Artificial intelligence has been around since 1956 and has made some giant leaps in that time: beating the best human at chess, the best human at US gameshow Jeopardy and recently beating the best human at complex strategy game Go.
Brands have only recently started adopting artificial intelligence for core consumer services. Google’s voice recognition technology now claims 98% accuracy and Facebook’s DeepFace is said to recognise faces with a 97% success rate.
IBM’s Watson, which uses artificial intelligence to perform its question-answering function, is 2,400% “smarter” today than when it achieved the Jeopardy victory five years ago.
There is no doubt that the relationship between men and machines is changing, and brands are on the cusp of making artificial intelligence an everyday element of their customer offerings.
For example, with a goal of reducing the amount of calls coming through to its call centres, Swedish retail bank Swedbank has integrated the technology tool Nina into its customer service strategy. Nina, an intelligent virtual assistant, delivers automated customer service via the brand’s website in a conversational manner. This enables self-service capabilities and quick and easy access to information for Swedbank customers.
“If you as a customer came onto the site and said to Nina that you needed a new card, she would acknowledge the request but also drill down to find out why, asking: ‘Have you lost your old card or has it been damaged?’” says Martin Kedback, head of channel management at Swedbank. “According to what the customer then says, Nina will come back with a different answer or maybe direct the customer to a different channel.”
Swedbank’s artificial intelligence platform has been operating for more than a year and now deals with up to 80% of its 30,000 monthly enquiries. The brand, which operates globally, has worked with Nuance Communications to develop Nina’s natural language understanding (NLU) technology.
There is no attempt to trick customers into the idea that they are talking to a human, though. The bank has intentionally designed the tool without an avatar or anything that tries to look like a person.
“At the beginning, we threw in what we thought would be the hundred most commonly asked questions and answers in the database and then let the customers do the work,” Kedback says. “As they continued to ask questions, we made Nina smarter.”
Machines will be responding to requests at hotel reception desks and in all sorts of environments
Calum Chace, Author of Surviving AI
Integrating AI and customer data
In the future, Swedbank plans to personalise every interaction and make them, arguably, more rewarding for customers by integrating Nina into the customer database.
When a brand integrates AI into the fabric of its core data the information it is able to access will be much richer. But such a development then poses questions about what a brand does with that data, whether it is appropriate ethically and how marketers retain trust.
“While there is huge potential for brands to capitalise on AI to create more personalised experiences, they also need to be careful not to cross the ‘creepy’ line when it comes to customer privacy,” says Rachel Barton, managing director of advanced customer strategy at Accenture Strategy. “If you integrate the technology into customer data for instance, it has the potential to run loose and make its own judgements, which could be to the detriment of customers who may feel a level of intrusion they didn’t necessarily sign up to.”
If businesses can strike the right balance between adhering to their brand values while allowing AI to access the right amount of data, it can be highly beneficial – it is very cost effective and can deliver real-time personalisation that may not be possible via a human. However, Accenture research finds that when customers want to complain or talk through a complex situation they want to talk to a human.
Hospitality is one of the sectors thinking more boldly about the opportunities of AI. Luxury hotel portfolio Dorchester Collection is using it to identify what guests want, not what marketers think they want. To enhance its customer experience, it is using the AI Metis platform, which allows it to eschew the standardised hospitality industry measurement techniques such as mystery shoppers and customer satisfaction surveys and, instead, tap directly into digital customer feedback.
“Metis reads thousands of customer reviews and tells us what really matters to our customers,” says Ana Brant, The Dorchester Collection’s director for global guest experience and innovation. “Think of Metis as a giant focus group that not only facilitates the sessions in multiple languages but also summarises key findings, puts the findings in [context] with competitors and tells us stories worth listening to.”
The brand recently completed a comprehensive, brand-wide Metis study, which contained 7,454 guests reviews from 28 different hotels and 10 major hotel brands across 18 cities and regions. “New information is available to us every second – the biggest challenge is the constant pursuit of meaningful analysis,” Brant says. “She [Metis] summarised all the findings in a 30 minute interactive video and provided us with invaluable insight, further redefining our competitive advantage.”
Also in the hospitality industry, hotel brand Hilton has made AI a key part of the customer experience in the Hilton McLean in Virginia, USA. In partnership with IBM, Hilton is using Connie, the hospitality industry’s first Watson-enabled robot concierge, to cater to guests’ needs for information about a hotel and the surrounding area.
Connie gives consumers quick access to personalised information through cognitive reasoning and robotics. Hilton argues that Connie enhances the customer experience. Examples of interactions guests have with the robot include business travellers asking Connie for directions to their conference room or a family asking when the pool closes.
“We’re continuing to test and perfect Connie based on guest and hotel operator feedback, which enables us to explore new ways smart technology can enhance our guests’ travels,” says Jonathan Wilson, vice-president of product innovation and brand services at Hilton. “We’re at the forefront of using cognitive reasoning in hospitality – this is just the beginning of what’s possible. Future meetings at Hilton properties can be smarter and more productive because of Connie’s ability to quickly sort through massive amounts of online data to field on-the-spot research and data requests.”
AI is being used across sectors to improve efficiency, reduce costs, increase revenues and boost customer satisfaction by improving on key areas of customer experience. According to Calum Chace, author of Surviving AI, this is a very interesting time for AI. “In the past few years, machines have got better than us at recognising images, particularly faces, and recognising speech,” he says. “Those abilities mean we won’t have people in call centres for long – machines will also be widely responding to requests at hotel reception desks and personal enquiries in all sorts of on and offline environments.”
Smarter stays at Starwood
The introduction of robots and artificial intelligence (AI) in other guises is firmly under way in the hotel sector. UK brand Edwardian Hotels has launched a ‘virtual host’ service called Edward. For guests who enjoy interaction via text message, Edward will check you in before you even get to the hotel, get you spare towels or reserve you dinner.
Meanwhile, the Henn-na Hotel in Nagasaki, Japan, has a front desk staffed by robots and image recognition in place of keys.
In 2014, Starwood hotels ‘hired’ Botlr, a robotic butler, to help free up front desk staff’s time at the Aloft in Cupertino, Silicon Valley. It was Starwood’s first experiment with a robot and is a marker of the hotel brand’s enthusiasm for technology and an AI future.
“We are experimenting with AI and testing ways for guests to ‘speak’ to their rooms to control things such as lighting, temperature, making reservations and more,” says Brian McGuinness, global brand leader for Aloft Hotels. “Guests want control and we believe that by eventually weaving AI into the hotel experience, we can make a room smarter, giving guests more control over their stay.”
The brand plans to differentiate itself by the high personalisation levels it will offer. It is exploring ways for rooms to “pay attention” to guests’ personal preferences, so their favoured entertainment channels and environmental conditions will be available instantaneously.
Starwood acknowledges that technology is only as good as the people who work alongside it – especially in hospitality. “We think the future will see the rise of truly ‘smart’ rooms that respond to guests’ wants and needs just like a human would,” says McGuinness. “But we believe the human touch is a critical part of the hospitality business and don’t think robots or AI can or should ever replace the human touch.”
Artifical intelligence in action
Health and fitness routines are informed by personal, physiological and behavioural factors. Sports apparel brand Under Armour’s UA Record fitness app integrates machine learning technology to get under the skin of users. The app analyses personal food, exercise and sleep data in combination with insights from other anonymised members of the community, providing timely advice and motivation.
Budget airline easyJet is using artificial intelligence to analyse mind-boggling amounts of data. The airline plans to use AI to provide a service to match consumer demand. It is already a part of easyJet’s revenue management strategy and the airline will apply it in areas such as predicting catering needs and optimising destinations and flight times.
Food brand Knorr has used cognitive technology, a component of AI, in its recent Love at First Taste campaign. A ‘Flavour Profiler’ is integrated into the brand’s site and offers tailored recipes. Behind this is a cognitive engine that can interact with humans ‘naturally’ and from which consumers can learn what recipes best suit them, based on their individual flavour profile.
When it was launched in 2003, Skype’s internet voice call software seemed like something out of sci-fi. Now it plugs into machine learning to deliver voice recognition and real-time translation. Users of the Skype Translate service speak into Skype, have the words translated into text and that text then synthesised into spoken Spanish, English, German, French, Italian or Mandarin.
WayBlazer is a travel app that uses IBM’s Watson AI technology to give personalised recommendations. The B2B company works with businesses that merchandise hotels, tours, cruises and activities. Using Watson’s natural language abilities, it provides a way to personalise recommendations for travel, including local insights and points of interest for consumers.