Using Chatbots to Enrich Client Experiences...However, not all chatbots are created equal. There is a fine line separating chatbots that provide good or bad customer experiences. More

AN EXL WHITE PAPER

Using Chatbots to Enrich Client Experiences

Baljinder SinghSenior Vice President and Chief Information Officer, [email protected]

Written by:

2 © 2016 ExlService Holdings, Inc.2 © 2016 ExlService Holdings, Inc.

However, not all chatbots are created

equal. There is a fine line separating

chatbots that provide good or bad

customer experiences.

More Messaging, More ChatbotsThe rise of chatbots coincides with

the surge in growth of messaging

applications. Both have become part of

advertising, commerce and customer

support ecosystems. Many companies

have established a presence on social

platforms such as Facebook, Twitter and

Instagram, all of which offer messaging

functionality.

In addition several companies have rolled

out chat as a channel that customers can

use to interact with them. The exploding

popularity of messaging apps and chat

creates an opportunity for businesses to

interact with their customers in relevant,

contextual and conversational ways.

Companies need several thousand chat

agents to handle this increase in volumes.

Customers desire an interactive platform

where they can engage and resolve

issues. Companies can deploy chatbots

to interact with customers using natural

language to meet this need without

requiring a person.

Chatbots use artificial intelligence to simulate human conversation when interacting with a person.

Companies can use chatbots to help provide customer support, answer questions or other basic tasks.

This can free up employees to help customers with more complex questions or those that that require

nuance and creativity.

Messaging Apps Have Surpassed Social NetworksMonthly active users for top 4 social networks and messaging apps

–– Big 4 Messaging Apps–– Big 4 Social Networking Apps

BI INTELLIGENCE

Mill

ion

s

– 3,500

– 3,000

– 2,500

– 2,000

– 1,500

– 1,000

– 5,00

0

Source: Companies, BI Intelligence

40.1

1 –

10.1

2 –

20.1

2 –

30.1

2 –

40.1

2 –

10.1

3 –

20.1

4 –

30.1

3 –

40.1

3 –

10.1

4 –

20.1

4 –

30.1

4 –

40.1

4 –

10.1

5 –

20.1

5 –

30.1

5 –


EXLservice.com | 2

3 © 2016 ExlService Holdings, Inc.

Automating and Enriching Customer Interactions with ChatbotsChatbot technology has advanced a long

way since Joseph Weizenbaum developed

his natural language processing program

ELIZA in the 1960s. Building these bots

used to be complicated, resource-intensive

projects. Advancements in artificial

intelligence (AI) allow developers to quickly

build and deploy chatbots that present a

reliable alternative to human support.

Chatbots have been successfully used

across various industries:

• Travel: Helps book trips and delivers their

boarding passes

• Healthcare: Assists with identifying

illnesses and providing healthcare advice

• Retail: Customizes shopping experiences

• Banking: Provides financial and money

management advice

Some can also function as a personal

assistant. Siri, Cortana and Google Home

can answer questions, set appointments

and buy things online. As interacting

with these bots becomes an everyday

occurrence, customers have grown more

accepting towards their use.

One example is a Facebook Messenger

chatbot introduced by KLM Royal Dutch

Airlines‎. Customers can give the bot

permission to automatically message them

regarding their boarding pass, check-in

confirmation and delay notifications. It can

also respond to basic traveler questions.

If the chatbot is unable to answer a

question, it connects the passenger to a

person. This bot integrates directly with

the Facebook Messenger app, eliminating

the requirement for users to download a

separate app.

When Chatbots Ruin User Experiences Poorly designed chatbots can lead to

detrimental, and sometimes absurd,


Consider Poncho, a chatbot designed to

provide weather updates over Facebook

Messenger. While its creators intended

Poncho to be “the first bot you want to be

friends with,” early interactions with the

bot were frustrating. Users found Poncho

unable to answer even simple questions or

required questions to be rephrased multiple

times. The bot required users to construct

questions in unnatural ways to understand.

Phrasing a question as “Precipitation San

Francisco Tomorrow” rather than “Will it rain

in San Francisco tomorrow?” didn’t create

the personal touch Poncho’s designers

hoped for.


EXLservice.com | 3


The conversational hiccups of Facebook’s

chatbot have been mostly ironed out

at this point. However, they show that

organizations must carefully consider their

chatbot’s design to avoid creating negative


Designing a chatbot should be part of

the organization’s digital transformation

roadmap.

An effective, complex chatbot should be

tied to a company’s digital capabilities.

The customer journey map should be

looked at in a holistic manner, focusing

on key customer interactions that provide

“moments of truth” to ascertain the right

type of mechanism to communicate and

engage with them. Further, as shown in the

KLM example, the digital interface (chat)

should be complemented with the right

human touch to provide compelling and

seamless engagement.

Initial experiences with chatbots could focus

on piloting rule-based bots that answer

simple customer queries. Companies with

more advanced digital capabilities, and

those provisioned to support decisions

based on large volumes of data can

create bots that adjust and improve their

interactions by using machine learning to

learn from the data.

Best Practices for Building ChatbotsAutomation is increasingly a part of

everyday life, from self-checkout lines in

grocery stores to touchscreen kiosks in

restaurants. It’s effective when it makes

the experience seamless to the customer.

Designing a good chatbot may seem

daunting, but a few best practices can

simplify the process:

Never Overextend a ChatbotWhatever function a chatbot is designed to

perform, it should perform well. A chatbot

that does a dozen things poorly will be less

useful to customers than one that does

three things perfectly.


EXLservice.com | 4


Make it snappy Don’t keep customers waiting for answers. A

good rule of thumb for chatbots is that they

should provide a customer with information

they’re looking for within 2-3 questions.

Add Emotions with AvatarsUsers don’t prefer interacting with a

faceless chatbot. Consider adding an avatar

to a chatbot to produce a more human

look and feel, which in turn can drive a

more empathetic response from its users. A

chatbot’s answers should also be written in

a conversational manner.

Provide Customers Ample OptionsNot everyone will want to speak with a

chatbot. Forcing customers to interact with

a machine when they wish to speak with a

human is a recipe for frustration. Provide

customers with options that allow them to

connect with a human agent immediately.

Making Chatbots WorkChatbots are becoming better, faster and

cheaper. More companies than ever can

provide cost-effective customer service,

rapid response times and simplified

business transactions using these bots.

However, these benefits will only be

achieved if a company has the digital

strategy and plan to integrate the bot into

their existing operations. These bots require

automation to rapidly pull information

requested by customers and machine

learning to cut down on the amount of

exceptions requiring human intervention.

By approaching chatbots as part of an

overall digital transformation agenda,

businesses will be well positioned to

provide enhanced customer experiences.


EXLservice.com | 5

GLOBAL HEADQUARTERS280 Park Avenue, 38th Floor, New York, NY 10017

T: +1.212.277.7100 • F: +1.212.277.7111

United States • United Kingdom • Czech Republic • Romania • Bulgaria • India • Philippines • Colombia • South Africa

Email us: [email protected] On the web: EXLservice.com

© 2017 ExlService Holdings, Inc. All Rights Reserved.

For more information, see www.exlservice.com/legal-disclaimer

EXL (NASDAQ: EXLS) is a leading operations management and analytics company that designs and enables

agile, customer-centric operating models to help clients improve their revenue growth and profitability. Our

delivery model provides market-leading business outcomes using EXL’s proprietary Business EXLerator

Framework®, cutting-edge analytics, digital transformation and domain expertise. At EXL, we look deeper to

help companies improve global operations, enhance data-driven insights, increase customer satisfaction,

and manage risk and compliance. EXL serves the insurance, healthcare, banking and financial services,

utilities, travel, transportation and logistics industries. Headquartered in New York, New York, EXL has

more than 26,000 professionals in locations throughout the United States, Europe, Asia (primarily India and

Philippines), South America, Australia and South Africa.

Using Chatbots to Enrich Client Experiences...However, not all chatbots are created equal. There is a fine line separating chatbots that provide good or bad customer experiences. More

Documents