While smartphones remain the most widespread type of hardware embedding assistants and voice technology, 2018 was a significant year for the rollout of smart speakers across the globe. Nine million units were shipped worldwide in the first quarter alone.[i] Google added Korea, Mexico and Spain to its list of countries where Google Home is available. Amazon launched its Echo in key markets like Australia and France. On Prime Day, the best sellers worldwide were the Fire TV Stick with Alexa Voice Remote and Echo Dot.
Manufacturers have also intensified their efforts to make their assistants available well beyond their sole platforms. In January, Qualcomm and Microsoft partnered to make it easier for hardware manufacturers to integrate Cortana.[ii] In April, Alibaba brought its assistant Tmall Genie to Chinese Mercedes.[iii] In June, Sony embedded Google Assistant in its TVs.[iv] In August, Amazon launched Alexa Auto SDK to bring its voice assistant to more cars.[v] Each month, new platforms join the already substantial lists of assistant-compatible devices (12,000 devices for Amazon Alexa[vi]and 5,000 for Google Assistant[vii] as of May 2018), bringing omniscience to the most popular personal assistants.
As digital assistants and voice technology availability skyrockets, now is the time for brands to enter the assistance world. Not only is the space highly connected, but it will reward early, innovative players. According to the iProspect 2018 Global Client Survey, developing a strategy for voice and digital assistants is the second fastest-growing priority for marketers in 2019 (vs. 2018), only second to maximising sales in online marketplaces. Voice search and assistants are the two priority emerging channels in 2019.[viii]
Although technology giants are very active in the development of voice technology and virtual assistants, do consumers embrace these new opportunities to interact with technology as they become available?
To better understand consumers’ actual views and expectations, iProspect surveyed more than 10,000 respondents across Europe, Asia-Pacific, and Americas during the last 12 months.
CREATING BRANDED INTELLIGENT AGENTS
As people get more and more familiar with assistants across the globe, availability is only half the battle. Assistants will need to resonate with users on a daily basis to really stick over time. This means not only demonstrating their utility, but also their potential to be trusted.
To establish safe environments, Google already offers some control to users by allowing them to erase records of interactions with its assistant. However, to build heightened levels of trust and intimacy, Emotional Intelligence will be critical. So what should marketers consider when creating their intelligent agents?
From call centres to open frameworks
Most brands have already started creating an intelligent agent without even realising it. Call centres are goldmines of recorded conversations that can be leveraged as training materials for an AI system. By converting these conversations into understandable labelled data, AI systems can learn what matters most to current customers, what issues they have, what makes them happy, and what makes them frustrated. These conversations can grow into a full dialogue system. As consumers constantly switch between channels, open frameworks should be preferred to closed ecosystems to make intelligent agents more useful and relevant wherever consumers are looking to converse with the brand. It is equally important to put a plan in place so that customers can easily discover the brand’s conversational technology and know how to use it.
Conversations that flow
When asked to describe the personality and character traits of a digital assistant, the first two adjectives users chose are ‘Helpful’ (60%) and ‘Smart’ (57%).[ix] Brands should remember actions are the main goal of a digital assistant and use chatbot UX designers to help diagram actual conversations and dialogue.
To create a natural flow, developers should feed their assistant a variety of responses (upwards of 20 for each scenario) so that it can piece together unique phrases. Keeping context top of mind is key to success. Although maintaining context in conversation comes naturally for humans, it’s still one of the hardest things for a machine to do, resulting in irrelevant responses and frustrating results.
Finally, customers should be able to respond in various ways, from Boolean yes/no, to images, to free form. Intelligent agents should be able to learn from customers’ responses over time and continue to adjust.
Emotional intelligence and personality matter
Interestingly, the third adjective users choses to describe the personality of a digital assistant is ‘Friendly’ (44%), a very human characteristic. Consumers don’t want to engage with robots, but with brands who know them, who help them, and who care. Through its intelligent agent, a brand speaks, sees, feels, and thinks. It welcomes customers and makes them feel special. What does a brand sound like when it speaks? Smart? Funny? Cool?
Myers-Briggs personality types (MBTI) can be used as a baseline for developing a brand’s personality and creating a consistent, human-like persona. If your agent talks, explore using voice inflection within responses to better communicate and avoid sounding robotic.
This article is excerpted from Future Focus 2019: Searching for Trust.Download Future Focus 2019 for key insights and success stories on navigating truth and authenticity in 2019.
Discover our other resources around Voice and Assistants:
And many more in our Resources section!