Introduction to Chatbots

The evolution of artificial intelligence-based technology is evolving today with a plethora of services being offered in different domains.

This decade saw the rise of many chat applications like WeChat, Line, Viber, Slack, Signal, and lot many. Well, how many of you have used the mentioned chat applications?

I have used all of them out of curiosity.

With the spread of messengers, virtual assistants that imitate human conversations for solving various tasks have become increasingly in demand. Chinese WeChat bots can already do tasks like setting medical appointments, calling a taxi, sending money to friends, check-in for a flight.

Gartner forecasts that by 2020, over 85% of customer interactions will be handled without a human.

It’s the 2020s decade and we can observe the scenario related to chatbot has been changing rapidly.

The opportunities provided by chatbot systems go far beyond giving responses to customers’ inquiries. There are other business tasks where chatbots are being used extensively like collecting information about users, helping to organize meetings, and reducing overhead costs. No wonder the size of the chatbot market is growing exponentially. Chatbots have become very important for all businesses because they are fast enough to resolve customers’ queries which in turn helps businesses address more customer queries.

Google’s Meena has created a lot of buzz on how these chatbots are becoming increasingly used. The fun fact is probably by end of 2020, lot more companies will be investing in chatbot and you would never know whether you are talking to a human or a bot 😛

So, what do you think is a chatbot? Is it a computer program? a service? a system ? or just a module?
According to Oxford Dictionary, a chatbot is

“A computer program designed to simulate conversation with human users, especially over the Internet.”

But I would put it rather in a practical way being a chatbot developer,

“A chatbot is a service which is powered by rules and sometimes AI, that we interact through UI”

In this article, I would like to cover an interesting case study of Swiggy. For Indians, Swiggy is one of the most popular platforms. For those outsides of India, I would say Swiggy is India’s largest and most valuable online food ordering and delivery platform.

An interesting short story of this unicorn startup for you!

Founded in 2014, in an office space in Koramangala, Bangalore, 1 neighborhood,6 delivery executives, and 25 partner restaurants are what Swiggy started with. Growing at 25 percent month-on-month, almost 4 years into the field, Swiggy now has a major presence in Delhi, Mumbai, Pune, Bangalore, Hyderabad, Chennai, and Kolkata.

They have partnered with 12,000 restaurants and over 13,000 delivery executives. According to the RoC data filed by the company, the net worth of the company is Rs 3,86,34,590 with a turnover of Rs 7,41,702 to be precise.

Aren’t these stats interesting?

At Swiggy, chat is one of the primary contact media for the customers to reach out to. Approximately 70% of the customers prefer to communicate via chat than on call.

How Swiggy got its third-party customer support system converted into an automated process?

In the year 2018, Swiggy became so popular that it expanded from 9 to 200+ cities and this resulted in approx. 10x increase in the number of conversations daily. As it is rightly said, “Necessity is the mother of Invention”, third-party software used by Swiggy wasn’t enough as Swiggy was planning to extract customer satisfaction scores, rate of order cancellation, etc. For the customers, the chat experience was not at par with Swiggy’s s highly recognized UX. Swiggy was keen on interactive messaging components and monitor customer interactions with each of these components.

The support executives were also not very thrilled with the increased traffic. They were solving similar repetitive complaints every time — whether it was order cancellation or updating the customers about their order status. That is when they started exploring ideas related to chatbots.
Swiggy started with a controlled flow within conversations i.e. customers wouldn’t be required to type in any text but are provided with options to choose from and proceed to the next step. This was actually “Decision Tree” working behind the scenes.

Over a period, Swiggy was able to gather some common reasons for order cancellation and created a chatbot flow which is something like:

  • Ask customer reason for cancellation
  • Predicted time delivery as deterrence
  • Show cancellation fee
  • Ask the customer for confirmation on cancellation
  • Details about refund using rich visualization.

This workflow has increased the speed of the cancellation process by 96.3%. The cancellation process which took around an average of 5 minutes, now can be done in 10 seconds.

This is probably one of the classical examples of how chatbots help businesses in solving customer queries faster and more efficiently. These chatbots are not only helping business in customer queries but it has become an important tool even in this critical times of Covid-19 by providing leads for Oxygen, Beds, Medicines. Of course, I do not require to recall which chatbots I am referencing here.

Click here to read more blogs.

Follow on LinkedIN

Leave a comment

Verified by ExactMetrics