Even for a casual observer of the tech industry, it’s easy to see that digital assistants such as Amazon’s Alexa, Apple’s Siri, Google’s Assistant, Microsoft’s Cortana, and Samsung’s Bixby are hot. Whether embedded into the latest smart speakers, built into the latest device operating systems, or evolving as independent entities, these new anthropomorphized chunks of code are receiving enormous amounts of attention from tech vendors, the tech press, and the general public.
But as hot as they are, there’s confusion among consumers about what digital assistants are and what they’re supposed to do. That’s largely because the various assistants offer an overlapping, but different, set of “skills,” to use Amazon’s term.
My firm, TECHnalysis Research, surveyed just under 1,000 people who own and use smart home and digital assistant products. The key takeaway of the research is that no one assistant currently has the ability to meet most people’s needs—more than half (55%) said they use multiple digital assistants. And just over a third of that group said they weren’t comfortable doing so, because of differences in the ways assistants currently work.
The various assistants understand different “wake-up” words, and different phrases to call up specific data or actions. If people have to learn how to address and ask questions or make requests of assistants in very different ways, it may cause confusion that ultimately leads to a slowdown in adoption of assistants in general.
The decision over what assistant to use depends on the device they live in. Those that live in smart speakers (“Alexa . . .”) are used more than those built into phones or PCs (“Hey Siri . . .”), the research shows. Interestingly, multi-assistant users most often select an assistant based on the device that’s closest to them.
It also depends on the task the user has in mind. Eighty-six percent of smart-speaker owners ask their device to play music, for example, while just 52% of smartphone users ask their assistant to do so, and only 46% of PC-based assistant users.
A New Concept
The path to consumer-friendly personal assistants includes a few milestones in computing (don’t forget Microsoft’s Clippy), but Apple’s Siri was the first to popularize the concept in 2011. It presented itself as a mobile, voice-driven interface (with a limited vocabulary) that allowed certain types of commands to be made and certain types of information to be requested. Even as it has expanded its applications, Apple has sought to keep Siri’s skills limited. Siri was also limited by the single basic microphone found on an iPhone or iPad.
The “big bang” in the digital assistant world came with Amazon’s Echo device, which was announced in the fall of 2014 and became widely available in the spring of 2015. The Echo used a specially designed microphone array optimized to get a clean “input” signal from your voice. More importantly, though, Echo’s assistant, Alexa, also brought with her a wider vocabulary of recognizable words than Siri. And a year into her life, Alexa learned even more words as developers created third-party skills for her.
Along the way, Microsoft introduced Cortana. Google first tried Google Now, then renamed the assistant “Google Assistant,” suggesting that the artificial intelligence-powered assistant would underpin lots of Google services. In addition to voice, both of these assistants also brought with them an enhanced AI focus. Each in their own way, they proactively learn things about you by looking (with your approval) at your productivity and personal data. They use this data to offer more personalized and contextually relevant data to you throughout the day. Siri and Alexa have steadily added these machine-learning capabilities, too.
Samsung’s Bixby and the Google Lens add-on to Google Assistant give the digital assistant the power of sight. The use the host device’s camera to capture images of people, objects, and places around you, then use sophisticated AI to recognize and manage them. This new twist adds an intriguing layer of contextual information. The assistant can see what’s going on around you.
Moving forward, we’ll see even more capabilities added to digital assistants, including more natural responses, the ability to sustain longer conversations, and other skills that haven’t even been thought of yet. In many cases, however, those advancements are likely to be rather subtle, such as moving from a two-question interaction to a three-question one, or providing a slightly more contextually appropriate response to one of your queries.
As assistants gradually get smarter, they’re likely to specialize in different ways to accommodate different users and use cases. Actually, we’re already seeing differences in approach among the various assistants. One assistant might be good at providing answers to random questions (“What’s the capital of South Dakota?), while another might be far better at helping a user organize a future conference call.
While the ultimate goal of all the assistant platforms may be the same—to provide the digital equivalent of a personal assistant—each of the big tech companies creating them is coming at the concept with a unique perspective and with its own agenda. Apple wants to provide links to its services, all while maintaining the privacy of individuals. Google wants to leverage search to gather lots of personal data to create a smarter assistant (and potentially open up new advertising channels). Amazon wants to provide an intuitive, natural language front door to its many services, and maybe gain a foothold in the tech platform wars.
In practical terms, that likely means the various assistants will end up asserting themselves as very distinct, and different, personalities.
This all matters because digital assistants represent the next great tech platform battle. All the major tech players know this, and are working to create platforms that can be built upon to sell additional services, provide deeper hooks into their respective ecosystems, or develop other types of products or business models that have yet to be completely figured out. Because of that potential, digital assistants now carry with them the gravitas of being inherently important for future business.
At the same time, all these major tech vendors also recognize that it’s still early days for digital assistants, so there’s a chance for almost anyone to release a product and grab some market share.
For consumers, digital assistants represent an intriguing glimpse into the future of more personalized, more contextualized, and more meaningful types of computing experiences. They show the potential for all the intelligent devices around us to go way beyond simply showing us pretty digital things on various screens. Assistants could enable completely new types of interactions with users. They could become like personal concierges that show us just the right information at just the right times. They might also provide a bridge between the digital world and the analog world we inhabit.
The era of the digital assistant is coming, but getting there—and determining where there is—will mean a rocky climb, as we search for the right words to speak to overlapping, incomplete visions of that future.
Bob O’Donnell is the president and chief analyst of TECHnalysis Research, LLC a market research firm that provides strategic consulting and market research services to the technology industry and professional financial community. You can follow him on Twitter @bobodtech.