So many people around you are busy implementing chat bots. Your bank is implanting it for customer support, your tech vendors are implementing it for problem solving and handle how-to-do queries. Your colleagues are talking about stories on how so and so bank or an ecommerce company is planning to save millions of dollars by deploying chat bots.
Essentially, chat bots are software programs that help an organization converse with the customers, service providers, employee etc. Chat bots typically have access to the organization’s database (customer profiles, transactions, knowledge documents), have ability to read and understand text as well as decipher voice, interpret what the user wants and then respond to the user appropriately. This basic ability leverages NLP (Natural language processing) algorithms to interpret the user intent, map it to a fulfilment resource, get the required information and provide it to the user. They can do most of the work that a typical customer service rep or frontline (L1) problem solver can do.
Key ingredients for chat bots are NLP, voice recognition and access to org databases.
First, NLP has been maturing over last few decades in developing its building blocks through research and in recent years some of the large players such as Google (AI), Microsoft (LUIS), Amazon (Lex), IBM (Watson) etc. have invested in binding & wrapping up the building blocks of NLP and making them available to application software developers as a platform to build applications rapidly.
Second, most corporates have implemented technologies ESB (Enterprise service bus), SOA (service oriented architecture), micro services etc. that essentially plumb the disparate applications and databases in the organization to provide data and information that can be served to the user quickly on various contexts.
Third, voice recognition too has matured in last 2-3 years and is able to fairly well decipher various accents. These 3 elements maturing and converging have thrown up countless opportunities to automate some of the mundane corporate tasks that till now need human intelligence to navigate through systems and communicate with users. Apart from these key elements, there are a few generic elements such as availability of internet, smartphone proliferation etc. have only come up to make the business case for chat bots stronger. There are few other interesting social factors that have helped and I will talk about them shortly.
First factor is instant availability and free form of query. Typically there is a few minutes hold time when you call up customer service, it involves carefully typing in the required menu option number (what you want always comes fairly late on the menu or not listed at all!), carefully typing in long digits of account numbers etc. You make one mistake and you feel like you are playing the snakes and ladder game and being unlucky! Start all over again! Also you need to be in fairly quiet place to be able to understand what the call centre rep is saying apart from being armed with data that you can never memorise! The calling technology was a boon over making a trip to the service provider’s office but times have changed and it is history now. In contrast chat bot is instantly available, usually knows who you are and you can freely type what you want the way you want. No anxiety of miss typing and playing snakes and ladders!
Second factor is the ability to control the pace of the communication. While chatting, you can multi task; you can put the chat window on hold for fairly long time and do other things (even like board a bus, pay at the teller, take a phone call, answer a few questions for your child, anything… ). This can’t be done on phone calls with customer care. The chat empowers you to do the task immediately instead of remembering to do it later in a quiet environment. Amazing!
Third factor is beating app fatigue. Availability of your service provider on popular messaging apps like WhatsApp, FB messenger, Twitter etc. helps you avoid downloading yet another app on your phone. For example money transfer service providers such as Western Union & TransferWise have integrated with Facebook messenger. Companies like MasterCard, Staples, Wall Street Journal, Pizza hut and many have a channel to provide services through Facebook Messenger and do not require the user to download their apps.
Few companies for example ICICI bank, are deploying chat bots only during the day time!, moreover they do not seem to have a mechanism to transfer the chat to a human if the dialogue is going nowhere. Some of the implementations appear to be very primitive that just provide may be the status of your transactions and nothing beyond that. Some are very slow like MoneyGram where the response is in minutes than in seconds. A few of them like Pizza hut, force login through host account than seamless integration through Facebook. Hope companies get over these initial hurdles and quickly build interdependent and seamless solutions that benefits the customers.
In the near future, many of the businesses would adopt chat bots as their primary channel of communication with customers. Lot of website and apps on devices would cease to exist. The generic chat bots would be like to the google search window used today with high degree of personalization. Chat bots would tend to be like a companion or a buddy that would keep track of all your business/ personal tasks and help you take care of lot of tasks automatically as you wish them to happen.
Chat bots are being humanized rapidly to induce humour in the conversation to start with and then the ability to decipher your emotions and deal with it appropriately. They would guess what you want at the moment and help you get it or do it much more quickly. They would also help remind you of your priorities and guide you to get back on the track in case you are slipping away from your goals! While businesses will give a big push to chat bot implementations and drive everything towards them (as it saves a hell lot of money as well as gives insights into customers) it would be challenging to handle a sizeable chunk of customers who would still find it difficult/uninteresting to talk to a machine and would want to interact with humans.
As the technology gallops, it will be very interesting to see the social impact the chat bots will have on the mankind. You never know, you might need a bot to talk to other people – scary!