Bots are the new apps

OPINION / 2nd March 2017

“Bots are the new apps,” said Microsoft’s CEO Satya Nadella in his keynote speech at Microsoft’s Build conference last spring. His vision of how people will interact with machines in the future is of “people-to-people conversations, people-to-digital assistants, people-to-bots and even digital assistants-to-bots.” And, almost 12 months later, his vision is materialising, as many businesses begin to introduce their ‘computer programs that simulate conversation with human users over the internet’.

The introduction of auditory or linguistic Artificial Intelligence (AI) is on the rise. The ability for machines to listen selectively and communicate using written or spoken language can now be seen in Amazon’s Alexa, Apple’ Siri and Google’s Now (among many more). These virtual customer/personal assistants are using Natural Language Processing (NLP) and machine learning to engage with humans in the form of bots or chat bots (see other forms of AI in the Gartner chart below).

Artificial intelligence (AI) types

Source: Gartner (October 2016)

Microsoft, which had a rocky start with its ill-fated AI experiment Tay, is now investing heavily in bots, machine learning and AI; and it has joined the tech giants Facebook and IBM in introducing new free tools to help developers create bots.  Since launching Wit.ai (Facebook) and Bot Framework (Microsoft) and Botkit (IBM) last year more than 30,000 bots have already been made, with many more expected to be launched this year.

Here we look at some of the latest bots in travel, financial services and retail, and selected some of our favourites:

Travel

  • Expedia.com has launched a bot on Facebook Messenger. It helps travellers search for and book hotels. Users enter their destination, travel dates, and how many nights are needed. They are then served five possible listings via Expedia.com which are all viewable within Facebook Messenger. Users can also choose a hotel and confirm the booking without leaving the Messenger app.
  • Skyscanner has teamed up with Skype to offer users the option of group bookings with friends or colleagues. By adding the Skyscanner bot to the chat, it will behave like it was another group member, suggesting flights, times and prices. Like its Facebook Messenger bot, it will give users a link to the Skyscanner site to complete the reservation.
  • KLM Royal Dutch Airlines launched the first chatbot in its industry last year to create a seamless customer experience. Through Facebook Messenger the KLM bot can communicate with its customers, send itineraries, boarding passes, check-in confirmation and even delay notifications.
  • Voyages-sncf.com, the leading online travel brand in France, was the first to introduce chatbots on its website in 2012. Since then it has been using iAdvize (a real-time customer engagement platform) to manage its real-time online customer interactions. 
  • Kayak rolled out a bot on workplace messaging system Slack. Users can search for flights by entering their travel destinations and dates and the bot will return with available options, but users are taken to the Kayak website or mobile app to continue the search and complete the booking.

Financial Services

  • Toshka Bank, of the Otkritie Financial Group (Russia’s largest private banking group), launched the world’s first financial services Facebook bot, last year. The bot lets its customer check their accounts, find nearby ATMs using geolocation, call the bank, contact customer support and make payments via Facebook messenger. Tochka’s Facebook bot is linked to the bank’s Facebook account. To use the bot, a client opens a dialogue window to get access to the necessary features.
  • The Royal Bank of Scotland (RBS) recently installed an advanced “human” AI, known as ‘Luvo’, to help its staff answer customer queries. The bank, which trialled the scheme for two months last year, has now rolled out the bot more widely to support employees.
  • PayPal has just launched its first bot that lets users move and manage their money in a conversational and casual style. Built on the Slack messaging platform, which has five million daily active users, the bot allows people to send money between PayPal accounts without leaving their Slack conversation.
  • Matter, a consultancy firm, has built a pensions advisor chatbot called Matilda which can have a natural language conversation. It can advise clients on pension options and users can chat with it like they would an actual adviser.
  • AJ Bell, a UK fund supermarket, became the first DIY investment platform in the world to complete a live share trade using Facebook Messenger last year, when it purchased £500 worth of Facebook shares. It is planning to launch a new trading service that allows its customers to buy and sell shares on Facebook Messenger later this year.

Retail

  • Sephora, the French cosmetic store, launched a chatbot on messaging app Kik last year. Kik users provide details about themselves via a quiz, then the bot offers personalised beauty tips, product recommendations and reviews. Kik users can purchase products that are referenced in chat without leaving the Kik app.

  • Victoria’s Secret is using chatbots in the messaging app Kik to offer fashion and beauty advice, allowing users to receive outfit recommendations and browse new styles in an artificial intelligence-enhanced experience.

  • Nordstrom launched its first chatbot on the Facebook Messenger and Kik messaging services in December 2016. The chatbot helps users select gift items by asking them a series of questions about who they are shopping for. It then chooses gifts from Nordstrom’s online store that match their answers.

ORM’s view

Voice recognition technology is becoming more accurate, and with machine learning it is becoming better at understanding the meaning behind the words and phrases we use. The current accuracy rate for auditory or linguistic AI is around 93%. Experts believe once this figure reaches 99% it will be a game changer, and will trigger mass adoption of this technology.

But for now, businesses shouldn’t jump on the bot bandwagon without a plan, blueprint or clear map for how it’s going to work and add value.  They need to take time to find out what their customers want, and how the bot will be most useful to them, in the moment, in real time.

We are working with many of our clients to help them realise the benefits and efficiencies of AI technology. If you would like to know more about how to use AI in your business, reach out to us at @ORMLondon, we’d love to chat.

Jason Bell Jason Bell Solution Consulting Director Jason Bell