Chatting up a girl or understanding chatbots College days once again in business?
The current crop of commercialised chatbots has not been optimally designed keeping in mind the functions they are to perform.
Numerous Bollywood songs have been dedicated to that ‘new girl’ in college whom we all wanted to walk up to and talk to but were always unsure as to how to go about it. Numerous bets plans with friends and failed attempts later, she still remained an enigma. The hopeful songs describing budding love were then replaced by sad songs of the 1960s, describing heartbreak and unrequited love, Guru Dutt style.
Cut to 2017. Most of us are well settled in our jobs, and the college days serve only as a pool of memories to dive into during occasional gatherings. But, the problem of not being able to understand how to approach a ‘new’ situation still persists. As is the case with Artificial intelligence and chatbots, for instance. Aakrit Vaish- CEO, Co-founder, Haptik, feels that the latest buzzword in the commercial space, AI is being considered as the ultimate elixir to all business problems, thanks to a romanticized image created in Hollywood flicks such as I, Robot, Her and several others. But the fundamental problem lies in the fact that without having a proper understanding of what these chatbots are all about, entrepreneurs are just rushing into adopting the technology, without actually analyzing it, and till what extent, their business requires its incorporation.
What is AI?
Essentially, AI refers to the development of cognitive intelligence in a machine that allows it to perceive its environment and choose actions to attain particular goals. However, unlike the image that popular culture might project of an AI developing conscious knowledge, it is in principle a programmed machine that has a highly adaptive programming which enhances its learning capacity, thereby updating its responses on a real-time basis as and when exposed to different stimuli.
Where do chatbots fit in?
Chatbots represent one application of AI. There are various applications of AI such as robotics, deep learning, neural networks, etc. Chatbots is just one of them. Unlike an average entrepreneur’s expectations of chatbots being able to take care of everything their customers ask for, the average customer has a confusing experience talking to bots which have not been well calibrated to the frequently asked questions regarding the particular product or service in question.
The ‘complex’ chatbot myth
One of the greatest myths regarding chatbots is that they are very adept at solving complex problems and dealing with open-ended situations. AIs cannot generate solutions by running logic functions and loops faster than the speed of sound. In fact, they are best only for solving simple, structured problems for which it would be an overkill to create dedicated apps or websites. Examples of such simple scenarios can be:-
1. Updating an insurance claim status
2. Setting a reminder
3. Location of a center/shop/ATM
4. Medical report status
5. Status of order
6. Credit card lead generation
It has been noted that 80 per cent customer queries for any business pertaining to a particular industry are from the same 20 per cent core group of questions, which are in effect the introductory ones. For instance, for a travel organization, the oft-repeated queries would be around the destinations covered, the duration, price per individual, itinerary etc. A chatbot should ideally be assigned to provide answers to such questions and once a customer is satisfied with the responses and wishes to proceed, an agent can be brought forth to answer specific queries and solve more complex concerns. This helps to reduce labour costs, reduce the amount of time taken, builds professionalism and improves quality of customer engagement, as agents are not burdened with rudimentary tasks.
Lead generation is another such instance where chatbots can be of great help. It has been noted that the typical form filling conversion on a website is 1 per cent. While on chatbots, the rate can go as high as 20 per cent as people find it easier to reveal information in a conversational format, giving one piece of information at a time. It is much easier than to go through a long and drab list filling mundane details.
It is important for industry leaders and entrepreneurs to realise the present scope of AI and chatbots and not harbour any unrealistic expectations. While AI is making giant strides every day, and into every core human function such as HR, it is still a programming-oriented, learning process that is yet to develop the ability to make on-the-spot choices, account for the emotional quotient or provide reassurance while tackling a particular situation. The current crop of commercialised chatbots has not been optimally designed keeping in mind the functions they are to perform. If they are designed keeping the specific tasks in consideration, chatbots can really perform a given task with high efficiency and accuracy, given their ability to rectify and update themselves on the process. This makes them a highly valued tool for customer engagement in the future, and unlike the new college girl in the example above, not entirely unpredictable.