When Apple adds a feature to its current product lineup, it means it’s ready for mainstream adoption. From September 17, the HomePod, Apple’s music-focused entry into the burgeoning smart speaker market, will be able to make and receive telephone calls as well as be able to tell if someone called when you were out or where you’ve left your phone if you’re unable to find it.
It will also receive Spanish support options in time for its official launch in Mexico and Spain, and French Canadian support as well.
But most impressive is its shortcuts feature. This gives owners the chance to school their HomePod and educate it about routines. For instance, when you say “good morning.” it will know to turn off the lights, place an order for a latte at Starbucks and play the news headlines. And it’s all thanks to growing capabilities in artificial intelligence (AI).
A week before Apple’s iPhone launch in Berlin, Germany, the venue – the 2018 IFA – was very different but the conversation was very much the same. From smartphones and smart speakers to smartwatches and smart-home companions, AI was the main attraction and the topic that had everyone talking.
“Artificial intelligence is generating a buzz,” says Stéphane Akkaoui, head of R&D at Novagile, Sitel Group’s software venture. “It’s a very media-friendly subject and one of those truly disruptive technologies impacting every part of our lives. For instance, every single Google product is already AI-powered. However, as exciting it sounds, it’s often too easy to focus on science fiction rather than scientific fact.”
Comparing and contrasting the current Apple product range with the biggest announcements that came out of IFA – one of the world’s biggest consumer technology events – provides a perfect tool for separating AI fact and AI fiction.
It was the venue where AI was being channeled for everything from composing photos via smartphone cameras to calculating whether calories burned were due to swimming, climbing, cycling or waiting for an Uber.
However, from Amazon and Asus to Samsung and Sony, the bet every company is making is that the technology’s ability to process natural language is ready for the big time.
Alongside new Google- or Amazon-supporting smart speakers from Bang and Olufsen, Harman Kardon and Technics launched new voice-operated routers, hard drives and Wi-Fi stations and in the case of Toshiba, Alexa-compatible smart TVs.
“The breadth and depth of new devices with integrated support for either Google Assistant or Amazon Alexa should leave no one in any doubt that the age of the vocal assistant is just months away,” stresses Geoffrey Boulakia, General Manager EMEA of TSC, Sitel Group’s digital customer experience (CX) agency. “It’s why at TSC we’ve spent 2018 building a dedicated team ready to support brands in identifying use cases and developing their own voice apps and voice bots so they’re ready for this vocal revolution.”
The popularity of voice as an interface is more than just talk. Google Assistant and Apple’s Siri each have 500 million users thanks to their integration in smartphones, tablets, and smartwatches. And now that these assistants are baked into smart speakers too, the market is exploding. In April, research from Voicebot.ia revealed that one-in-five American adults – that’s 47.3 million people – have access to an Amazon Echo, Google Home Assistant or Apple HomePod.
“The rapid growth is not surprising. A host of consulting companies has carried out in-depth research and the findings are always the same,” Boulakia comments. “Just 10 percent of U.S. consumers are unaware of voice technology. This means that unlike say, the smartphone, which needed a group of savvy early adopters to drive it into the public consciousness, devices like smart speakers are owned and used by all generations and age groups.”
A PwC study conducted earlier this year found that among owners, voice is already the preferred interface for web searches and for texting friends and family because it feels more natural than any other interface. Moreover, once a smart speaker is in place, it opens the doors to other connected and smart appliance purchases.
Amazon’s VP of smart home, Daniel Rausch, revealed at IFA that Alexa now works with 20,000 devices, from notebooks and games consoles to TVs and smart locks and that since January: “Alexa has sung ‘Happy Birthday’ millions of times to customers, and she’s told over 100 million jokes.”
But in terms of business, the most interesting statistic is that between January and August, the number of brands using Alexa has jumped from 1,200 to 3,500.
“If done right, voice offers the greatest opportunity to remove friction from the customer journey,” Explains Boulakia “The best interface is no interface and the best journey is a personalized journey. And as the technology matures and the data feeding it grows in volume, the ability to predict a request and to respond with the right emotional tone as well as with the right information also increases. This is the era of conversational business.”
But this era has arrived because of the quality of data that already exists. It’s the same data that has enabled the launch of over 300,000 chatbots on the Facebook Messenger platform alone and that is now allowing brands with foresight to find the right words to have a spoken conversation with their customers.
“As a global organization dedicated to customer experience, over the past 30 years Sitel Group has built up a considerable volume of interaction data around different touch points and different intentions along the customer journey,” Boulakia proclaims. “And it’s that level of understanding that’s needed in order to really leverage AI in a way that does add value for our clients and for our employees.”
For example, Novagile is currently developing a system to enhance contact center agents’ capabilities.
“As one of a number of AI-focused R&D sprints we’re undertaking, we are building a tool that gives our agents the right answer to a specific question in a specific conversation with a customer, regardless of vertical or brand,” Akkaoui begins. “Just for the first prototype, we need a dataset of 1.4 million interactions to analyze and cluster into a pool of intentions or responses.”
The suggested answers to customer questions would be drawn autonomously from this pool and while it’s a large enough dataset to develop an initial solution, millions more interactions would be needed to improve the system sufficiently so that it could be industrialized and widely used.
Luckily, historical data is available and with each iteration of the tool, the time to deployment shortens – from months to weeks. But, more importantly, if the solution works and starts to deliver the value intended, it becomes self-propagating, generating the data needed for constant improvement.
“AI can create a virtuous circle,” Akkaoui says. “If something works, it gets better and better. As more people use it, more data will flow and that data improves the tool or solution. This is why a vocal assistant will only get better. It’s also why many of the home robot companions on display at this year’s IFA failed to impress.”
No one wants a robot in their home until it is the finished product. However, no company can develop a smart home companion unless they can generate data, via user interaction to train it. In other words, the AI hype around robots is for the moment artificial. As is the concern that automation will lead to the end of work.
“Technology will free people from the mundane and repetitive elements of their jobs,” Boulakia says. “Not just at TSC, but right across Sitel Group we are pioneers at leveraging AI to enhance – not limit – our employees’ capabilities, particularly in contact centers. Intelligent automation is letting them focus on the crucial, moment-of-truth elements of their role, where emotion and empathy are paramount, while tools such as chatbots help to deflect low-value interactions and empower customers.”
All of which means that the only smart assistant in the typical home or office for some years to come will be of the voice-activated, virtual variety.