The use of voice assistants is becoming increasingly popular and offers companies a great deal of economic potential. However, their implementation must be carefully considered to ensure that users receive service that provides real added value.
Ever since Apple integrated the first voice assistant into a smartphone with Siri in 2011, interest in this type of “conversational interface” has steadily grown. According to a study conducted by the management consultancy Capgemini, 51% of the 5,000 responding consumers in Germany, France, Great Britain and the United States regularly use voice assistants such as Amazon’s Alexa, Samsung’s Bixby, Microsoft’s Cortana, Google Assistant or Apple’s Siri. Most often, they do so to search for information, play music and video content or to reach out to companies.
Voice assistants, which can be activated by using voice triggers (e.g., “Ok, Google”) or by pushing a button, transmit the user’s spoken requests to cloud-based software the assistant are often based on. This software, in turn, processes the request and then outputs information via a simulated conversation, or initiates activities or transactions, as the case may be.
Controlled via readily available smartphones, dedicated loudspeakers or laptops, users, have become increasingly acquainted with this sort of interaction. Hence, the days when most people perceived speaking to devices as unnatural and unpleasant seem to be long forgotten.
Increasing user base through intuitive usability and an expanding ecosystem
One reason for the growing use of voice assistants is their increasing ability to precisely recognizing user requests. Moreover, compared to interactions in, for example, a chat that requires text or information to be typed into a web-based or mobile interface, the use of natural spoken language is considerably more convenient. For the user, interactions that use natural language are easier, barrier-free and save time – which just happen to be the main advantages of using voice assistants.
Other reasons for the increasing adoption of voice assistants are the evolving ecosystem of services and devices connected to this technology. These services are linked to voice assistants via an increasing number of Alexa skills or Google actions that continuously enhance these wizards’ capabilities. Speech input can control more Internet-of-Things-enabled devices such as household and entertainment appliances. Moreover, the automotive industry, for example, is embedding voice assistants in its connected cars.
Commercial potential of voice assistants particularly strong
In addition to information retrieval and voice-based device control, an Adobe Digital Insights study found that 22% of U.S. consumers is already using voice assistants for shopping purposes. While the interest in “V-Commerce” in Germany is less pronounced (12%), in three years’ time it is expected that 13% of all consumer spending will be made using voice assistants. This could go hand in hand with a significant increase in turnover, as predicted by Gartner. Increased customer satisfaction through the use of voice assistants as well as continued loyalty and positive word-of-mouth recommendations can further enhance the effect. According to a study conducted by the University of Oxford and Reuters Institute, 58% of decision-makers in the digital sector are therefore considering providing content and connectivity for this channel in 2018.
Long-term use depends on technological advancement and continuous benefits for the customer.
Whether voice assistants establish themselves as a communication and sales channel will depend on the technological progress they make. According to a benchmarking study by the company Snips, there is currently no solution that does not misunderstand or misinterpret at least some concerns. To reduce the dissonance between humans and machines, and to optimize the technology involved, particularly the following areas must be developed further:
- Computer-based natural language processing (NLP),
- Machine learning
- Computing power and telecommunications infrastructure
Computer-based processing of natural language or NLP includes the techniques and methods that enable the recording of spoken language, its structuring, understanding, and response to it. To realize this process, two additional components of NLP are essential:
- Natural language understanding (NLU)
- Natural language generation (NLG)
Since humans and computers employ different languages, NLU is required. It uses algorithms to transfer unstructured content in the form of natural language in a structured way (speech-to-text), which can be understood and processed by a computer.
Much simpler than the first part of processing natural language is generating synthetic language, NLG. To do so, the information that is supposed to be translated into text must first be defined. It is then re-structured based on a set of rules and translated into a natural language that the user can understand (text-to-speech).
Regarding this process, it must be borne in mind that high-performance voice assistants require large amounts of training information in the form of real customer service data and continuous content editing. Contrary to the promises of some providers, voice assistants do not automatically work straight out of the box. They require the careful input of information and continuous training to deliver a positive user experience.
Outlook: A promising future for voice assistants
Besides achieving flawless voice processing, additional obstacles have to be overcome in terms of data protection and security. Only if the user and his or her data are protected, and long-term added value is created, it will be possible to establish voice assistants as a sustainable channel.
If you listen to and believe in the opinions discussed in business, science, and media, digital assistants will become even more present and versatile in the future. And they are set to also extend to professional contexts. Gartner, for example, predicts that by 2020, virtual assistants such as voice assistants or chatbots will provide approximately 25% of support for customer service centers. This will also affect other areas, such as self-service and support for sales processes.
In addition, internal business processes may be accelerated and simplified using voice assistants. From office and appointment management, the processing of simple, repetitive tasks to performing research, a multitude of tasks may be assumed by voice assistants in the future.
In order to orchestrate this channel with other communication instruments, and to harmonize its content and technical aspects, it is advisable to include voice assistants in strategic and operative planning. Moreover, to successfully integrate voice assistants in organizations’ activities, it is important to consider both the opportunities and risks this technology presents carefully.