By using this web site you accept our use of cookies.

Chatbots — GenNext Customer Engagement

Chatbot suggesting sushi

Technology is changing at lightning speed. In the era of digitalization, the importance of conventional business functions and tools such as telesales, tech support and informative websites is diminishing. Many retailers’ survival is at stake due to digitalization.

This era of digitalization is demanding a new phase of retail called chat commerce where bots interact with end customers to drive business using audio, text and emoji signals. Bots have picked up the pace as a digital trend and it will continue to penetrate the retail digital market.

Chatbots are classified into two categories:

  • Rule-based Chatbots: Bots are built based on some set of rules or tasks to be performed that are limited and repetitive e.g. interactive voice response, tracking responses, grievance / complaint registration and tracking.
  • Artificial Intelligence (AI)-based Chatbot: Complex bots that are trained to take intelligent decisions based on users’ inputs. These bots will show better performance over time as they make use of historical data powered by Natural Language Processing (NLP) and Machine Learning (ML).These bots are technologically challenging and complex to build and train. However, they are estimated to provide efficient process handling and quicker ROI.

Technical functioning of a chatbot

Technical functioning of a chatbotFigure 1: Schematic representation of a chatbot

In technical terms, bots are web service calls which communicate via interfaces or channels such as chat window, Skype, Facebook, VoIP, etc. A chatbot functions as follows (refer Figure 1):

  • The user sends a request in terms of e-mail, chat, image, emoji or audio on interfaces or channels.
  • A web service call is made with the help of a connector.
  • A connector establishes a connection between the business application and the interactive channel.
  • The translation and interpretation of human language input are performed by a combination of NLP (Natural Language Processing) and ML (Machine Learning) at the Application level.
  • In AI-based chatbots, the business application is powered by an artificially intelligent program either custom built or built using existing frameworks such as LUIS, or
  • The response is sent back to the user via the requested channel.

Why chatbots?

Chatbots offer commercial and non-commercial benefits such as:

  • save time, labor and money for tasks which can be easily automated
  • help in manpower reduction, especially in customer-facing processes
  • enhance ROI by keeping the operating costs low
  • reduce the difference between “conversational commerce” and online shopping
  • enables an interactive and simple shopping experience

What can chatbots do in the retail segment?

To enable interactive commerce, chatbots are used as virtual assistants or support agents. They provide:

  • personalized assistance for enhancing customer service
  • product recommendations
  • order confirmation
  • updates on order and shipment enquiries
  • help in managing social media pages for increasing customer engagement rates
  • an interactive product user manual
  • notifications by sharing / pushing updates and advertising new products
  • product location maps and assist in locating a product in the store

Implementations of chatbots across Industries

Many companies are using chatbots to converse with their customers for assistance in various forms:

  • H&M’s chatbot on Kik acts as a personal stylist to customers by helping them choose the right H&M outfit based on their preferences and purchase histories. The overall customer experience is benchmarked at a higher level due to the seamless interaction with their bot.
  • Tacobot, an order taking bot by Taco Bell, allows customers to order food and track their deliveries. Tacobot also answers enquiries on food and provides customer specific recommendations. Customers fondly describe Tacobot as a witty and fun bot.
  • Uber integrated the Facebook Messenger chatbot with which customers can book a cab.
  • Mastercard helps customers with details on “account balance”, “purchase history” and “recommend card offers” using a recently adopted bot.

Why chatbots are not yet popular?

Chatbots are not widely used for a number of reasons:

  • If the efficiency of AI is unstable, it leads to confusion and may lead to loss of business.
  • New strategies must be adopted to enable users to engage with the bots more frequently.
  • Although the chatbots are proven ways of improved user engagement, not many companies use them as this technology lacks media attention.
  • Companies are reluctant to experiment with chatbots as there is no major breakthrough in the market yet.

Chatbot is still a "work in progress"

Be it the stability of existing technologies, job market conditions in the IT industry, the rush towards process automation or openness of industries to adopt bots over humans, the new wave of commerce has begun and will continue to take over the market, though at a slower pace.  

So next time during a chat with Customer Support Team, try not to lose your calm as chatbots do not understand emotions (until they are programmed to).
 Mohammed Saqib 
Senior Software Engineer with over four years of experience in the retail sector.
Copyright © 2017 by EVRY India.
EVRY India acknowledges the proprietary rights of the trademarks and product names of other companies mentioned in this document.

Contact us to learn more