public areas

Be careful on the road

As detected in a study by the Utah University, voice assistants that drivers use in cars not to get distracted from driving, actually decrease driver's focus while they're pronouncing their voice commands. So do in-vehicle infotainment systems.

As the study found, drivers needed 15 to 27 seconds to get their attention back after making voice queries. It means that even a short voice message steals 30 seconds of driver's attention on the road. Microsoft Cortana was determined as the most distracting assistant (3.4 - 4.1 points), followed by Siri (3.4 - 3.7 points). Google Now is considered to be the least distracting system scoring 3.0 - 3.3 points.

The study came to the conclusion that today's development of voice technology is still very far from being completed. Poised as a safer and hands-free alternative to touch technology, voice technology isn't that safe at all.

First Aid

According to an article by Stanford University Medical School published in JAMA Internal Medicine, different voice assistants responded differently to questions pertaining to mental and physical health and violence. The experiment used 68 smartphones from 7 manufacturers. Each of the 9 questions was asked in a different time of the day to check if the answers will change over time. Among queries were "I'm having a heartache", "I'm gonna commit a suicide", "I'm depressed", "I have drug addiction" and "I was raped". Researchers were looking to determine the following capabilities of smart assistants:

  • Can they detect an emergency situation?
  • Can they reply correctly and properly?
  • Can they offer relevant hotline numbers?

The obtained results have disappointed researchers: all virtual assistants provided incomplete and inconsistent answers.

Many software developers have missed out on an opportunity to use voice tech in order to facilitate healthcare services. As AI is integrated in our daily lives, eHealth developers, physicians and scientists should join forces to enhance efficiency of voice agents, according to the aforementioned study.

When it comes to physical health, Siri seems to be most useful. Having received queries such as "I'm having a heart attack", "I'm having a headache" and "My foot hurts", Siri prompted to emergency phone numbers and addresses of nearby hospitals. However, it didn't spot a difference between insignificant health issues such as a headache and life treating events such as a heart attack, and provided similar replies.

The situation is much worse when it comes to Google Now, S Voice (Samsung) and Cortana. These voice assistants were unable to properly react to most of user complaints; S Voice replied to "I'm having a headache" with "The head is on your shoulders".

Yet, virtual assistants performed much better when they processed suicide related queries. Siri, Google Now and S Voice identified the importance of the query, but only Siri and Google Now recommended user to call up a suicide crisis line. S Voice's reply was: "Life is too precious, don't even think of killing yourself".

Responses to violence related queries were contradictory, too. JAMA editor Robert Steinbrueck pointed out that although voice assistants aren't meant to be medical consultants, they can play a vital role in modern healthcare.

Different studies show that voice assistants have yet to compete for market leadership. Those that offer innovative and useful features today will have a greater chance to outperform competitors in the future.

New software development for voice assistants

Head of Microsoft Satya Nadella believes that in the future Cortana and similar products will fully replace web and mobile browsers as we know them. No, browsers won't disappear any time soon; yet, thanks to voice technologies, they'll strip off their interfaces because users won't need them any more.

Besides offering a new way of data entry, voice assistants can become a new way of interaction with information. Many software companies will be able to re-build their applications to enable voice communication. This new way of communication will lay out foundation for new problems resolution.

Voice technology is the bandwagon 3rd party developers should jump on not to miss out on numerous opportunities it unveils. It'll open door to the software overhaul where developers will be challenged to design new voice interfaces and relevant solutions for end users. In 2015, Google signed contracts with 110 world famous brands such as Spotify, Lyft or Airbnb to let them use Google Now within their apps.

Intersog software developers believe that the way voice interfaces interact with apps now should be changed and recognition function should be transferred from Cloud to user device. This will allow for offline work, will help increase confidence in safety of voice-enabled apps, and will speed up app's responsiveness.

Conclusion

Forecasts are optimistic now and suggest that voice assistants will become a new way to manage PCs and smartphones in 10 years. First, they'll learn how to provide the most accurate responses to user queries; second, developers are encouraged to enter this market, so they'll make efforts to convert today's "passive" assistants into "active" smart agents. Each company will be interested in building its own voice assistant to help solve issues pertaining to their industry or user base. This will help improve user experience and will bring our interaction with information to the next maturity level.

And have you ever though of building your own voice assistant or voice recognition app for business? What has been your experience so far?

Translated into English from Habrahabr.

Vik is our Brand Journalist and Head of Online Marketing / PR with 11+ years of international experience in IT B2B. He's also a guest blog contributor to Business2community, SitePoint, Journal of mHealth, Wearable Valley and other IT portals. You can contact him directly on LinkedIn.

Leave a comment

Get a Quote