Are Voice Interfaces Finally Making Themselves Heard?

Voice is the most biologically natural form of communication among humans. Texting may be popular, but very few of us can type 160 characters in 17 seconds like Guinness World Record holder Marcel Fernandes Filho. 

As a species, the evolution of speech and language has enabled us to form complex communities, understand the mechanics of our environment, and reach the current state of our civilization as we know it today. But for 30-plus years, the interface into computing has not been our voice–it has been a plastic ball and two buttons (the mouse), combined with a rectangle of plastic keys (the keyboard) that we press to make words.

Voice assistant technology, in its latest form, has been around for several years now, with Apple’s Siri, Google’s Google Now service, and Microsoft’s Cortana. The technology has more recently caught fire with the increased adoption of Amazon’s Alexa-based devices (the Amazon Echo and Amazon Dot) and Google’s Google Home device.

Just last month, Samsung also announced its own voice-based assistant for the upcoming Galaxy S8 smartphone line of devices. And Facebook has its “M” service, not to mention a host of smaller voice services, such as Ozlo and Viv, that are available.

These recent advances in voice recognition technology are demonstrating that natural language interfaces may finally be ready for mass adoption. And with that milestone comes a number of exciting marketing opportunities.

Brief History
Voice technology has been around since the 1950s when Bell Labs created “Audrey,” a technology that could recognize digits that were spoken. As the rest of the 20th century unfolded, voice technology expanded from digits to dozens of simple words to a vocabulary in the thousands.

In 1987, World of Wonder released the “Julie,” doll which could be trained to understand a child’s voice.

In 1990 Dragon released its first speech-recognition product, Dragon Dictate, which promised to translate speech to text directly to a PC. The first version cost a whopping $9,000. Dragon continuously improved the product over the following decades, but it required a training sequence in order to replicate a user’s words accurately. In the early 2000s, voice recognition stalled at around 80% accuracy as companies focused on other technology priorities.

Google breathed life back into voice technology when it launched “voice recognition” in 2010 as part of the Android smartphone software platform. Apple followed with Siri in 2011 for the iPhone 4s. Next, Amazon introduced Alexa via the Amazon Echo in 2014, and Google Home became available in November 2016.

How Big Will It Be?
The voice-assistant or voice-activated space could potentially be huge, but it is still very early. Market researchers are offering forecasts over a wide range of outcomes. For example, Tractica suggests the market will be $5.1 billion by 2024, while Grandview Research has forecast a $127 billion market seven years from now. 

On the marketing side, Juniper research has forecast that $12 billion will be spent on voice-assistant targeted advertising by 2021, mostly through mobile assistants such as Siri, Cortana, and Google Now.

What Is The Use Case?
In 2014 Chris Messina coined the term “conversational commerce” to name the space defined by combining voice interaction with machine learning and chatbots so that consumers can interact with brands on both a transactional and affinity basis.

Voice assistants create a more compelling level of interactivity between users and brands because they leverage a more organic form of communication. Brands can offer a host of new potential value when voice assistants are combined with machine-learning technology that powers the growing number of chatbots (e.g., small software programs that power natural conversations between humans and computers and often connect multiple services together).

Here are a couple of recent examples of how brands are experimenting today:

Domino’s Pizza
Amazon Echo users can add a “skill” (Amazon’s term for an app) to an Alexa device that enables users to order a pizza directly from the pizza chain using just one voice command. The skill requires that you configure a preset order on Domino’s mobile app first and then say the exact phrase, “Alexa, open Domino’s and place my Easy Order,” which is somewhat more complex than just saying, “Order a pizza.  But it’s an intriguing early experiment into what voice assistants are capable of in terms of marketing transactions. Taco Bell, Wingstop, Burger King, and Pizza Hut are all experimenting with different forms of this kind of voice-enabled transaction.

Macy’s On Call
Macy’s provided consumers with an in-store shopping assistant via a smartphone chat app. The assistant is being tested in 10 stores and provides users with answers to questions in natural language about where items are located, what services are offered in a store, and, in a couple of stores, to request assistance.

What Is The Marketing Potential? 
As with the desktop internet and mobile devices, many of the same campaign treatments can be applied to voice-enabled devices and apps:

  • A request for a sports score might result in an ad for a future game for that team.
  • Asking when a certain live theater event or musical performer is coming to town could result in an offer to purchase tickets for the event.
  • Tie-ins to media content could unlock new forms of commerce. For example, if someone is watching an Amazon Prime show on her TV and asks what outfit one of the actors is wearing, the voice assistant could identify the outfit and offer to purchase it for the consumer through Amazon.
  • Brands might be able to combine data from different graphs to create a richer picture of consumer households and pre-emptively provide offers. By combining location-based data with calendaring access for a household (with permission, of course) Domino’s, for example, might be able to offer or even pre-emptively order a family’s favorite pizza if the underlying assistant network determines that the family is overcalendared for the day and won’t be able to cook dinner.

Risks To Consider
Of course, privacy is a big concern for voice assistance technology. A new U.S. study from BusinessInsider Intelligence suggests that a majority of those surveyed have serious concerns with device security. After all, in order to be voice-enabled, today’s generation of devices need to be “always listening.”

Voice assistants create a new level of intimacy between users and these devices, so marketers will need to consider how they create interactions in the household with even more sensitivity than they have with attribution techniques on the desktop web and on smartphones.

What Should Marketers Do Next?
While neither Amazon nor Google have disclosed home many in-home voice assistants they have sold, most analysts estimate between 15 million to 25 million devices are in service today. Some analysts forecast that Alexa-based devices could top 60 million by 2020.

That’s a substantial installed base for a new device class, but is still very much an early-adopter level of market penetration. NetPop conducted a study of typical Amazon Echo buyers in 2016–if your marketing targets this demographic profile, you may want to consider aggressively experimenting with conversational commerce in 2017.

If your marketing falls outside this target, feel free to allocate some funds and talent to experimenting in the conversational commerce space with clients and brands that could benefit from being associated with the latest technologies. The experiments will provide valuable learning for a time in the future when brand dollars begin to flow more meaningfully into voice based marketing.

In the meantime, stay tuned. Conversational commerce is shaping up to be a major marketing development for the rest of this decade and beyond.

Up Next

Customer Story

T-Mobile: If You Love Your Customers, Set Them Free


RQ Media’s Salzman: Connecting Brands To Celebs Who Really Care

From the Blog

Dig Deeper