The Next Bot: the future of Chatbot, AI and Messaging

IQUII
IQUII
Published in
7 min readJun 14, 2017

More than 2.5 billion people use at least one app to exchange messages and today spend 90% of Mobile time between email and messaging platforms.

When people began to spend a lot more time on messaging apps than social media, Bots became (and are becoming ever more so) the instrument through which to access and interface with all types of equipment.

Credits: Chatbots Magazine

Facebook then initiated the birth of a new platform for applications that offer users an ecosystem of micro-experiences from immediate access.

In fact, platforms such as FB Messenger and Slack have transformed those that were previously (mainly) entertainment tools into powerful tools for users and businesses.

But what are Bots?

A ChatBot, in simple terms, is a software governed by a set of rules (and Artificial Intelligence) designed to simulate a smart conversation with the user through a conversational interface such as a chat.

It offers a service — which can be fun, playful, informative, and supportive — available to users on the major messaging platforms (Slack, Telegram, Facebook Messenger, etc.).

We usually consider Bots as a unique category capable of performing a specific task: conversing.

In fact, Bots are positioned within a spectrum of possibilities and functions, to the extreme of which two macro-categories can be identified:

  • Simple Bots, rule-based and designed to accompany the user along a specific and defined conversational path;
  • Intelligent Bots, equipped with Artificial Intelligence and machine learning, able to learn over time and understand each query.

With over 11,000 Bots already developed for Facebook Messenger, examples of best practices from businesses, from retail and eCommerce to Banking to Sport are beginning to emerge, which we analyze in more detail in our free Business Bots eBook.

Vision

The evolution started by Mobile and Messaging Apps has not yet finished, and the shifting of focus and attention from users to conversation, to instant and on demand responses is fueling new needs and new requirements that technology and business are starting to answer.

The direction we look to in order to outline the (very near) future of our interactions is that of the big platforms, or better still, big ecosystems, micro-experiences enclosed within a single complete platform.

An example? WeChat.

The WeChat application, developed in 2011 by Tencent, which today has more than 700 million active users per month, is much more than an instant messaging app: it is a mobile micro-world with powerful features and services.

Social Networking, instant messaging, groups, and photo and video sharing are not the only functions enclosed in one application.

Follow the official accounts of famous brands and celebrities, share Moments, book an appointment with a doctor or dog groomer, shop at your favorite retail stores, and make payments and bank transactions without ever leaving a single app: this is the evolution of a context of fragmentation and dispersion in which over 4 million applications in the Store make it difficult to retain consumers.

Today, even the large western companies are moving in this direction. Facebook, Apple, Skype, Google, Kik and Telegram, to name but a few, are focusing their attention on the inclusion of services and functions within a single platform and the creation of conversational systems that can learn through Artificial Intelligence.

Inspiring

“We are on the threshold of something that could cause a profound change […], and that something is artificial superintelligence.

Artificial Intelligence, at one time, consisted of giving commands to a box. There were human programmers who developed “manually”, with great fatigue, pieces of knowledge. These expertly built systems were useful for some purposes but were very “fragile”, and difficult to expand. Essentially, you only get what you insert.

But since then there has been a revolution in the field of Artificial Intelligence. Today, we are focused on Machine Learning. Instead of handwritten representations and characteristics of knowledge, we create algorithms that learn, often from unrefined perceptual data. In practice, the same thing a child does.

The result is an Artificial Intelligence that is not limited to a single sector: the same system can learn to translate from one language to another, or learn to play every game of the Atari console.”

What is Artificial Intelligence?

AI, by definition, is the ability of a machine to artificially reproduce intelligence and assume behavior as a rational agent, making decisions, and performing actions resulting from context perception.

The expression Artificial Intelligence was coined in 1956 by US mathematician John McCarthy to indicate a new discipline whose goal was to “make machines do things that would require intelligence if they were done by man”.

The arrival of Mobile has projected us definitively into the era of Artificial Intelligence, giving people the chance to chat with anyone in real time through simple textual platforms. At the moment in which conversation reaffirmed itself as the dominant interface, Bots emerged as the first expression of AI, to enhance interactions with companies and meet their daily needs without leaving a single platform.

In recent years, several steps have been taken in the field of AI, so that developers today are able to embed Artificial Intelligence, without great complexity, into their own ChatBots.

Bots equipped with AI are based on machine learning, so they are able to understand the language and not just a specific set of pre-set commands. In this way, Bots learn from interactions and become more intelligent.

If the interaction interface has its strength in total simplicity, creating a satisfying experience for the user is not an insignificant task.

Some factors need to be considered, such as:

  • Absence of a standard navigation;
  • Flow optimization;
  • Integration with the platform and with APIs;
  • Error checking;
  • Analytics.

To create a ChatBot that can handle conversations with your users and optimize the management of your customer service for example, it’s necessary to know the operation and characteristics of the platforms, the needs of the target user, the type of requests and information to be handled and so on.

Strategy

So if the starting point is the purpose for which the Bot is created, from helping the user in the purchase process to offering real-time support at any time with their requests, to create a Bot that is truly functional and effective the most important focus is on User Experience.

While the operation of a ChatBot and its stream of communication is the simple implementation, creating a user-conceived system that has the functionality to generate value for the company and for consumers requires an in-depth study of the data, needs and conversations that allow for constant optimization.

“The most successful Bots will be the ones that users want to come back to regularly and that provide consistent value.” — Matt Hartman

Implement a Bot for Facebook Messenger

The creation of a ChatBot for Facebook’s messaging platform is developed in 4 macro steps:

  • Set perimeters for what to respond to and how the Bot responds before then installing it within a Facebook Fan Page after approval by Facebook;
  • To install the Bot on Messenger, the Bot must first be associated with a Facebook Fan Page to create an FB App ad hoc through the Facebook Developer Program;
  • Once installed, the Bot will be able to intercept every single message that a user sends privately to the Facebook page;
  • If the Bot recognizes a template that has been defined in the message sent by the user, it will respond immediately to the user by giving them the answer to their question.

The development of a Bot is also an evolutionary process that allows you to progressively optimize and sharpen the intelligence and rules. In managing the first level of service and support, this approach allows you to drastically reduce your workload by issuing the Bot’s answer to established questions.

The basic approach of implementing a Bot is of an iterative type and starts from an analysis of macro patterns and then evolves and refines by defining the rules, analyzing the results obtained, and consequently redefining the rules.

insideIQUII

In this period we have been able to work on and deepen our expertise of the world of Bots (not even summer has stopped us!) And we are putting together several different projects that we will tell you about in updates to follow.

We also produced an eBook focused on the world of Bots, in which we analyze:

  • The state of the art;
  • The perception that consumers have of these new interlocutors and the possibilities that come from them;
  • The best Business Bots in different markets, with examples and best practices;
  • How companies are embedding this relational tool into their strategies;
  • Why Bots can influence Customer Service and consumer brand experiences;
  • What Bots can do for companies and how to best use them;
  • How to Implement a Bot for FB Messenger;
  • How Bots impact on business processes.
http://chatbot.iquii.com

--

--

IQUII
IQUII
Editor for

We’re not another Digital Company. We are forward thinkers: we humanize businesses and make engagement and loyalty real in #Finance #Sport #Retail