Chatbots — A Crash Course for Newbies

Altaf Rehmani
Not So Technical
Published in
13 min readAug 2, 2018
Introduction to Chatbots — Source expertsystem.com

This article is for anyone who wants to get upto speed on chatbots, specifically for a lot of non-technical managers and marketing folks who 1) want to know what chatbots are and 2) look into implementing chatbots for their division 3) who would like to tap the power to reach the millions who are hooked onto messaging apps.

As we already know a lot of us are spending a massive amount of mobile time in messaging and chat applications. 1.4 billion people used a chat app in 2015 — and young people were the most active among them.

Source — BI Intelligence, Google : 2017

With chat bots, you can reach consumers in fun, useful, and meaningful ways in an environment that presents the least friction possible. There’s not much to download, no icons to add to the homescreen, and no memory hogging. And your chatbots can have access to have direct access to more than 100’s of millions of users on popular messaging platforms. Are chatbots the new apps?

What are Chatbots?

Here is the textbook and Classic definition of a chatbot:

A chatbot is a software service, powered by rules and sometimes artificial intelligence, that you interact with via a chat interface (text or voice).

In simple words — “bots” are essentially software applications with human conversation interface, mostly via text. Unlike apps with several buttons or links as interface, you can navigate bots through conversational ways via texts. Bots don’t need to be downloaded, they live on the servers, and typically run on messaging apps.

It’s highly likely you’ve engaged to a chatbot, even if you didn’t realize it was one. From iPhone’s Siri to mobile messaging services, chatbots are surfacing everywhere with increased frequency. Chatbots are applications that communicate and perform basic tasks, whether it’s answering questions or helping customers purchase a product.

Their core responsibility is to streamline interactions between services and people. Chatbots are beneficial for both parties: developing chatbots is cheaper than training and hiring human customer service agents for the company, and customers often prefer a brisk mobile interaction over talking with someone in person or with the call center.

Consider this statistic from Gartner,

Artificial intelligence will amount for 85% of customer relationships by 2020.

From planning your next vacation to finding the right dress for your wedding party, no matter what you want to do, there’s likely a chatbot for that.

A bit of history and evolution of Chatbots.

It all started in the 1950s, when scientist Alan Turing started exploring the idea of having “meaningful conversations” with machines. But today we’ve actually been able to combine the technology and artificial intelligence to create effective chatbots that people want to use. Some versions of the chatbot are still powered by certain keywords, meaning that if someone didn’t ask a question featuring one of these specific “trigger words,” the bot was unable to respond to the person. However, advances in machine learning have made chatbots better listeners, and more accurate.

History of Chatbots

Today chatbots can learn conversational and written cues to know what someone is trying to ask when said or written in a variety of ways. A consumer can ask a chatbot the same question in a variety of ways and the chatbot will understand what they’re trying to say.

Few examples of how brands use bots.

Are chatbots just a fad? What are enterprises and brands doing with chatbots? Lets touch upon how companies are investing in this new conversational interface which enables a new line of communication with leads and customers.

  • The CNN chatbot helps users retrieve news. You tell the CNN bot what you are interested in, be it the Olympics or US presidential election and it will happily oblige with a selection of news stories.
  • Disney keeps releasing bots along with its movies, the Zootopia bot, for instance, lets you discover and solve crimes. You can talk to the Miss Piggy bot to know more about her!
  • PitchBot is a bot designed to simulate you talking to investors about your venture to raise funds.
  • The Hyatt hotels bot can talk to you about your reservation status, check-in times, available facilities and price.
  • Hello Jarvis bot is a personal assistant and Healthtap answers questions about medical issues (yikes!).
  • Banks are racing and even competing against each other to roll out chatbots to make day-to-day banking easy.
  • TravelFlan is an upcoming travel bot from a HK based company which promises hassle free getaway.

There are chatbots to talk to about weather (Poncho), stock markets (Finance bot), shopping (Nordstrom), travel (Kayak) and almost anything under the sun. You can even use a Chatbot for random chats to create stories or just to kill time.

Types of chatbots:

Typically there are two broad types of chatbots in terms of how they work behind the scenes.

Rule based chatbots: One is a type with predefined rules, trigger words and conditions which are used to respond responses and solve the specific problem they are designed to solve (or information they are supposed to lookup). This is what the majority of the chatbots are today.

AI based chatbots: The other types are one which deploy AI technologies behind the scenes — learning what the users ask over time, try and understand the context and sentiment and deploy “machine learning” techniques to predict and analyse the intent of the user interacting with them. These bots tend to become smarter over time and learn from their right and sometimes incorrect responses and self-improve. Typically these bots try to be more generic and serve as personal assistants answering a broad range of questions and able to have varied conversations.

The Technology behind AI based Chatbots:

AI based chatbots may use a number of technologies behind the scenes in order to function more intelligently.

Natural Language Processing (NLP)

A Chatbot typically only understands only the words that are fed into its knowledgebase. A bot, unlike humans, can’t understand different contextual usages of a word if they are not programmed into it. Natural Language Processing (NLP) is the ability of computer programs to make sense of common written and spoken language.

One of the best examples of NLP today is its application in Google Search. Type something like, “How is the weather in Hong Kong” in Google and Google will provide the weather information in a nice card at the top of the page. Type, “what are the chances of rain in Hong Kong” and it will give you the same result. This is an excellent example of natural language processing in practice. The NLP programming of the Google bot is mature enough to understand that although the question and wording are slightly different, both cases require it to pull up the same set of information.

To achieve good NLP, chatbots are fed all variations of the frequently asked questions and answers manually. Algorithms are then written to help the bot fit in synonyms and scope is left for fuzzy matching as people generally ask the same questions in a number of ways.

Machine Learning (ML)

Machine learning is the ability of computers to learn by themselves without any hard coding. The vast majority of chat bots on the market today are not using machine learning. Most are using a retrieval based model where possible questions and answers are hard coded into a flowchart-like structure. Thus most of today’s chat bots are not really learning to answer questions themselves using machine learning but are still learning to understand questions better with the help of NLP. For chatbots that are using the machine learning like IBM’s Watson or Google Assistant, the difference is they have a lot more data to work with and train.

Sentiment Analysis

Although machine learning is essential for a chat bot to expand its responses over time and NLP assists in understanding the questions and statements entered by the user, these two alone are not enough for a chat bot to have full contextual understanding in a conversation.

While NLP takes a sentence and breaks it down into codified elements to allow the chat bot to understand the content, sentiment analysis seeks to rate the context of a sentence or series of sentences on an emotional scale from perfectly positive to perfectly negative. When a user adds the sentence, everyone is going to hear about the kind of service your company has provided me, it is of no consequence in NLP. However, sentiment analysis will take the statement into context with the preceding conversation and decide whether to respond with thank you, or go into damage control mode.

Note that none of these are absolutely required — It is very much possible to create bots which still use rigid rules, trigger keywords and still serve useful information via lookups or function calls to their users. You can even have bots programmed to do something in between set rules and implementing AI technologies.

Quick Predictions behind Chatbots:

  • Chatbots will be responsible for cost savings of over $8 billion annually by 2022, up from $20 million in 2017 (Juniper Research)
  • Chatbots will power 85% of all customer service interactions by the year 2020 (Gartner)
  • By 2020, the average person will have more conversations with bots than with their spouse. 30% of web browsing will be done by voice (Gartner)
  • 47% of consumers are open to buying items through a chatbot, and 37% would buy items from Facebook (HubSpot)
  • 67% of consumers worldwide used a chatbot for customer support in the past year (Business Insider)

Chatbots — Enabling conversational commerce.

Conversational commerce is a new opportunity that brands can leverage to deliver commerce and customer service. Using widely popular messaging apps, it is a personal, fast and convenient way to communicate with your customers.

Conversational commerce fits into a larger trend of humanizing the interactions and relation between consumers and businesses.

H&M chabot : Source Yopto Voice

So what can you achieve as a business via conversational commerce?

  • Deliver convenience and decision support on the go, at scale
  • Improve your customer satisfaction
  • Integrate (e)commerce into your customer service

The personal nature of messaging apps in combination with the simplified conversational interface and the on-the-go availability make messaging into a perfect platform for commerce and customer service.

In the context of a ecommerce / retail business, bots can enable conversational commerce by implementing the following:

  • Product recommendations
  • Convenience of reservation or purchase
  • Relevant and actionable offers in real time
  • Alerts and updates on inventory
  • Easing the payment process

How do you plan or approach your plan for a Chatbot implementation:

While chatbots are becoming increasingly popular, that doesn’t mean you can build one type of chatbot for everyone. 60% of chatbot users are between 13 and 19 years old, with more females than males. The service seems particularly suited for Millennial and Gen Z-ers, who grew up using many of these on-demand technologies, but it can be difficult for brands to reach other, older demographics through chatbots.

When developing a bot, companies need to be aware of their target customers and who they are trying to reach with a new technology.

Chatbots are amazingly efficient for both users and brands — they cut costs for companies by winding down the expensive interactions with traditional call center agents or customer service representatives, and they also make it easier and more convenient for customers to communicate with brands, streamlining interactions.

Most customers would rather get a quick recommendation and purchase an item on their mobile phone instead of sifting through online reviews and trying to buy products in person or by calling a representative. Of course it always depends on the product but more and more we’re seeing customers wanting to make quick and easy purchases on their phones.

While some types of businesses naturally lend themselves to bots such as service-oriented businesses, there are plenty of other companies that can reap immense benefits of bots. A general rule of thumb is that if your company is regularly communicating with customers or could increase communication to improve sales and customer experience, using a chatbot could be a good option.

Suggested Approach to build your next Chatbot.

  1. Find your audience

More than 3.2 billions consumers were using messaging apps by the end of 2015. The top four messaging apps have outgrown the top 4 social media apps (in monthly active users). Consumers are now ready for conversational commerce via their messaging app but you will need to define which messaging apps your customers are preferring today.

The most simple route to this massive audiences on messaging apps is to connect directly with the leading messaging apps. You can use the tools and interfaces provided by each of the messaging apps themselves (for instance your standard Facebook Page admin interface).

2. Define the best use cases

The possible applications of conversational commerce differ according to our industry, company and market situation. For a retailer it can be about giving styling and product advice, while for a travel business a full fledge concierge service via messaging might be the most desired conversational support. Pick your use case carefully and make them complimentary to your other channels. Start small on your use cases and learn and improve before scaling up to wider use cases or audiences.

3. Integrate your services and systems

“App fatigue” is a new challenge for many companies as consumers are downloading less apps and using only a few of them frequently. Mobile messaging are one of the most popular platforms in time spend and number of active monthly users.

With conversational commerce you can mould your services and open them for usage on the messaging platforms. Most messaging platforms allow you to bring your services into the conversation via bots and commerce features.

4. Add intelligence and scalability

Chat bots are a good example of how you can add intelligence and scale to your conversational commerce. While platform like Telegram, Kik and Wechat already opened their bot functionality , Facebook recently joined the with the Messenger Platform.

To allow more natural conversations with your chat bot, NLP and AI solutions allow you to train your bot in natural language. Many frameworks and technologies will help you to build and teach natural language chatbots. While simple menu-driven chatbots are fairly easy to design and develop, natural language driven chatbots will require sufficient expertise and training by specialists in NLP and AI.

5. Build a Human + Bot team

Although bots can take over large parts of the conversation, access to human live agents might still be required to complete your conversational services. Human agents and bots can perfectly work aside. In our conversational commerce platform we have integrated a unique bot-to-human feature which allows consumers to escalate from a bot to a live agent. Once the conversation is completed between the consumer and your live agent, the bot continues the conversation again with the consumer.

6. Learn and improve

Conversational commerce is still in the early stages and therefore many learnings still need to emerge. Using a mix of live human experts and bots you can gradually launch your conversational commerce. Clearly define your objectives and success criteria for your conversational commerce. On the consumer side, communicate your service features and restrictions well to manage expectations and service levels. Conversational commerce requires a ongoing process of optimisation based on the conversations and the intelligence provided via dedicated conversational platforms.

Business benefits and Limitations of chatbots:

Business Benefits:

  • Future of customer service
  • Efficiency improvements
  • Conversational commerce
  • Lower costs for business
  • Easier to develop and deploy

Limitations:

  • Speech recognition limits
  • Language interpretation limits
  • May not be able to cover all possible cases
  • Learning the wrong things
  • Security
“A robot named Pepper holding an iPad” by Alex Knight on Unsplash

Platforms to build Chatbots :

Here is a list of some popular Chatbot plaforms for getting your feet wet without being too technical or having any coding knowledge.

Chatfuel — no code platform for marketers and non-technical people to create bots for the facebook platform.

Motion.ai — Now part of Hubspot suite of tools provides a similar drag and drop interface with facility to call your own services to respond to messages. If you are a marketing — you should definitely explore this package.

Flowxo is a platform which allows you to create and manage chatbots easily.

Microsoft bot framework provide a set of tools and API to do end-end delivery and deployment of bots using a combination of Azure, Bot connector and developer portal.

Amazon provides “Lex” the tech that powers Alexa and combination of services which make similar things possible and Google / IBM have their own services with similar offerings.

A list of top bot platforms can be found here

Future of Chatbots:

Bots allow users to interact with services as if they’re sending a message to another person. There’s a smaller learning curve for the user, they are simple to connect to, and they live right where 100’s of millions of users already spend a lot of their time — inside messaging apps.

Conversational commerce is bringing back the business-to-customer dialogue where it’s been missing most: shopping online.

While the chatbot hasn’t evolved to the point where it could completely pass for a real person yet, the messaging app has evolved to improve a part of business that never quite carried over from brick-and-mortar to the online store. Using a combination of machines and humans — the user experience of interacting with a brand can be enhanced at the same time lowering costs and performing the same functions over a period of time.

And it’s only possible now because messaging apps as a whole have experienced explosive growth compared to most social apps, offering users a more engaged and private communication channel than any social network.

  • Chatbots will be here to stay and will be increasingly utilized by companies across all industries and are only going to get more powerful
  • Important to understand the current limitations of Chatbots
  • Chatbot created must be useful and serve specific purpose for your business or customer.

Blogs and resources used for this article

  • Letsclap.io
  • Rocketbots.io
  • eMarketer.com
  • Chatbotnewsdaily.com
  • Venturebeat.com
  • Shopify.com

Hopefully I have given you enough to get started in the right direction on chatbots. Connect with me on linkedIn, follow me on twitter or just drop a line to me via email with any feedback on the article.

Altaf Rehmani is a technology Innovator , helped various businesses with Digital transformation projects, Agile Evangelist and a champion of applying technology to enable business growth. He lives in Hong Kong and can be reached via email or twitter. Please leave your feedback and a clap if you have liked this article.

Other articles which may be of interest:

Managing High Performing teams

Common Mistakes in Agile Implementations

Scaling Agile in Enterprises

Applying AI in The context of eCommerce

--

--

Altaf Rehmani
Not So Technical

Technology Innovator,Digital IT Mgr and Agile Evangelist | Certified Scrum Master. I love innovation,startups and help businesses with their digital strategy.