Source: Eze. S. E. How Chatbots Can Help You Increase Conversion (2018)


A Chatbot, at one level, is merely a computer program which responds to a query by a user using auditory or/ and textual methods.

Financial applications have started to use bots for bill payments, fund transfer and deposits. E-commerce companies use bots for personalized methods of selling to their customers. Bots reduce the wait time for order fulfillment in Restaurants and Retail industry. Enterprise applications are looking at for enabling employees to claim reimbursements or apply for leave. It is just a matter of time before bots find use-cases in every single industry vertical and business.

Van Baker, research vice president at Gartner says that — by 2020, over 50% of medium to large enterprises will have deployed product chatbots. [1]

A Juniper report predicts a cost saving of $0.70 per interaction by using chatbots which would trim business costs by $8 billion per year by 2022. [2]

Generally, there are two types of chatbots: rule-based and AI-based.

A rule-based chatbot is provided with a list of set answers for a set of queries a user might use the chatbot for. For example, if a user asks question X, the chatbot will answer Y. [3]

In a rule-based chatbot, if a question is asked in a different way or if the user asks a question which the chatbot has not been programmed to answer and puts up the standard “sorry, I did not understand you”, the experience of the chatbot will suffer and the user might not want to come back to seek the help of a chatbot for his queries.

AI-based bots, on the other hand, learn independently and can also adapt the way they respond based on previous interactions. [3] Though AI-based bots would learn to address most queries posted by a user over time, a chatbot has to be ‘’ to take the chat experience a notch higher for simulating how it behaves with a human as a convincing partner.

Conversational UI

A Conversational UI is a chatbot experience that processes language and responds in a way as if one was interacting with another human.

The next generation of interactive technology products or applications which are driven by AI, Chatbots, Personal Assistants and Voice Activated/Enabled gadgets are going to be conversational. Conversational systems — either voice-based or chat enabled — are on their way to become the primary interactive medium of choice for users replacing touch and mouse-based interactions. [4]

Conversational systems, though, should not be perceived as a new direction for interactive technical products or applications, but as merely the next level in enabling the user to easily and efficiently find information or complete his tasks. [4]

In a Conversational System, our choice of words while we respond to user’s needs will influence how people perceive the customer experiences we design for them because there would be no accompanying visual cues to serve as a guide. [5] It, therefore, makes sense to look at ‘Conversation Design’ to craft these set of customer experiences which are completely dependent on a set of words as the primary and only medium of interaction.

Conversation Design is a design language based on a human conversation to enable interaction with devices. Conversation Design is about teaching computers/ devices to be fluent in human conversation and its conventions. [6]

A chatbot cannot become a convincing conversational partner unless the people who build it understand the basic rules of how human conversations are structured. Individuals who architect such conversational experiences have come to be known as Conversation Designers.

Conversation Designers work at different user-facing touchpoints enabling interactions with applications and devices. A Conversation Designer would have to be someone with a technical understanding of writing copy to use the basic elements of conversation such as, turn-taking/ switching, threading, use of abbreviations, ability to repair broken conversations, validate user input and manage expectations, to structure a meaningful dialog. [7] Conversation Designers refer to an enormous body of existing literature in linguistics and psychology for deducing how people make meaning out of words and for making conversations work.

Strategies for Chatbots

An important characteristic of a chatbot to come across as a convincing online conversational partner is that the bot should be endowed with a personality. Without a personality, the bot can never come close to making the kind of connect with the user to solve queries in a way which are fluid and natural.

Instead of the standard ‘sorry, I did not understand you’, a conversational partner would ask for clarifying questions, rephrase queries and provide new prompts for solving errors.

The tasks/ queries which a potential chatbot attempts to address could be tested with real users to understand user expectations and how they would want to go about completing the task aided by the chatbot before actually building the chatbot.

Non-verbal responses like emojis and gifs could convey the right emotion which might otherwise get lost in interpretation could be another way to humanize a bot. Using filler words like ‘umm…’s and ‘aah…’s or using ‘colloquial slang’ would also go a long way to make that personal connection which only humans are capable of to the conversation.

Chatbots could also use different greetings instead of the boring ‘Hi’ every time in response to a chat query. Something like ‘Hey, How are you doing today?’ or even a ‘Hello, what’s new with you?’ could add another dimension and connect with the chatbot.

We need to make sure that a bot’s behavior and personality should be consistent throughout not just in one instance of the conversation but through every interaction with the end user. We don’t want a bot with multiple personalities showing up because of inconsistent responses to chat queries.


There are several pieces in the jigsaw puzzle for a Conversational experience which would not just address the end user query but would also be a delight to use.

A crucial step in the direction of designing an efficient and delightful chatbot is the awareness of what goes into building a conversational experience and the kind of people that are required to build one.

Humanizing a chatbot would make users feel that there is someone actually listening to them and responding appropriately — instead of just reading or relaying information. Humanizing a bot would also make the experience of a chatbot more personal and natural than detached and pretentious. [8]

One thing is clear though — a chatbot is only as good as how it is programmed to respond. If AI is involved, then a chatbot can learn and improve over time. It takes a conscious effort to humanize a chatbot by designing Conversational so that a chatbot can become a convincing conversational partner.


[1]Goasduff, L Chatbots Will Appeal to Modern Workers (2018)

[2]Gilchrist. K.Chatbots expected to cut business costs by $8 billion by 2022 (2017)

[3]Stephanie. 8 Myths About Chatbots & Artificial Intelligence (2018)

[4]Junaid. A. A. A Technique to Obtain Feedback for Designing Conversational UIs (2018)

[5] Amunwa, J. The UX of Voice: The Invisible Interface (Retrieved July 2018)

[6]Google Inc. Google Conversation Design

[7] Junaid. A. A. The Unicorn Skill in User Experience Design: Three Perspectives (2018)

Source link—-819cc2aaeee0—4


Please enter your comment!
Please enter your name here