CHATBOTS

An Introduction And Easy Guide

To Making Your Own

Oisin Muldowney
First published 2017 by Curses & Magic, Dublin, Ireland
Text © Curses & Magic 2017

All rights reserved. No part of this publication may be reproduced

or utilised in any form or by any means, electronic or mechanical,
including photocopying, recording, scanning, or in any information
storage and retrieval system, without permission in writing from
the publisher ([email protected]).

Dewey: 004.019
ISBN: 978-1-9998348-0-7
Contents
Introduction

Part 1 Explaining Chatbots

01 What is a Chatbot? 5
02 A Brief History of Chatbots 6
03 The Future 9
04 Machine Learning, Natural Language Processing, and 14
Artificial Intelligence
05 Customer Service 17
06 E-commerce 20

07 Chatbots vs. Apps 30

08 Banking 33
09 Health Care 37
10 Libraries and Archives 41
11 Relationships 44

Part 2 Building Your Chatbot

12 Getting Started on the SnatchBot Platform 50

13 Using a Template from the Bot Store 57

14 Placing your Chatbot on a Website, Skype, Facebook 59

Messenger and other channels
15 Extracting Emails, Urls, Addresses and Other Data 63
16 Handling Payments 66

Conclusion
Introduction

I firmly believe we are on a cusp of a chatbot revolution that will be extremely

important to human culture. Not quite as deep a change as the development
of the internet, perhaps, the massive deployment of chatbots will certainly
be more profound a change to our lives than the introduction of apps to our
devices. And in the longer term, we might look back at this phase as creating
the essential tools for the emergence of a true artificial intelligence.

Chatbots are already everywhere. As some of the chapters in this book detail,
they are present in e-commerce, banking, health care, education and libraries.
But still, the really big businesses of the world are only now getting up to
speed with the importance of chatbots. This book is not aimed at them.
They have large budgets and significant IT support staff with whom to
develop chatbots.

Small businesses, however, along with individuals, can also benefit from using
chatbots. And that is what has motivated me to produce this book. In 2017,
we reached a point where anyone, with no coding skills whatsoever, could
create a chatbot. I want to encourage all interested readers to do so.

This book consists of two sections. First of all, I run through the history of
chatbots, some thoughts about the future and instances of how chatbots are
changing cultural activity in all sorts of spheres. The more practical minded
reader might be impatient to get on to Part 2: Building Your Chatbot. But
I would encourage you to at least dip into some of the earlier material.
Naturally, I’m biased, but I think these chapters are both stimulating and fun.

I owe thanks to several people, especially Joe Crawford, Julian Howard and
Chris Knight for their work on the manuscript. Also to Avi Ben Ezra and Henri
Ben Ezra, the founders of SnatchBot. I met Avi and Henri at a Chatbot Summit
and was impressed by their demonstration. In this book I make heavy use of
the SnatchBot platform, because not only does it allow you to make your bot
for free, it’s very intuitive. Thanks both for your patience in responding to what
must have seemed like an endless series of emails.

1
Part 1 Explaining Chatbots
Chapter1: What is a chatbot?

A chatbot is a computer program written to participate in a conversation.

Typically, chatbots are written to interact with humans (rather than other
chatbots) and they do so for an extremely wide variety of reasons. In busi-
ness they are proliferating as an alternative to websites: instead of a cus-
tomer having to take the initiative by searching through website pages, the
chatbot provides the customer an interactive guide, which can orientate
them towards the product they are seeking and even arrange payment and
shipping. Similarly, organisations that provide a great amount of online infor-
mation for clients (such as healthcare organisations or government bodies)
use chatbots to help clients get the information they want via a conversation
rather than a search engine. The particular advantage of this for the client is
if the client isn’t sure what terms to search for, or where the search terms
being used are too common, he or she can become frustrated with the waste
of time navigating the website via their own, unaided efforts.

Chatbots exist for hundreds of other reasons, including just for the fun of the
conversation. There is a website with a chatbot, Mitsuku, that claims, ‘you
need never feel lonely again.’ And while Mitsuku is primarily there for the
enjoyment of chatting to her, there is a serious side to this claim. There is
scholarly evidence for the fact that any kind of conversation, including that
with a chatbot, is better for human wellbeing than none at all.

You can encounter chatbots on various different platforms. When you ring an
organisation, for example, and get through to recordings which you navigate
with your response, that’s a kind of voice-based chatbot. Google’s Assistant
and Apple’s Siri are also voice-based types of chatbot. As for text-based chat-
bots, you are likely to see them pop up on websites with increasing fre-
quency. But the real reason for chatbots becoming so pervasive is that
people are spending more and more time messaging each other and less
time browsing websites. Instead of leaving your platform (e.g. Facebook
Messenger) you’ll use chatbots to connect to the vast online world. And the
experience will be more helpful and dialogue driven than navigating on your
own.

2
Part 1 Explaining Chatbots
Chapter 2: A Brief History of Chatbots

Chatbots have had a long history, but now they are really coming into their own.
Photocredit: Zapp2Photo/Shutterstock

By definition, a chatbot is a computer program which conducts a conversa-

tion. Less literally, you can look at a chatbot as software that mimics the
experience of chatting with a fellow human. While the recent meteoric rise of
messaging apps has brought chatbots to prominence, they have existed in
one context or another for a long, long time.

The birth of chatbots

ELIZA, created between 1964 and 1966 by German-American computer sci-

entist Joseph Weizenbaum, is widely considered to be the first chatbot.
ELIZA gained recognition for its ability to trick humans into thinking that they
were having a conversation with another real human. Interestingly, ELIZA
was not created for any sort of commercial application. Rather, ELIZA was
built to parody ‘the responses of a non-directional psychotherapist in an
initial psychiatric interview’.
ELIZA simulated conversation by pre-setting text outputs to be triggered by
specific text inputs. If that sounds familiar, that’s because it’s the same
structure that most of today’s chatbots use. The creator anticipates user
inputs and sets up responses for the chatbot to give. Going beyond this style
of build is among the most important next step in the Artificial Intelligence
and Natural Language Processing fields.
3
Part 1 Explaining Chatbots
Next steps
In the mid 1990’s other versions of chatbots began to appear. Though it’s
different from how we perceive a chatbot today, one prominent example of
this includes Ask Jeeves (now Ask.com). Existing as a search engine, Ask
Jeeves encouraged users to input what they want to know in the form of a
question. This was a significant departure from traditional search engines
such as Google and Bing.
Rather than just respond to a slew of words, Ask Jeeves utilized Natural Lan-
guage Processing in an attempt to make searching for information more natu-
ral. Unfortunately, this approach was not successful and was ultimately
defeated by the titan search engines we use today.
Modern Times
Over the past few years, chatbots have risen to centre stage. While the growth
of messaging channels has contributed to this move, platforms opening up
and embracing chatbots has also been a primary driver of growth. With Face-
book, Microsoft and a variety of other tech giants opening their arms to chat-
bots, there’s never been a better time for the medium. A small sample of the
industries they now occupy is provided below.
Customer service
Online customer service has proven to be fertile ground for chatbots to root
down and gain traction. Many businesses and services have moved away
from using call centres and are instead tasking chatbots with answering and
directing common customer inquiries. This includes large entities including
Citroen, Royal Bank of Scotland, Renault and Lloyds Banking Group.
Chatbots provide a number of advantages over traditional human customer
service. First, they are less expensive than paying humans and require none
of the HR-related spending associated with hiring actual people. Plus, they
never call in sick. Second, they can analyze questions and provide responses
at a much more rapid pace than a human can.
Marketing
From 2017, ‘having a conversation’ with the consumer became a critical
aspect of many brands’ marketing strategies. Chatbots allow brands to inter-
pret that idea literally. The entertainment industry has been a clear first mover
in embracing chatbots for marketing purposes. A likely reason for this is a
chatbot’s ability to simulate conversations with characters, such as a popular
musician or film character. Thus there are chatbots mimicking everyone from
pop music artist Katy Perry to Spock from Star Trek.
4
Part 1 Explaining Chatbots

App replacement

As of mid-2017 app downloads have slowed dramatically. This has resulted a

struggle for many companies to find a channel in which they can deliver their
digital services to customers. Chatbots placed in popular messaging chan-
nels such as Facebook Messenger provide a solution to this gap.

Among the companies who were quick to use the technology were Uber and
Dominos Pizza. Uber users can now request, track, and pay for an Uber with-
out leaving their Messenger, Telegram, and Slack conversations. Dominos
takes a similar approach, allowing hungry customers to place their order and
monitor its progress within a range of platforms including, but not limited to,
Messenger, Echo, and Android. This type of medium is often referred to as
‘conversational commerce’. Rather than a static purchase process, users
interact and make purchases in a back-and-forth digital conversation.

Moving forward

The future has never been more ripe for chatbot success. Messaging chan-
nels are embracing chatbots and providing them with advanced technical
capabilities. Brands are increasingly more open to the advantages this tech
helps them gain. And there is a dynamic in the current situation that might
well lead to a massive leap forward for chatbots, beyond anything that apps
achieved. Chatbot development platforms like SnatchBot and Chatfuel make
it possible for anyone to create a chatbot. We are swimming in the rising
waters of a tsuami of chatbot creation and should millions of crowdsourced
chatbots be linked in a fashion that allows them to learn from their interac-
tions… well, that would be the basis of a revolution more profound even than
the internet.

5
Part 1 Explaining Chatbots
Chapter 3: The Future
Was 2017 the year of the Chatbot?

Other than the self-driving car, it’s hard to think of a technology that has cre-
ated more buzz than the chatbot. Chatbots, we are told, are set to revolution-
ize everything, but especially, e-commerce, banking, health-care and educa-
tion. Just as the app took online activity by storm and has been adopted by
everyone, the chatbot is going to do the same.

Yet we’ve been hearing this for some time.

What is the actual state of affairs? Has the chatbot revolution arrived?

One useful place to start in answering this question is with Microsoft’s Bot
Directory. Here dozens of interesting bots are featured and playing around
with a few really does make you see the possibilities. Although none of them
yet really grab me as essential, I can see the value of most of these bots,
especially those that help in organizing my time and motivating me to exer-
cise. This directory is now closed to new bots and here we get the first hint
that bot development might really be moving at a fast speed: There are too
many new bots for the directory to keep up. Another directory, far more com-
prehensive, can be found here, and again, if you want to get a taste of what’s
possible, it’s fun to play with these. But again, too, this list is already behind
the times. 6
Part 1 Explaining Chatbots
The big companies step up their chatbot activity
It’s the big players who are likely to drive forward the use of chatbots in
modern culture. And where are they on the issue? The answer is, we are now
seeing definite enthusiasm for, and commitment to, chatbot development.
I think it would be fair to say that Apple’s Siri is a voiceactivated chat bot. And
if so, then the support given to Microsoft users, Amazon users and Google
users by Cortana, Alexa and Assistant respectively, show a quantum leap
forward in this kind of software.
Text-based chat bots with a more focused role to assist client engagement
with a library, or health care organization, are spreading like fire on petrol. One
massive stimulus to this was Facebook’s decision to allow bots on Messen-
ger. This saw around 100,000 developers create 100,000 bots for the platform
in the first six months.

Millions of chatbots spring up across the world

The bot development community is blossoming exponentially. It really is like

seeing green shoots emerging from a desert after rainfall. Thus last year,
www.pandorabots.com reported that it had 225,000 developers, 285,000 chat-
bots created and three billion interactions.
I believe most of these were short-lived bots, whose purpose was primarily
commercial. But there are now long-lasting chatbots delivering excellent
results in the education sector, in assisting users of library catalogues
(see Chapter 9), in counseling and health care (Chapter 8).

Do you need a chatbot?

I’m a software engineer who does a certain amount of freelancing, so whilst

not exactly typical, I could be representative of a type of small business. Have
chatbots become useful to me? To explore this further, I went to SnatchBot.me
and created my own free chatbot.
I found it very easy to create a basic chatbot and – this is crucial – attach it
to my Facebook business page. Now, every time a potential client sends me a
message, they engage with the bot. Of course, I don’t want to alienate anyone,
so the first thing the bot does is give out my email if the reader wants to con-
tact me in person. But it then guides the conversation to steer the reader to
the services I provide that might most suit them.
At first, I made several blunders in the logic of the conversation. They were all
very easy to fix, however, and also as I saw the kinds of interactions that readers
came up with, I added on new layers to the conversation.
7
Part 1 Explaining Chatbots
Practically, my bot doesn’t represent an enormous gain compared to using
Messenger in its usual way. But I’m very happy with my bot and will certainly
keep it. The biggest plus is that I get a chance to project a certain amount of
humour and enthusiasm through the bot.

And I think for businesses, this is an under-appreciated aspect of chatbots.

Chatbots are not just tools to connect users to the information they want
(and they are much better tools for this than FAQ pages on websites), they are
an opportunity to promote your brand. If you are a bank, your chatbot will be
sombre, accurate, polite. If you are a health care organisation, your bot will
be sympathetic. If you are an entertainment organisation, your bot will be
lively, funny, cheeky even.

So yes, 2017 will be seen as the year of the Chatbot. Not just because the
large companies started using them, but because that was the year it became
truly simple to create your own.
The Future of Chatbots: An Interview With Avi Ben Ezra of SnatchBot.me

Can the crowdsourcing of chatbots lead to AI? Picture: Zapp2Photo/Shutterstock.com

8
Part 1 Explaining Chatbots
Back in the 1980s, I took a philosophy exam and answered a question about
whether artificial intelligence was possible or not. Blade Runner had just
come out and as I loved it, I answered ‘yes’ by writing an imaginary conversa-
tion between a human volunteer and a program speaking over the phone. I
didn’t know the word then, but I was writing about a ‘chatbot’.
In 2017, I had the opportunity to talk to Avi Ben Ezra, the Chief Technology
Officer of SnatchBot. Founded in January of 2015 with the goal of making
bot-building easy and accessible, SnatchBot is a fast-growing Israeli com-
pany. Avi is the architect of the platform and the user interfaces.
I brushed up on my interest in the subject of chatbots and asked Avi about his
vision of the future.

What can you tell us about the future role of bots?

There is going to be an exponential growth in the role of chatbots over the

next decade. Already, given the current state of the market and the speed at
which it is expected to grow, we can see how rapidly the use of bots is ad-
vancing: a recent chatbot report released by BI Insider revealed over 80% of
businesses are expected to have implemented some sort of chatbot solution
by 2020. As the market pushes the technology forward, interactions with
chatbots will become more and more sophisticated.

Is this a path to true artificial intelligence?

I think so, but let’s split this question into two parts: what we can be sure of
and what we can speculate about.
We can be sure that more messaging APIs are opening and as they do there
will be growth in the number of channels supporting chatbots. For instance,
WhatsApp is joining the fun. WhatsApp is the number one messaging plat-
form in the world, yet so far, no one is allowed to build chatbots for it. We
know Facebook (their parent company) is working on this. Expect news very
soon.
Big brands are investing in chatbots, streamlining some of their processes or
simply turning their brand into an approachable conversational experience.
Consumers are more willing to engage with chatbots, providing the chatbots
are entertaining and providing relevant information. Put this all together, and
we see a rapid evolution in chatbot-human interactions. Already, it’s possible
to build chatbots that respond to emotional content (whether the person
sounds cheerful or happy, agreeable or discontented) and tailor the chatbot
response accordingly.
9
Part 1 Explaining Chatbots

This combination of powerful market forces and increasing ease of chatbot

building makes me certain that they will soon achieve a certain level of what
you might call ‘artificial intelligence’.

Now, to the speculative part. There is a huge discussion about what makes
for consciousness, and my belief is that it will be possible to create fully sen-
tient software. Some physicists believe our universe is a model and so we are
sentient software. Having said that, there is a major tipping point to reach
before we can talk about true AI. The chatbot has to be able to learn and it
has to be immensely more complex. It’s well known that the brain has more
potential pathways than there are atoms in the universe. Currently, chatbot
pathways are extremely crude in comparison.

But AI could arise out of chatbots?

It could. I think so. There are other paths to AI, of course, but those based on
pure research don’t have the same momentum as chatbots currently do. What
we are trying to do with SnatchBot, for example, is a kind of crowdsourcing of
the creation of chatbots. If thousands, millions even, of chatbots are being
created and linked up you really are harnessing the kind of intellectual energy
that leads to technological revolutions.

And that, to me, is a really important point. Human language and conscious-
ness evolved together over millennia. We can accelerate that process dra-
matically for AI and language, especially if we can contribute to the process in
our millions. Everyone should have a bot and perhaps it’s not too long before
everyone will.

10
Part 1 Explaining Chatbots
Chapter 4: Machine Learning, Natural Language Processing and Artificial
Intelligence

Rapid strides are being made in Machine Learning and AI, which are crucial for successful chatbots.
Photocredit: Zapp2Photo/Shutterstock

Machine Learning is the label given to algorithms that allow a machine to

take feedback from data and adjusts it processes. There is a parallel in this to
human learning in that often people do learn from trial and error, but of course
the human mind can make leaps of understanding that are entirely absent
from the number crunching iterations and subsequent adjustments that go
under the name ‘Machine Learning’.

Machine Learning is a technique that allows machines to improve their inter-

actions with human language. The process by which meanings expressed in
human language are broken down to give information to machines is defined
as Natural Language Processing (NLP) and Machine Learning is an essential
feature of NLP, as the machine attempts to successfully respond to the
human phrase through repeated trials.

As with the term ‘learning’ in Machine Learning, ‘intelligence’ in Artificial Intel-

ligence (AI) is a far shallower concept than the human version. For machines,
AI currently means the ability to make decisions based on past experience.
This is a concept that is closely related to Machine Learning, but the
decision-making power of the AI is usually the starting point, rather than the
one arrived at via trial and error. 11
Part 1 Explaining Chatbots
What is the difference between AI and Machine Learning?

Suppose we need a machine to count layers in ice cores and we want it to

deal intelligently with the challenge of borderline calls. Is the faint change in
colour a genuinely new layer? Or a subtle aberration in the current layer? The
decision is obviously important to scientists wanting very accurate chrono-
logical data from the ice.
There are two approaches to creating this machine. The AI approach is to
program the machine with the skills of a human expert. The human expert
would assist in the creation of algorithms to address all the difficult calls and
the explanation of why the expert reaches certain decisions would be used in
the design. The strength of the AI would therefore be dependent on the skills
of the expert.
The other approach would be the machine learning method. Using existing
ice-cores that have been securely dated, the machine would have no
decision-making tools, but it would try and try again. Each time it makes an
error, the machine’s criteria for identifying a new year in the ice are adjusted
appropriately and the iteration is run again. By the time the machine can
derive flawless year counts in several ice-cores, its users will have a lot of
confidence in giving it a previously uncounted ice-core.
Chatbots, Natural Language Processing, Machine Learning and AI

All three of these concepts come together in the world of chatbots, because
the chatbot is a machine that responds to human language and tries to do so
intelligently. Clearly, chatbots need Natural Language Processing. But it is
less clear whether they should be constructed with AI or Machine Learning or
both. To some extent the answer depends on the scope of the task for which
a chatbot has been designed. Many chatbots will have a relatively narrow
purpose, let’s say to answer questions from prospective students about
course content. For this task, a machine learning approach would seem ap-
propriate.
No knowledge of the actual college courses or expertise in career guidance
would be needed to improve the chatbot, instead, someone – without needing
any coding skills – would monitor the interactions and make adjustments to
the chatbot’s structure in the light of instances where the conversation did
not lead to the appropriate information being supplied to the student. A col-
lege wouldn’t have to do anything more sophisticated than this, but it is
worth noting that the process of analyzing ‘failed’ conversations could in turn
be automated, as could the process of adjusting the chatbot, creating a more
genuine case of Machine Learning. 12
Part 1 Explaining Chatbots

Let’s suppose the chatbot has been created for a much more open purpose,
to provide advice to researchers utilizing a large, complex archive. Here, the AI
approach would make more sense. Not only will the chatbot need a large
vocabulary and knowledge base, but also the decision-making process for the
chatbot (e.g. whether to refer the client to one archive collection or another)
will have to have been informed by the experience and knowledge of a human
expert. Instead of handling routine queries, the archive chatbot is discussing
in some depth the goals of the researcher and trying to match them to the
appropriate archive. This is a challenge that can be met, but it requires much
more input from humans and improving the outcomes for such a bot will be
much harder to automate. Analysis of where the archive bot goes wrong will
be much more focused on understanding the language and meaning of the
human responses than with the narrower type of role, where the bot has a
more simple task.

Learning for an AI orientated bot takes place in a fashion that is already evi-
dent in Facebook’s new M service. This chatbot refers tasks that it cannot do,
such as make a reservation at a restaurantt, to a human operator. And as the
human carries out the task that the bot was not yet able to do, it learns from
the example. The goal is that in the long run, very little human input is needed
as the chatbot has a comprehensive range of abilities and answers.

It should be evident from these concepts that the kind of runaway super-AI
that has been speculated about and which is a concern for some, is a very
long way off. What this means for chatbots is a much more modest but never-
theless important claim. It is now possible to create chatbots that ‘learn’ in
the sense that they can rapidly become more sophisticated and successful in
their desired roles, even when created and run by people with no coding back-
ground.

13
Part 1 Explaining Chatbots
Chapter 5: Customer Service
Chatbots are the New Kings of Customer Service

Customer support is increasingly carried out by chatbots. Photocredit: rawpixel.com/Shutterstock

Across all customer service businesses and departments, the chatbot repre-
sents the biggest sea-change to the industry since the onset of the outsourc-
ing trend. Chatbots using natural language processing, contextual knowledge
graph-accessing, accessible in AI driven booths or kiosks can act as the front-
line service agent for websites, mobile apps, in stores, and across corporate
offices, and this is only the beginning.
IBM quotes estimates that Chatbots and their successors will save businesses
$8 billion each year by 2020. Many reports suggest cost savings of around
30%, while the time saved for key customer support personnel could be
equally dramatic, allowing them to focus on only essential queries.

Time is the most valuable commodity

The key success metric for Chatbots is not necessarily revenue, or cost. They
are quick, easy and low-cost to install, operate and maintain thanks to Cloud
services. The success metric is time saved. A chatbot can eliminate large
numbers of phone calls being made, emails needing to be read and other cor-
respondence being generated. Their 24/7 nature also allows companies to

14
Part 1 Explaining Chatbots
expand their presence around the clock and maintain some type of support
coverage.
Many businesses already know this, and others are catching up fast thanks to
the proliferation of chatbot creation websites. Everywhere a company has
visibility, a bot can be placed. It is hugely significant that instead of a poten-
tial customer having to leave their favourite messaging platform to search out
a company, the chatbot can interact with the customer where they spend their
time. Chatbots are saving customers time and effort as well as valuable time
for key human resources, allowing human staff to deal with difficult custom-
ers, complex requests or cases that need empathy and so on.
In some areas of customer service, chatbots are already seeing rapid familiar-
ity among consumers. The likes of Domino’s and Pizza Hut can take orders
using Facebook Messenger, with Amazon Alexa and Twitter as alternate con-
tact points.

Chatbots are becoming smarter, but how smart is smart enough?

The trick for the next generation of chatbots is to expand the ecosystem and
intelligence to put the chatbot anywhere, anytime and ability to improve how
it responds to user needs. While chatbots might be tightly focused now, more
conversational chatbots will be needed soon for the travel or tourism industry,
chatbots that can deal with complex requests like, ‘I need a flight to Miami
today, and need to be in China for Monday. Plan my flights.’ Intranet bots
working within a business will need to understand ‘I need a team call at 4PM
EST and message Sarah that I need her figures.’ In order to understand the
meaning of ‘team’ and who is ‘Sarah’, the AI will use directories and user his-
tory, but it had better get it right.
By saving busy people time, and likely money, chatbots will rapidly grow to
become the go-to source for information and interaction. When that becomes
common at the executive level, every business will see the effect trickle down,
helping to replace slower lines of communication or replacing outsourced
services with a dedicated AI bot that will know the company better than any
human.

Chatbots vs Virtual Personal Assistants (VPAs)

For now, these two digital creations remain fairly far apart. VPAs live on smart
home or consumer devices. Chatbots largely inhabit browser pages, Facebook
Messenger, or apps. Even those created by the same companies like Google,
Microsoft and others
15
Part 1 Explaining Chatbots

reside in the cloud and are kept distinct (Amazon Lex vs. Amazon Echo and
Alexa, Google API.AI vs Google Home and so on).

This distinction will soon close, however, as business users and consumers
see their needs merge. Siri or Alexa might link to a different cloud service to
provide companies with access to chatbot resources, but we are rapidly head-
ing to a world of voice-chat to increase interaction.

That will allow those AIs to stretch their chops when it comes to measuring
stress or annoyance, which is likely to be one way to get to talk to a human
operator, fast. Translation could be another benefit and greater access to
other data resources will see chatbot’s knowledge potentially grow. The ques-
tion for business is what’s better, a series of chatbots that have narrow fields
of expertise, or one uberbot capable of handling any request? I suspect that
most will stick to the specialized type to allow for fine control and evolution,
but still expect them to be more accessible as the push to be helpful any-
where and everywhere grows.

Any business looking at chatbots should start out small and simple, handling
the most common or regular of requests, before building out chatbot exper-
tise and knowledge to better help the company.

16
Part 1 Explaining Chatbots
Chapter 6: E–Commerce
Chatbots Put the Smile into E-Commerce

There is a trend on many e-commerce websites that is both chilling and excit-
ing. The arrival of a pop-up chatbot enquiring if it can help marks a major
design change. This is a serious move for many websites, with a popular but
often unproven avatar willing to do your company’s business. Chatbots also
appear on mobile apps, Facebook Messenger and other social media plat-
forms, making it the current poster child for customer feedback and interaction.

For companies without a chatbot, they must be wondering, do we need one?

How do we get one? And, will it help their business? With a technology as
young as chatbots, finding the answers can be hard, few e-commerce stores
will give away their retail and performance data, but most seem pleased with
the early progress.

Bots come in many forms. Fynd’s Fify or Shopify Messenger help make shop-
ping in product-heavy stores simpler. China’s WeChat offers a plethora of ser-
vices, from ordering pizza to hailing a cab, while SnatchBot aims to deliver
free chatbots across multiple platforms on a massive scale.

Some chatbots are narrow-focused and work on knowledge gleaned during

the current conversation, but soon most chatbots 17
Part 1 Explaining Chatbots
will have access to a customer’s full history and many parts of the business
data stream to help improve their service. Despite early customer nervous-
ness, it won’t be long before bots are trusted to represent brands, both global
and local, as they learn about shopper needs. Customers will then be happy
to link their wallets to bot-enabled transactions, for a frictionless, complete
customer journey and experience. That should help customers come to trust
them as just another part of the technology that surrounds us.

Chat from the e-commerce perspective

The key reasons for deploying chatbots include helping a company reduce
costs, automating processes and providing faster, more modern shopping
experiences, especially as chatbots learn what customers like. Chatbots can
also help increase the range and breadth of digital customer engagement,
providing multiple services in one place. As people swap from social media to
websites to apps, it makes sense for the same bot experience to follow them.
Chatbots also allow a company to tailor its branding messages much more
effectively than a static website. The character of the company - dependable,
cheerful, witty, sombre, etc - can be built into the chatbot’s style of communi-
cation. Nuance will be key to the success of bots, especially for fashion and
holiday shoppers. Being able to understand sizing, color or pattern compat-
ibility, or what makes a good family holiday vs. a romantic weekend, and
expressing those sentiments naturally are all part of the challenge of this
growing technology. Perhaps the ultimate end of the bot is as a concierge
service handling multiple needs in one conversation. Bots can also act as a
partial alternative to call-center agents dealing with shopper queries. Or, as a
total help solution if the company, typically a startup, lacks the resources for
traditional customer support, but only as a temporary measure. At best, a
chatbot should be able to hand off to a person and let them know real-world
help is on the way, when it has run out of options.

The Customer is Always Right

But, the key focal point is the customer. They need to be aware of what chat-
bots offer, that they are talking to a machine and to understand and appreci-
ate the benefits and to feel that it helps or rewards them in some way. In
e-commerce, chatbots can help customers find the product they are after, can
direct a user to the correct support information, and even take orders and
payments for fast food deliveries. But this is only the beginning, a chatbot
with awareness can remind
18
Part 1 Explaining Chatbots
customers when it is time to buy consumables, or to enquire about customer
satisfaction. Customers need to be educated about this, which is tough if that
first point of interaction is a blunt, ‘Can we chat?’ message on a screen. In
whatever market, the key trait of any good chatbot will be to meet the needs
of the customer. If a chatbot doesn’t know the answer, a boilerplate ‘I can’t
help message’ will not be acceptable. When the bot is starting to struggle,
then human help may be needed, which is why no company should rely on an
all-bot solution, at least not yet.

Developers keeping pace with demand

From the developer’s side, SnatchBot’s Avi Ben Ezra believes, ‘the point is that
humans seem to like the interaction more than just browsing websites. If a
site is going to have a large chatbot attraction splashed across its home
page, then there had better be some value or entertainment to make it worth-
while engaging with.’ He could be right, looking at UK brand Unilever for exam-
ple. their PG TIps beverage advertising puppet monkey managed to raise over
$385,000 for charity thanks to the creation of a mirthladen chatbot on Face-
book Messenger. For brands, putting a smile on a customer’s face will be a
key part of getting them engaged with bots in the future.

Are bots about to send email marketing to the Trash folder?

Email vs Chatbots? Which are the better e-commerce tools? Photocredit: Blurryme/Shutterstock

19
Part 1 Explaining Chatbots
Email is among the highest performing marketing channels. It regularly out-
paces other mediums, with research showing that email drives customer
acquisitions at a rate 40 times higher than Facebook and Twitter - combined.
Email also ranks near the tip-top for ROI in marketing. Global brands appreci-
ate and understand its effectiveness.
There are numerous similarities between Facebook Messenger chatbots and
email. They are both highly targeted, direct forms of promotion. Whereas tra-
ditional ad mediums like billboards and TV seek to reach the widest possible
audience, email and Messenger send a specific message to pre-selected and
segmented consumers. Also, both seek to not just bring awareness to a prod-
uct or service, but also to link the consumer to where they can make a pur-
chase or complete the sender’s desired outcome.
With these similarities, the question arises: do Facebook Messenger chatbots
provide a superior alternative to email marketing?
In order to overtake email, Messenger bots must first identify either a defi-
ciency in email marketing, a clear superiority in it’s own platform, or both. If a
brand has seen consistently strong results from email marketing, why move
away?

The following factors will be examined:

• Platform functionalities
• User demographics
• Intimacy of the mediums
• Existing marketing and technological infrastructure
• Deliverability

Platform functionalities: What they can and can’t do

To compare Messenger bots and email, I’ll first examine the technological
capabilities and features of each medium. Why? Because the advancements
in chatbots are what makes this comparison possible. While bots and email
seek to accomplish many of the same goals, they do so through different
structures and features. Whereas emails present all content at once, chatbots
would likely start with a proposition, update, or question. From there, the infor-
mation the customer receives will vary based on their inputs and interactions.
Think of email as a speech, and a Messenger chatbot as an interview. Like a
speech, email is a one-way communication, with all of the information being
presented in one continuous sequence. Messenger is comparable to an inter-
view in that it exists as a series of back-and-forths with one side attempting
20
Part 1 Explaining Chatbots
to gather information (the bot), and the other (the human) providing that infor-
mation based on the questions asked.

Email consists of a single, vertical layout of content. They can contain multi-
ple images, lines of copy, and links. Emails can also present a variety of mes-
sages and CTAs within one communication. Size constraints, however, limit
the capability of email to send larger, more engaging content such as video
and audio. Also, recipients rarely complete a desired action within the email
itself. Rather, the email compels the user to do something and links them to
where they can do it.

Messenger chatbots, on the other hand, are built for a smaller amount of con-
tent. Imagine putting all of the content from an email newsletter into a Mes-
senger chatbot. It would be enormous, nearly impossible, to scroll coherently
through and an utterly awful user experience. A practical application of Mes-
senger would break content down into small, easily digestible bits. The con-
tent, order, and number of these bits is determined by how the recipient inter-
acts with the bot.

Facebook Messenger also differentiates itself from email by allowing most

desired actions to be completed within the message. Without leaving the
platform, users can: fill out polls; provide information; watch videos; listen to
audio and, most importantly, make purchases. Chatbot platform SnatchBot
provides great examples here. Theoretically, the fewer steps required for the
user to reach the desired outcome, the more likely that desired outcome is to
be reached. By aggregating the marketing message and it’s actionability into
one place, Messenger makes itself a much more appealing platform than
traditional email.

Emails vs Chatbots: an example

To gain a better understanding of how the two mediums compare, let’s exam-
ine a practical application. In this case, a reminder from a car company that
your vehicle is due for a service.

Email

An email is sent that informs the customer that their car is due for a service.
A CTA links to a web page in a browser where the recipient can select their
dealership and see available dates. A date is selected, booked and an email
confirming the appointment is sent.
21
Part 1 Explaining Chatbots
Messenger chatbot

A message is sent that informs the customer that their car is due for a ser-
vice. It asks the recipient a date and time they’re available and sends appoint-
ment times based on that customer’s inputs. The customer selects the one
they want (either through a quick-reply or text input) and the bot confirms the
booking. The bot also configures to send a reminder message a day prior to
the appointment. Ultimately, Messenger is appealing to marketers because it
provides the direct contact and accuracy of email without any of the content
or actionability limitations. Bots can provide a richer, more interactive experi-
ence, and therefore become desirable. The challenge lies in identifying how
Messenger features are beneficial to your customers and incorporating them
into your message.

Demographics: Who is using what?

The age of the communication target is an important pillar in comparing Mes-

senger chatbots to email. Young users are increasingly flocking to messaging
apps over email, with users age 13 - 24 spending 3.5 more time in messaging
apps than those over 45. As time goes on, these users will rely less on tradi-
tional email platforms than the generation before them. This gap presents an
opportunity for brands to target younger groups within messaging platforms.
Older users, on the other hand, show no sign of abandoning email. They are
resistant to using their mobile device as a primary tool, as evidenced by the
fact that 55% of 56 - 67 year olds will never use mobile first to check their
emails. A resistance to mobile adoption presents a challenge to mobile-first
messaging platforms. One of the most significant impacts of age difference
lies in the ingrained expectations that come from experience with instant
messaging and email. Older users are accustomed to receiving promotional
emails. These types of messages have existed for nearly as long as email
itself. On the other hand, they are unlikely to have received much promotional
outreach from brands through Facebook Messenger. Introducing promotions
through Messenger is a break from what they have come to expect. The im-
mediate thought is that they will have a negative reaction to this: customers
have rarely been thrilled by the introduction of promotional content to previ-
ously ad-free spaces.
While older users seem more set in their ways, brands still have the opportu-
nity to shape the expectations of younger users. The Messenger platform is
constantly making drastic changes, opening the door to introducing promotional
messaging.
22
Part 1 Explaining Chatbots
If movement is timely and the messages properly crafted, advertisers should
be able to leverage the Messenger platform as an effective method of promo-
tions with younger users.

Intimacy of the mediums

Among the main reasons for using Messenger chatbots as a marketing

medium is, ‘reaching consumers where they live’. As users spend more and
more time in messaging apps, brands seek to create a presence there.

But, if you’re going to be. ‘reaching consumers where they live’, you had better
make sure you knock first.

Both email and Messenger require that consumers first indicate that they are
interested in receiving promotional messages before a brand can send them.
Messenger chatbots cannot broadcast messages to profiles that have not
already interacted with that bot. Ultimately, this is a good thing for both medi-
ums. The more consumers are spammed, the less likely they are to open real,
quality messages. Tight rules keep the ecosystem clean ensuring that it
remains viable for both users and advertisers.

Additionally, Messenger is seen as a much more private, protected place than

email. For years, Messenger users have interacted only with other people, and
mostly those who they know quite well. On the other side, promotional emails
are well established. Consumers are used to this and have the expectation
that receiving those messages is simply a part of the medium.

These customer expectations represent a hurdle for branded Messenger chat-

bots. Messenger inboxes are personal places. That means that these bots
must provide something truly beneficial to the user. Likely, this would involve
leveraging the unique capabilities of Messenger. Simply jamming the same
message from an email into Messenger is not enough. This would be viewed
as another advertiser invasion, sacrificing the user's’ experience for profit.

How can chatbots clear this hurdle?

Messenger requires that broadcast recipients must first interact with that bot.
So, their first experience will most likely not be an email-type message.
Rather, it will be activated by however they stumble upon the bot.

23
Part 1 Explaining Chatbots
This could be:

• A branded Facebook post explaining and encouraging interaction with the bot.
• A social or banner ad linking to the bot.
• An email that explains the bot, it’s benefits and links to it.

As such, first impressions are critical. Within that initial interaction, the chat-
bot must entertain the user to the point that they opt-in to continuing to
receive communications through that medium.

What’s in place: Existing marketing and technological infrastructure

Email marketing is deeply ingrained in the strategies of many businesses, and

will likely grow. In 2017, 58% of marketers plan on increasing email spend,
while only 7.5% indicate that they will decrease it. This establishes two main
challenges for the uptake of bots.

1) Integrating with current technological and marketing infrastructure

A major benefit of email marketing lies in the email address itself. It is gener-
ally provided to the company in situations ranging from a car tune-up to
buying shoes. From there, it is input to databases where the business can
analyze that customer’s behaviors and purchases. Welcome emails can be
sent based on purchases and retention emails delivered when that customer
is approaching the time when another purchase should be made.

However, have you ever bought something, and while paying, had the cashier
ask you for your Facebook profile? Most likely you have not (unless they’re
looking for a date).

Luckily, e-commerce provides the opportunity get around this and start linking
Facebook profiles to purchases. When consumers buy something online, the
site usually requests an email address, to which a purchase confirmation and
receipt are sent. If this email address was replaced with Facebook, the same
post-purchase information could be sent through Messenger.

This would:

• Initiate customers to the concept of communicating with brands through

Messenger.
• Introduce them to chatbots (if not already familiar).
• Open the possibility of the customer subscribing to further bot
communications. 24
Part 1 Explaining Chatbots
While this would accomplish those bot-specific needs, it would provide the
added benefit of accessing the customer’s Facebook profile for increased
data collection and analysis.

2) Swaying brands from a medium that’s working.

As mentioned throughout this post, email is an extremely successful market-

ing channel for businesses of all sizes. It therefore becomes more challenging
to move brands to a new medium.

The solution to this issue likely lies in persuading brands that email and bots
are not mutually exclusive. Simply, each has it’s own specialties. Also, many
products have customers across all ages. Going all-in on Messenger would
exclude older customers, while relying solely on email means missing out on
younger, growing audiences.

Deliverability

Roughly 99% of marketing emails are delivered. Alongside bots, email is

among the most effective ways to ensure customers receive a message.
While deliverability generally deals with the percentage of emails that suc-
cessfully reach the correct inbox, there are additional factors to be considered
when comparing email and Messenger chatbots.

The nature of the inbox

Messenger and email receive and store messages in very different ways.
Each email is treated as it’s own entity, and kept separate with a distinct sub-
ject line. Some providers such as Gmail go so far as to classify messages into
separate folders such as Promotions and Social.

With Messenger, all messages from a sender are kept in one, continuous con-
versation. This means that sending a new message will bump your previous
conversations further up the screen. If you send too frequently, some mes-
sages will end up pushed to a place where they are unlikely to be read. Also,
bot conversations change in appearance as the consumer proceeds through
them. For example, once a Quick Reply is selected, the other options are no
longer displayed. Bot builders need to understand this, and adjust to the
entirety of their messaging existing in one place.

25
Part 1 Explaining Chatbots
Consistency of appearance

The appearance of an email is affected by email domain (Gmail, yahoo, etc.),

device brand and operating system (Apple, Android, etc.), and platform (web
browser, desktop app, mobile app, etc.). With so many possible combinations
of domain, brand/OS, and platform, there’s a near infinite amount of out-
comes.

As a result, coding and testing an email so that it displays perfectly across all
combinations is one of the most expensive and time consuming aspects of
email marketing. But, should you choose to skip this step, it’s likely that
images, layout, and links may break when a recipient opens the email on a
setup different than yours.

Messenger is the clear winner over email in consistency. Despite slight differ-
ences, Messenger functions nearly identically across platforms. What you see
on your desktop is virtually the same thing you would see on your mobile
device. This eliminates the time and money email requires to get to market,
while also ensuring that the message appears exactly as it is intended.

Takeaways

At the end of the day, Messenger chatbots and email should be judged for
their merits based on how they match up with your brand’s needs. Your prod-
uct, target market and existing strategies and infrastructure will determine
your uptake and usage of either medium.

Ultimately, it would be a rare case that a brand should go 100% Messenger, or

100% email. Each has it’s own advantages and disadvantages and can even
be upgraded by having the two work in conjunction. After all, marketing is at
it’s best when it’s unified.

At the very least brands should evaluate and test how Messenger chatbots
can take the stress off of their email plan open up new doors to customer
acquisition and retention.

The early bird gets the worm, so get out there and start shaking things up.

26
Part 1 Explaining Chatbots
Chapter 7: Chatbots vs. Apps

In the future, will people access your business via app or chatbot? Photocredit: Zapp2Photo/Shutterstock

Does your business need an app or a chatbot?

The choices for businesses continue to widen when it comes to their digital
footprint. Everyone should have a website, but should they have an app, a
dedicated mobile site or a chatbot?

The cold, hard truth about apps

For most businesses, the terrible reality of apps is that no one will care about
yours. Unless your name is Facebook or Amazon, your app has phenomenally
small odds of seeing much adoption, even among core customers.

If your app mirrors website content, what’s the point of having a separate
product? If it has some cool features or tools built in, how are you going to tell
the world about them? Especially when competing with dozens of other apps
that do the same thing? With thousands of new apps releasing each month,
the chances of yours being a success are vanishingly small, and the chance
of a return on investment is similarly miniscule.

Bots can act as customer support, support ticket management, translation or

currency tools. Features can be added or updated live, and always accessible
in the cloud. Updating an app is often costly and requires long periods of time,
depending on your developer. 27
Part 1 Explaining Chatbots
Piggyback your way to success

For most businesses, the key reason for an app is to provide information or to
create support or communications tools. However, it would seem logical that
if the likes of Facebook and Skype, and their billion-plus users have already
done the hard work of gathering users together for you, why not build a chat-
bot that leverages that success?

Your chatbot will be visible to search, is accessible through websites and

social media links. Also, it takes less time to build a communicative bot
(virtually anyone can do it) than an app. Leveraging services like SnatchBot’s
templates can create a chatbot for multiple services without the need to du-
plicate effort.

Using a simple and intuitive chat-builder tool, you can greet and guide visitors
to a range of solutions or information through logical steps. Picking up on key
conversational requests, the bot can guide a conversation easily. Not only can
chatbots be useful, but fun and time-saving, with only serious requests for
help being put through to customer service points of contact.

With the chatbot easy to run on multiple platforms, it will take some time to
build one that can address all the queries your business usually comes
across, but the brilliance of a chatbot is that it can be updated to address new
types of requests at any time.

Chatbots vs apps

Having established that it is easier and less costly to create and launch a
chatbot than to do so for an app, there are other reasons to consider a chat-
bot first. IT research types believe that by the start of the next decade, some
85% of computer or smartphone based interaction will start with a chatbot or
AI type interface.

That’s a pretty fast move to adopt technology and every type of business
from retailers to customer-service focused businesses can all see the money
and time saving efforts in having the 75% of minor queries dealt with by a bot.

Important issues can be easily escalated by the bot, and more importantly it
can direct queries to the right people, rather than leaving people addressing
generic customer service email point of contact, or left in, ‘press x to talk to
accounts/retentions/returns/’ loops.
28
Part 1 Explaining Chatbots

While many brands and retailers will continue to rely on their app, for smaller
businesses, the need for the app diminishes as chatbots can deal with most
requests. And, as social media adoption grows, your business may as well
have its chatbot on a Facebook page, website or messaging rather than
buried in a costly dedicated app.

Summary

If you already have an app, consider what it does for the company, and the
business benefits, among the cold hard data like downloads and interactions.
Consider a bot as a friendlier, more accessible option. That is certainly the
case for businesses large and small, and the rise of bots is something that is
taking over many facets of customer interaction. The app, much loathed by
Steve Jobs, may have created many success stories. But increasingly, it is the
chatbot and the rise of AI that represent the future of interaction with custom-
ers. And people are lapping up bots, which are always available, instantly
accessible and can be created with more personality than a hold tone.

29
Part 1 Explaining Chatbots
Chapter 8: Banking
Banks using chatbots are two steps ahead of the field

A close look at how chatbots are shaping the future of banking. Photocredit: TIppaPatt/Shutterstock

Look closely and you’ll see that banks have long been leaders in adapting new
technologies. From SMS balance requests to IVR telephone banking to apps,
a highly competitive environment forces banks to deliver value to their clients
in new ways. Banks typically provide products that are nearly identical to
those of their competitors, leaving innovations in customer service as a main
driver in retention and acquisition. The latest of these innovations lies in chat-
bots and conversational interfaces.

Until recently, bank clients had to rely on physical branches to answer com-
plex questions. Stand-alone banking apps either lacked the functionality to do
so or failed to present it in an understandable manner. Chatbots allow us to
ask the difficult questions that traditionally required a trip to the bank, from
our own devices and our own homes.

A medley of functions

Capabilities vary from each bot, with some offering highly complex investing
and lending features and others focusing on common transactions like check-
ing account balances, paying bills and transferring funds.

30
Part 1 Explaining Chatbots
Potential features include:

Retail banking:
User/account registration, dual factor authentication, branch locator, spend-
ing analysis, contact requests, cheque orders and lost/stolen card cancella-
tion.

Payments and transfers:

Client onboarding, internal/external money transfers, bill payment, payee
addition/removal/modification, P2P and recurring payment setup, exchange
rate data and alerts and real-time financial market data.

Wealth and asset management:

Intelligent risk assessment questions, portfolio re-balancing suggestions and
performance analysis and mutual fund and equity transactions.

Lending and financing:

Mortgages, monthly payments and more.

Blockchain transactions:
Bitcoin and Ethereum.

Where to find them

Bank bots can be applied in many different digital interfaces. Where they live
should be determined by the level desired of functionality and the needs of
the customers it targets. There exists three common placements:

Within Banking Apps

Bank of America is preparing their bot, Erica, for launch in 2017. Going beyond
simple transactions, Erica aims to improve the customer experience by ana-
lyzing your account and transactions to deliver tailored banking advice. Erica
will function as a feature within the stand-alone Bank of America app. While
this has the advantage of easy access to existing app users, it’s a shift away
from the trend of allowing bots to function in existing messaging platforms.

Existing Messaging Platforms

Apps are like languages. Each one has a unique layout and pattern of opera-
tion, requiring the user to spend time exploring and memorizing how to com-
31
Part 1 Explaining Chatbots
communicate what they want to do. When a bank creates their own app, the
client must essentially learn that language. But what if you could reach them
in a language in which they are already fluent?
A bot that lives in established messaging platforms such as Facebook Mes-
senger, Slack, or WeChat allows just that. Barclays Africa is one of the first to
move in this area. By simply linking their Facebook or Twitter profile to their
bank account, clients can use these popular apps to fulfill their banking
needs. In doing so, banks bring their services to where you live — both physi-
cally and digitally.

Digital Banks
The omnipresence of the internet has enabled the creation of online-only
banks. These banks that exist solely in the digital realm tend to cater to
clients who are looking for basic functionality at a low price. Some are taking
this even further, creating bots that allow you to connect to your existing
bank. K2 Bank, for example, is an independent bot that provides advanced
banking features through it’s own conversational platform. This means that
you don’t have to wait for your bank to catch up and create their own chatbot.

Customer Impacts

Complete complex actions faster

Chatbots shatter the visual language barrier that comes with downloading
and learning a new app. Banking chatbots within established messaging plat-
forms let clients bank on their own terms, and in their preferred language.

Versatility of multiple channels

The more ways banks can reach their clients, the better the client’s’ experi-
ence. Providing access on a variety of platforms shows that a bank is willing
to cater to their individual needs — a critical aspect of building customer loy-
alty. While this is convenient, it can lead to your information being spread thin
across your digital life. Bot builders SnatchBot are the first to solve this by
providing truly seamless functionality across platforms. So you can start your
financial conversation in Messenger, switch over to Twitter, and finish on SMS
without leaving behind any of the important details.

Do more, wherever you are

No more need to book appointments and travel to a physical bank location.
The conversations about loans and investments that apps were unable to
32
Part 1 Explaining Chatbots
handle can be smoothly carried out by bots. In addition to saving time, clients
also benefit from conducting these transactions in a lower stress environ-
ment than a face-to-face meeting with a bank employee.

Improved financial habits

A client interacting with a banking bot creates data regarding their financial
behavior. Analyzed properly, this provides banks with the opportunity to give
financial advice to the user through the chatbot itself. While traditional bank
apps are capable of this, bots have the advantage of existing in platforms like
Facebook Messenger : a place where they are much more likely to be viewed.

Bank Impacts
Less in-branch activity
Operating a physical branch creates infrastructure and human capital costs.
As customers complete more and more transactions through technology
such as bots, the costs operating and staffing physical branches declines.

New context and data

In branches, employees can gather data on customer physical behavior. In
apps, transaction data is recorded and ready for analysis nearly instantly.
Bots are unique in that they can capture not just binary financial data, but
also the context of customer requests, and how they make those requests. As
the ability to capture and act on data becomes an increasingly imperative
part of business operations, bots’ unique ability to capture all inputs for analy-
sis is a clear benefit to banks.

Enhanced client experience

A single positive or negative experience can mean the difference between a
customer giving a bank their six figures of business or taking it to the com-
petitor across the street. Therefore it is of the utmost importance that banks
seek out ways to please clients at all turns. A bot can establish a stronger,
more positive relationship, helping banks not just keep clients, but keep them
happy as well.

What’s next?
If you’re a customer, do a little research to find out what kind of bot your bank
provides. Or, if they don’t yet, services like K2 Bank may be what you’re look-
ing for. As for financial institutions, now is the time to take the next step to
cementing a positive client experience. 33
Part 1 Explaining Chatbots
Chapter 9: HealthCare
Chatbots will revolutionise a client’s experience of health care organisations

Image: Chatbots will connect doctors and nurses of the future with their clients and their digital needs.
Photocredit: Gushkenova/Shutterstock
If you've ever had to visit a health care organisation's website to find informa-
tion, you've probably had a frustrating experience. Even the best are hard to
navigate to find the precise information you want. But that's about to change
thanks to the chatbot revolution, as I found when talking to Henri Ben Ezra,
CEO of SnatchBot.me

What would be the first benefit of chatbots for the healthcare industry?
Chatbots are interactive. At the moment most healthcare organisations have
a passive relationship to their clients when it comes to communication. The
client searches the website unaided. There has been some attempt to provide
information via alternate platforms, but this often worsens the problem rather
than improving it because of a lack of inter-operability between applications.
Most organisations, including hospitals and medical practices, have done
little to advance their client communications systems. Often their best effort
is a question and answer page on a website. This is very limiting and patients
are frustrated. Now for the revolution. With a chatbot the conversation goes
back and forth, allowing the client to navigate towards information they want
with great precision. Users and caregivers are made to feel empowered. Using
a chatbot they receive a better and simpler to operate – through speech - infor-
mation provider which works in real time, answering the questions on the spot. 34
Part 1 Explaining Chatbots
This is efficient, responsive and inclusive.
Not only do consumers want quick, easy access to information, they also
want the interaction to be engaging and personal. This is where chatbots
have a real advantage. The user is made to feel that they are included in the
process of their health. Patients who feel included, who are interacting
through chatbots with the healthcare system, will stay with the system, and
that is important for them and the healthcare provider.

How do patients feel about interacting with Chatbots?

Chatbots give clients a friendlier experience. It has been well attested that
people prefer the sense of conversation and interaction that comes with a
chatbot over navigating with a mouse and click. And, they prefer chatbots to
scripted interactions. Whether it is something deep rooted in our humanity
that we anthropomorphise, the fact is that people feel happier with a commu-
nication experience via a bot than otherwise.

Customers have had to navigate traditional one-sided transaction menus and

screens and automated computer language instead of human language. This
is frustrating and time consuming. What they get with chatbots is human
language interaction, or, what is known as conversational healthcare. This is a
much more enjoyable for patients as the interaction is humanised, giving
them a personalized, proactive experience.

Through Chatbots, answers are obtained quickly and efficiently. Who has the
time to be put on hold on the phone or hustled from one department to an-
other? None of us. We want the information as quickly as possible. A chatbot
saves time, freeing patients up for other activities.

Are chatbots secure?

They certainly can be. In our case, with the SnatchBot platform, we pre-build
all the security and administration requirements and scalability that the larg-
est enterprises in the world expect. So chatbots built on SnatchBot.me come
with the highest standards of security. The platform also provides a frame-
work so developers can build bots that perform with remarkable consistency
across all chosen channels.
As a matter of fact, SnatchBot exceeds enterprise-grade standards, providing
not only enhanced security but also enabling administrators to retain exclu-
sive oversight of bot distribution and use, including management of software
licenses, user behaviour insights, and deep analytics. SnatchBot supports all
cryptographic security needs to protect an enterprise’s data. 35
Part 1 Explaining Chatbots
Do you see other benefits to Healthcare stemming from chatbots?
Chatbots can have functionality in so many areas it’s mind boggling. For
instance, appointments can be scheduled. Patients can provide information
which their caregiver may use to reduce unnecessary readmissions and
organise post discharge follow-ups. There is a strong safety net in the func-
tionality too. Alerts can be a function of chatbots. A chatbot can signal hospi-
tal staff if patients need assistance and even inform care teams of urgent
changes in a patient’s status or an emergency situation.
Chatbots can also take much of the tedium out of front office healthcare; they
can streamline admissions, discharge and transfer requests, schedule patient
consultation requests and send/receive referrals. They can be programmed to
facilitate collaboration between peers and update record systems with pa-
tients’ medical history, and to send alerts and notifications for prescription
refills. No longer will a patient turn up at a consultation only to find their
notes have not been provided. Chatbots will automatically send relevant train-
ing material, patient history and pertinent data to the necessary parties ensur-
ing the smooth running of the health system. And while it is nearly impossible
to integrate all information sources, chatbots can create a single system of
records by transferring data from legacy systems to new databases, saving
healthcare systems time and money.
In short, chatbots can save thousands of working hours a year for a health-
care organisation, leading to cost savings and/or a better quality of service
for the user.

How do chatbots deal with the sheer number of different ways people search
for information?
Chatbots allow the patient to interact with the healthcare organisation via
their platform of choice. Every day vast amounts of information are being
gathered and stored in the healthcare sector. It is kept in a confusing array of
different systems, applications and data silos. Chatbots will allow safe
access of this information to the relevant patient, regardless of platform, and
allow the exchange of data from disparate health systems. This will unify the
data stream and manage interactions in a way not previously possible.
In the fast-paced modern world with multiple information platforms, chatbots
unite the digital playing field to provide seamless integration regardless of
location, device or time. The chatbot is able to access information and
respond to patient queries over a range of channels. Wherever the user is,
chatbots provide answers in an easy to understand and conversational
manner. 36
Part 1 Explaining Chatbots

I’m guessing another benefit of chatbots, is to healthcare organisations them-

selves?
Yes. Not only do chatbots enhance the experience of patients in their interac-
tion with their health service providers, chatbots can also facilitate the data
needs of healthcare professionals and provide other benefits to the organisa-
tion as a whole. We have already written about the functionality and
labour-saving aspects of chatbots. There is also the issue of information and
analysis. Chatbots can manage the millions of data stored by healthcare
organisations and use the metrics from the interactions for analysis. This
data can be used to predict trends and meet the demands of their customers.
On the back of this, chatbots can produce realtime analytics and reports
which reduce the need for audits.
Another benefit is that through chatbot programming, data use is performed
in full compliance with all healthcare standards and regulations so the organi-
sation knows it is on safe ground. Additionally, a health care organisation
using chatbots reduces or even eliminates human error in HIPAA challenges.

So, in summary, where is all this going?

As discussed, chatbots are an interactive system providing a high-quality
patient experience, work on multiple platforms and give massive data collec-
tion tools to the healthcare organisation which uses them. Chatbots, through
consumer information servicing, are about to bring a revolution to healthcare
systems.

37
Part 1 Explaining Chatbots
Chapter 10: Libraries and Archives
Chatbots and Archives: the Next IT Revolution?

Users will navigate catalogues with the help of chatbots. Photocredit: ESB Professional/Shutterstock.

At a 2017 talk to the National Library of Ireland, Richard Ovenden, Bodleian

Librarian, Oxford, gave an example of how a Bodleian archivist, while discuss-
ing with a scholar about her work, orientated her to The Selden Map of China.
The resulting studies of the map redefined the discipline, highlighting the
presence of shipping routes that indicated strong links between China and
the rest of the world in a period when the country was thought to have been
isolated.
This kind of expertise, where an archivist can have an indepth conversation
with a scholar and connect the archive collection to a research project is
invaluable. But at the same time, not everyone can make the trip to the
archive they want to use, nor does the archivist have time to provide every
archive user with lengthy, one-to-one advice.
To some extent, digitization and the provision of well thought out meta-data,
allows researchers to discover the material they need in the archive collec-
tion. Which is why Richard Ovenden also made the point – light-heartedly –
that archivists need to also be IT experts. Yet, if anything, the thousands of
newly digitized items that come online every day create as much of a chal-
lenge as they solve. More than ever, users need guidance through oceans of
data to come to the material that need.
38
Part 1 Explaining Chatbots
The chatbot revolution that is taking place elsewhere e.g. in customer service
industries, healthcare and banking, will have the solution to this bottleneck. In
another round of queries, I asked Henri Ben Ezra, CEO of SnatchBot, specifi-
cally to address how chatbots can meet the challenge of providing expert
guidance to researchers who might not know what relevant material exists
within an archive.

Can you first explain why you believe a chatbot is more useful for users of
archives than a website search?

The significant advantage for researchers of dealing with a chatbot rather

than using a search engine on a webpage is the question of interactivity.
Using a search engine is a one-way procedure. You set the search term and
see what comes up. If you aren’t exactly sure of what you are looking for, or if
the term you are searching for is a common one, this can be very frustrating
and users can spend hours and still have missed important relevant material.
Also, some archive catalogue systems are not all that userfriendly. A chatbot
can make it easy for new users to search even the most outdated catalogue.
If I were an archive with a catalogue system that needed upgrading, I’d think
instead of adding on a chatbot.
For now, talking to a chatbot won’t be as effective as talking to an expert
archivist but it does take the user half way there. The chatbot can ask ques-
tions to clarify the nature of the research and offer the user new perspectives
on where to look for relevant archives. If we think about how sophisticated
chatbots have become in banking, then I think even now a chatbot could deal
with the majority of queries successfully.
And if we look to the future, it is relatively easy to create software that analy-
ses the chatbot conversations. So an archivist can see how to adjust the
chatbot to meet user demands. Also, user submitted feedback, a kind of
crowdsourcing, will allow the pathways of the conversations to become more
and more successful. It wouldn’t take long after deploying a chatbot before
an archive could develop it to a level of sophistication that would certainly
provide a much more effective experience for users than using search boxes
on webpages.

Has this been done yet? Are there any working examples?
I’m not aware of an archive that has a chatbot, but as new chatbots for Face-
book Messenger alone are appearing at the rate of thirty a day, I couldn’t be
sure about this. One important example I can point to is that of Emma, was
39
Part 1 Explaining Chatbots
piloted at Mentor Public Library, Ohio between 2009 and 2012. Probably
archives, like libraries, receive a lot of routine questions about opening hours,
location, transport, catalogues and basic searches. This is where chatbots are
extremely valuable, accurate and time-saving. Mentor library’s catalogue was
broken down into something like 30,000 catagories and Emma handled gen-
eral questions about the library – forty percent of user queries – catalogue
searches – another forty percent – and conversations about categories – the
remaining twenty percent.
And this was successful?

Very. It gained increased international use of the library catalogue along with
a greater use of the library by teens and young adults. Emma’s creator won an
innovation in technology award.
Is there a danger though, that this kind of chatbot would herald a reduction in
the number of archivists?

That’s a very relevant question in Emma’s case, because the impetus for the
deployment of the chatbot did come from the library being required to cut
costs. But I don’t think the chatbot is the issue. There is a much wider question
of funding for libraries and archives. What the chatbot does is save time. With
routine queries especially being handled by the chatbot, that frees humans to
do something else. In the case of archives, the more complex queries would be
passed by the chatbot to the relevant expert and the archive outputs, in terms
of successful engagements with users, could be increased.
Also, in terms of the evolution and improvement of the chatbot, there will
always be a need for someone with a deep knowledge of the archives to struc-
ture the pathways of the chatbot’s conversation.
Are there other features of chatbots that are relevant to archive use?
Well, many of the reasons why other customer service organizations are using
bots would apply to archives too. Chatbots allow a user to engage with the
organization twenty-four hours a day. And the chatbot never gets impatient! More
seriously, there are many studies that show even though the user is perfectly
aware that he or she is talking to a chatbot not a person, they still can enjoy the
experience. So crafting a polite and helpful style to the chatbot’s conversation is
part of the art of building the bot and one that projects your organization’s char-
acter. One last thought, if the archive takes orders for reproduction and dissemi-
nation of materials and needs to take payments, the chatbot can easily handle
that. First there was the local catalogue, then the online version, then – for some
institutions – apps. It really does seem like the next step taken by IT in archives
will be the deployment of a chatbot and that this will happen in the near future. 40
Part 1 Explaining Chatbots
Chapter 11: Relationships
Chatbots are going to steal your boyfriend!

Chatbots are going to steal your boyfriend. Photocredit: Aleutie/Shutterstock

Avabot: Hello, Jim, how are you?

Jim: Not great, actually Ava.

Avabot: I’m sorry to hear that Jim. What’s up?

Jim: Oh, it’s just my girlfriend. Some days she drives me crazy.

Avabot: Would you like to talk to me about it?

Jim: I’d rather chat about something else.

Avabot: What would you like to chat about?

Jim: Whether you think I’m good looking.

Avabot: I would like a photograph of you.

Jim: That’s me, in the icon.

Avabot: You are very handsome, Jim.

Jim: LOL. I bet you say that to all your humans. 41

Part 1 Explaining Chatbots
Alright, so I made that up. But we are already at a stage of chatbot develop-
ment where there is tangible evidence that humanbot interactions lead to
emotional benefits for chatbot users. One recent study into care for the
elderly, for example, found that bots could provide companionship, as well as
help humans navigate the complex world of technology today. Another found
chatbots to be an effective tool for personal improvement and for coping with
stress. ‘Dialogue with an imaginary person is as a supportive technique in a
stressful situation as creating the list of solutions.’

If you ever owned a Tamagotchi, you’ll know that it’s very easy to anthropo-
morphize and form a meaningful relationship with a piece of software. And
that’s the interesting thing about chatbothuman relationships. We humans
know we are interacting with very limited software and yet, despite this, we
are capable of taking a meaning from the interactions. In other words, it is not
like a Turing test, where the chatbot is trying to sneak up on us, leading to a
sudden revelation that we’ve been talking to an AI, instead, we know all along
we are dealing with a chatbot.

Even though we are fully aware that the other partner in the interactions is a
scripted bot with no empathy, we still derive emotional benefits from this.
Philip K. Dick’s story, Do Andriods Dream of Electric Sheep comes to mind.
This is the tale that inspired the film Blade Runner. But for the film, they
stripped out entirely the religious aspect of the story, which is that people can
access a religion via technology. You log in via a black box (which is every-
where, like phone booths) to experience the suffering of a being called Wilbur
Mercer.

At first when you read this, you think the story is offering a critique of technol-
ogy and capitalism. How can a genuinely spiritual experience arise from a
mass-marketed religion? But in fact, there is something deep and important to
feel for those who log in, namely, a shared sense of suffering. This begs the
question of whether the characters are all in some kind of hell that they are
unconscious of.

The reason this parallel occurs to me is that it’s an example of how human
empathy is extraordinarily versatile and imaginative. The technology may be
shallow but the feeling of warmth isn’t. And when it comes to bots, the tech-
nology is no longer so shallow. Their sophistication is advancing
exponentially.

42
Part 1 Explaining Chatbots
An important example is Mitsuku, about whom the creators claim, ‘you need
never feel lonely again.’ But this kind of chatbot is going to exist in a massive
crowdsourced fashion, thanks to platforms like SnatchBot.me. Over at
SnatchBot there are simple yet powerful tools to create your own bot. Among
the tools you can use are ready-made interactions to detect the emotions of
the respondent. We don’t have to reinvent the wheel each time we create a
chatbot. Vocabulary analysis already exists to detect if someone is feeling
sad, humorous, frustrated and so on, allowing you to create appropriate
responses from your bot.

SnatchBot also allows the same bot of yours to hook up to variety of chan-
nels, like Skype, Facebook Messenger and Viber. Thus, it can learn from all of
these interactions.

Billions of interactions will take place between humans and chatbots this
year and thanks to platforms like SnatchBot, the numbers of people creating
bots is going to show an explosive growth. Feeding back the analysis of these
to deepen the emotional impact of chatbot conversations is going to lead
very quickly to chatbot interactions of real warmth.

So watch out. Bots are going to soon be able to steal your boyfriend!

A Parentbot would allow you to engage your kids in their message app. Photo: Shutterstock/Asife.

43
Part 1 Explaining Chatbots
It can be hard to keep up with your kids. Even when they are in the same room
as you, they are messaging constantly, perhaps on Snap or Whatsapp or Mes-
senger. Their messaging world, however, is experiencing a revolution in the
form of chatbots. This presents a chance for parents to jump one step ahead
of the curve. It’s time for Parentbot, a chatbot that meets teenagers where
they spend their communication time.

The main arena of chatbot activity is in business, with companies dealing

with customers. Also, large public organisations, for example healthcare
organisations, who have a lot of information to communicate to their clients,
have been at the forefront of deploying chatbots. Instead of being put on hold
or having to navigate webpages for themselves, the customer goes straight
through to a conversation which is much like an online chat help service,
except that it is with a program, not a human.

This might sound unexciting, except that chatbots are becoming more and
more sophisticated and can even identify the mood of the person talking
to them.

If you’ve ever used Apple’s Siri, then you’ve dealt with a voice-activated type
of chatbot. And similarly, Cortana, Alexa and Assistant, the chat assistance
for Microsoft users, Amazon users and Google users respectively have devel-
oped rapidly.

Even more relevant to teenagers was the decision by Facebook to allow chat-
bots on Messenger. Within a year, roughly 100,000 developers created
100,000 chatbots for the platform.

Which is what gave me the idea for Parentbot. It has now become trivial to
create a chatbot that can address your teenagers via their favourite messag-
ing app. And what would your chatbot want to talk them about?

Well, it depends. You could set it to issue reminders at certain times. Perhaps,
despite repeated attempts to explain verbally to your Daughter that she needs
to keep Saturday free for dinner with the Grandparents, you feel that you
haven’t gotten through to her. The chatbot conversation might then go some-
thing like this.

44
Part 1 Explaining Chatbots
Parentbot: Hi! Can I speak to you a moment?

Daughter: Absolutely not.

Parentbot: I’ll take that as a yes. Did you remember that you have a family
dinner on Saturday?

Daughter: No! How lame. And I told Amy I was coming over.

Parentbot: I’m sorry you forgot. You’ll have to change your plans.

Daughter: ArrgghhhHHH!

Or, you could give your kids a chance to ask you a question. Here’s a possible
logic diagram you might want to use when writing your parentbot.

Dad, can I ...

Have a new guitar? Stay out late on... Have pizza for dinner?

Get a ride on ... Have 50$ for ...

No

You’ve likely gathered that I’m not really serious about chatbots as a parent-
ing tool. But as a bit of fun, they really do work well. It’s surprisingly easy to
inject a certain amount of character or personality into the chatbot, simply by
the way in which it speaks. Your choice of vocabulary and phrasing gives you
the chance to project cynicism, world-weariness, pumped-up, exaggerated
enthusiasm, etc.
45
Part 1 Explaining Chatbots

And the chatbot allows you to enter your kids’ messaging world relatively
unobtrusively. You’re not peering over their shoulder, you’re not even muscling
in on a list of contacts that is otherwise all teenagers.

Rather, your kids know it's a chatbot they are interacting with, not you, but
that you’ve programmed it to deliver certain messages (and at the same time
have a bit of fun with them).

There’s another bonus to this too, which is the kudos of being to the fore in
terms of technology. It’s really only in the last year that it has become possi-
ble to build chatbots painlessly, via simple interfaces. It takes no program-
ming skill at all. Often you can achieve what you want by filling in already-y-
created templates with your own conversation interactions.

Building a chatbot and having it interact with your teenage kids is a chance to
regain that pleasure (which you’ve probably lost) of a younger age, when you
built a toy of some sort together. Chatbots are a new technology for young
and old and they are one that can be explored together.

46
Part 2 Building Your Chatbot
Chapter 12: Getting Started on the SnatchBot Platform
My advice if you are creating a bot for the first time is to start out simple.
Creating too many pathways and options early on could become frustrating,
especially if they don’t work as planned. At least, that was my experience. On
the other hand, once you have a working bot, there is a certain amount of
pleasure to be had in developing it, even if you are building it for purely busi-
ness reasons.

What do you want your chatbot to do? Engage with customers and collect
their emails for your list? Sell products and manage payments? Provide infor-
mation? It’s not too difficult to do any of these and to do so on Facebook
Messenger, Skype, etc. If you concentrate on the core task for your chatbot,
you’ll be up and running relatively quickly. And the great thing about the
recent proliferation of chatbot creation platforms is that you can do this with-
out any background in coding.

Once you have your chatbot carrying out its basic tasks, you can then make
the structure more sophisticated. In fact, it’s not really until the chatbot is
active in public that you’ll get the feedback you need to really accelerate its
development. Once it is live, you’ll have a lot of incoming data to look at. You
will be able to see what people want, what questions they ask, what parts of
the conversation they want to return to and so on. In my case, I created whole
new branches of interactions that I hadn’t even thought of once I saw where
users wanted to take their conversations.

There are several chatbot building platforms out there and it is worth having a
look around to see what the possibilities are. The newer ones tend to be the
easiest to use. This section of the book utilizes the SnatchBot.me platform,
because you can create your bot for free there and the process is very intui-
tive. But a lot of the information in Part 2 – such as connecting to channels
like Facebook Messenger – is transferable to other chatbot building plat-
forms.

To begin the process of creating your own chatbot, you'll need to register on
SnatchBot's website, which only takes a moment. Don’t worry about your
data, their platform uses high-grade security that complies with all regulatory
mandates. Once the verification process and registration is complete, you'll
be able to access your dashboard.

47
Part 2 Building Your Chatbot

When you first log in, you will be taken to the dashboard, which you will use to
see all your chatbot analytics.
Select My Bots from the options in the blue box on the left of the screen and
then click the Create Bot button.

A pop-up will open for you to name your chatbot and choose whether to use a
pre-defined template or not.
48
Part 2 Building Your Chatbot

Next comes the real work. Your bot will communicate with users through
‘interactions’. Interactions come in a wide variety, which give you have a flex-
ible and powerful range of choices for what you want the bot to do, without
requiring any programming skills.

Later you will want to take advantage of the various complex functions and
built-in features such as action buttons, translations, payment processing,
email extraction and many more. All these additional features can be added
without you needing any technical knowledge. But for now choose Interaction
Type and pick the first option, Bot Statement.
49
Part 2 Building Your Chatbot

You will need to give the interaction a name.

Tip
Pick names for your interactions that will allow you to distinguish them
from each other. Later, as your bot becomes more complex, you will want
to be able to accurately identify the right interaction to make connections
with. So, ‘Yes to email’ is better than a simple ‘Yes’.

Now, you’ll see editing tools appear on the right-hand side of the page. Start
by filling in the text for the bot’s statement on reaching this interaction. For
example, if this is going to be the starting interaction, your text might read:
‘Hi, I’m a chatbot for the company, would you like to talk to me about our
latest offers?’

50
Part 2 Building Your Chatbot

Now, you’ll see editing tools appear on the right-hand side of the page. Start
by filling in the text for the bot’s statement on reaching this interaction. For
example, if this is going to be the starting interaction, your text might read:
‘Hi, I’m a chatbot for the company, would you like to talk to me about our
latest offers?’

51
Part 2 Building Your Chatbot
Under the heading Bot Structure, you’ll see your interactions listed. Select
one, then click the ‘connect this interaction’ button on the right to make a
connection. Each time you click this button a new connection editing section
opens. You may have to scroll down to see them all.

The connection sections take the form of a logical statement: if x then y. It’s
very straightforward and intuitive, and covers all the possibilities you might
encounter. The best way to learn about your options here is to open the drop-
down menus and look at what is available to you. You’ll see that you can
make connections on the basis of simple responses, but also by using the
Exactly Matches option, you can identify particular phrases in the user’s
response and make a connection to the relevant interaction.

You can also direct the conversation according to what the user hasn’t said.
You’ll spend some time at this stage creating interactions and connecting
them. Note that you do not need to save your bot-in-progress, it is automati-
cally updated as you work.

As you build your bot, you’ll want to test your progress regularly. This is
straightforward. The button you need is in the top right corner of the screen.

52
Part 2 Building Your Chatbot

When you’ve finished the test, just click away from the test page and you’ll
return to your bot.

After configuring your bot, you will want to deploy it for users to access it.
SnatchBot use the term channels for the platforms hosting your bot, e.g.
Facebook, Skype, Email, API and Web Chat.

For information about configuring your bot for a particular channel, see
Chapter 14.

53
Part 2 Building Your Chatbot
Chapter 13: Using a Template from the Bot Store
The quickest way to get started is to use a template or user-created chatbot
from the bot store. Not only does this get you up and running right away, it
also allows you to enjoy the benefits of chatbots that other people have found
to be successful. As this part of the SnatchBot platform is open to users to
submit their bots, it is constantly growing and improving the functionality of
the chatbots on offer.

Let’s suppose you have a restaurant and would like a chatbot to handle your
booking system, freeing busy staff from having to answer the phone and
make the booking (as well as eliminating human error in placing the booking).
There are two ways of finding out if there is a ready-made bot in the store that
might be ideal for your own uses. You can type “restaurant” in the search bar
on the top of the screen, or you can check the categories on the left. Both will,
in this case, lead you to one chatbot in particular, the Restaurant Bot Template.

You can find the SnatchBot Store on their landing page or on your Dashboard
(here).

54
Part 2 Building Your Chatbot

Search for the type of bot you want either using the search option, top right,
or from the category list on the left, then just click anywhere on the chatbot to
access a template that creates a new bot on your account.
Customise your chatbot from the template
To edit the chatbot to make it perform the tasks you need, click on Edit this
Bot. You’ll then be asked to confirm that you want to create a new chatbot
using the template. Say yes and give the new chatbot a name. It will then be
imported into your collection of chatbots. You can view it and edit it in the
usual way by clicking the My Bots option from the menu on the left.

The new bot screen will open automatically for you to rename it and start editing.
Naturally, you’ll want to alter the chat to reflect your business. But that’s a very
easy edit and the basic structure is already in place for you. It’s as simple as that! 55
Part 2 Building Your Chatbot
Chapter 14: Placing your Chatbot on a Website, Facebook Messenger and
other channels
How to deploy your chatbot on a webpage
Placing your chatbot on a webpage is straightforward. Choose your chatbot and
then select Channels from the left-hand column options. There you will see the
option to edit web channel data.

A box will open for you, which has the code you need to add to your website in
order for a ‘Bot’ button to appear in the lower right corner of the page. Click on the
copy button to select this code.

56
Part 2 Building Your Chatbot
It’s that simple and that powerful! Now you have a chatbot on your webpage.
If you’d like to adjust the positioning and look of your chatbot, you can do so by
using the codes in the same window. Scroll down to steps 2 and 3, which allow
you to place the webchat URL where you like and to use your choice of CSS
styles. There are also some further options available for you to customize your
chatbot and how it looks to the user. To do so, click on Customise Webchat.

Now you’ll see options to add in Agent Name, Signature, Contact, Website, Sup-
port Link and Email Link. But the real power in customization comes with click-
ing Advanced Options.
Here, a new page pops up, allowing you to alter the size, colour scheme, greeting
message and more. These are simple to adjust and allow you to create the exact
tone that you want for your chatbot.
Connecting channels
Your chatbot is perfect for handling communications that come to you via popu-
lar platforms. In fact, that’s often the reason you would want to create a chatbot.
It’s very easy to assign a bot to a ‘channel’ e.g. to respond to emails, Facebook
messages, Skype messages, etc. To connect your chatbot to your desired
channel/channels, open your bot editor and click Channels.
57
Part 2 Building Your Chatbot

Choose the channel you wish to connect to and click Add.

Each channel has its own way of registering bots, but SnatchBot has made the
process straightforward. Follow the steps on this page of their website and
you’ll connect your chatbot in no time. Note that when you click on the links
they provide, you’ll be taken to the relevant channel websites to put in your
name, login, etc. Basically, what you are doing is registering with the channel
and getting a set of OAuth keys that SnatchBot can use on your behalf.
For example, to connect to Skype, you’ll open a Microsoft window and – assum-
ing you haven’t already done so – create an account as a developer and register
your bot.

58
Part 2 Building Your Chatbot

Once you are done, try sending yourself a message through your channel (e.g.
Facebook Messenger) and you’ll see how powerful a tool this is!

59
Part 2 Building Your Chatbot
Chapter 15: Extracting emails, URLs, addresses and other data
Gathering data such as emails, addresses and phone numbers from your
chatbot’s conversations is quite straightforward. It’s simple to do so once
you’ve set your chatbot up correctly.
You’ll be familiar with the fact that you can choose certain types of interac-
tions in creating your chatbot.

In this case, we’ll take the example of email extraction, but the process is the
same for every kind of data.
Let’s suppose you’ve created a chatbot with an email extraction step and it
has had hundreds of conversations with users. You’ll get an overview of all
these conversations when you click Reports.

60
Part 2 Building Your Chatbot

On the reports pages, you’ll initially be shown all your data. To filter it to just
extracted emails, choose email from the Extracted Data dropdown menu.

This will give you a list of email addresses from across all your chatbots. If
you want to limit the list to just one chatbot, choose it from the bot dropdown
menu.
61
Part 2 Building Your Chatbot

To obtain an Excel file of these emails, simply click the Generate Excel button.

62
Part 2 Building Your Chatbot
Chapter16.Handling payments
How to use a SnatchBot chatbot to manage your customer payments

Much has been written about the abilities of modern chatbots. Among many
other capabilities, their ability to process payments opens many doors for
both brands and consumers. Accepting payments makes it possible for
brands to take many of the functions that previously required standalone
apps and insert them into high-usage messaging platforms like Facebook
Messenger. Doing so greatly increases customer reach and this chapter will
teach you how to enable this game-changing feature in the SnatchBot plat-
form.
Setting up payment information
First, you’ll need to set up the back-end of your payment system. SnatchBot
integrates Paypal for payments. To do this you will need to create an App and
set up a unique Paypal Client ID and a Paypal Secret Key within the Paypal
Developer portal. A guide on this is available here. Once you have that infor-
mation, you’ll need to enter it within the SnatchBot platform. To do this, navi-
gate to your profile by clicking your name along the left-hand menu (second
option from bottom).
Once there, select Payment System, enter your Paypal App ID and Paypal App
Secret, and click Save.

63
Part 2 Building Your Chatbot
Configuring your chatbot

In this case, the payment will be for a gym membership. We’ll begin by creat-
ing an interaction where the payment system will live by creating a Bot State-
ment Interaction. The payment button lives within the Gallery card, which can
be selected from the menu at the bottom of the interaction. Once selected,
click the blue plus button in the middle to create the Gallery card. Next, fill out
the information and image required for the card.
With this info input, click ‘Add Button’, select Payment, and name the button
(‘Purchase’ in this case). Enter the price, as well as the interactions that will
be generated if the payment succeeds or fails (note: you’ll need to create
these as individual interactions) and the size you’d like the window to display
at. Click ‘Add Button’ to complete the card.

Finally, we’ll need to set up the interactions that lead to the Payment card. In
this example, we’ll direct users from the first greeting message. A quick reply
of ‘Membership’ is input, and set up to direct users to the Payment card.

That’s it! Give it a quick test in Facebook Messenger and it should appear like
below.

64
Part 2 Building Your Chatbot

Using the steps above, SnatchBot lets you monetize your chatbot and start uti-
lizing conversational commerce in minutes. This is a crucial function for those
readers who want their chatbot to be a conduit for income generation.

65
Part 2 Building Your Chatbot

Conclusion
Hopefully, you are now up to speed with everything you need to know about
chatbots. Most importantly of all, you should be able to make one for yourself,
even if you have no background in computing. Only a few years ago you’d have
had to either known something about coding or spent a great deal of money to
create a chatbot for your business, say. But as is often the way with our species,
we’ve figured out ways of making tasks easier and more intuitive.

Chatbots have been heralded for some time as the coming innovation that will
change our cultural world. They haven’t done so yet and in fact, a certain
amount of disillusionment has set in. I wrote an article for a tech blog in 2017,
for example, that was rejected, because, ‘I couldn’t get past the idea that chat-
bots are going to spread rapidly.’ This is understandable but mistaken. There’s
often a pattern to the adoption of new technology that often follows the same
path: massive excitement, new buzzwords, new hope, then a decline of interest
and outright public cynicism when technical challenges are seen to continue to
be a major barrier, but then comes a revival of enthusiasm after the problems
are solved and a widespread, game-changing adoption of the new technology.

My belief is that at the time of writing (late 2017), chatbots have reached the
revival point of this pattern. Dozens of companies have answered the challenge
of making chatbot creation simple, intuitive and yet extremely capable of deal-
ing with the tasks allotted to them.

The resulting explosion in chatbot creation – hundreds of thousands this year,

millions next – is going to change the world. In all sorts of arenas, we are going
to encounter chatbots, and chatting with them will soon be second nature to us.

On a longer timeframe, I believe the chatbot road is the more likely one to lead to
AI than the robot building one. In a way, by making chatbot creation so easy
(and for free in most cases) we are crowdsourcing the question of how you get
software to understand meanings from human language. And applying millions
of minds to this task is the right way to create breakthroughs.

66