Chatbots: A Complete Guide (2020)

What are chatbots? Why are they such a big opportunity? How do they work? How can I build one? This in-depth guide is all you need to learn more.

We’ve heard a lot about Chatbots in recent years. Businesses have found numerous purposes for chatbots – in customer support, sales, marketing, HR, and other services.

Let us now dive a little deeper to understand what chatbots are, what they can do for us and how they are built.

What is a Chatbot?

Chatbots are software programs created to engage with users automatically. A chatbot can respond proactively to messages containing specific words or phrases by offering a predefined response. It could also use natural language processing and machine learning to analyze and understand an incoming message and provide an appropriate response in real-time.

Much of the data that we have suggests the rise of chatbots in various aspects of customer experience. In a paper published in The International Conference on Internet Science, one participant described the chatbot experience as:

It saves me the hassle of having to place a call, have to wait to speak to a person, then trying to get the information from that person. It also saves time in having to look through tons of text to find answers.

According to Gartner, AI chatbots will handle up to 85% of customer service interactions within just a year. Convince and Convert reports that 15% of American adults have used a chatbot last year; 37% see themselves using one to get a quick answer in an emergency.

Benefits of Chatbots


Chatbots streamline various customer interactions with your services while enhancing customer satisfaction. It is an essential part of delivering customer responses quickly, accurately and efficiently. The revenue impact – both the savings in human time along with revenue from happier customers – can be significant. That said, if or when a conversation gets complex for a chatbot, it can be handed over to the human support agent.

Chatbots help you scale customer service with fewer employees. More and more customers are turning to e-commerce for their shopping needs. The complexity and volume of customer inquiries are increasing constantly. Chatbots help process basic inquiries without involving human support agents, leaving them free to deal with more complex issues. The greater the inquires deflected the greater the time and money saved.

Chatbots help in decreasing overall ticket volume. 70% of customer queries are usually the same repetitive questions that can be easily handled and resolved by a chatbot. At a basic level, chatbots deflect common queries by redirecting the customers to support articles. But a custom chatbot solution can also automate common transactional and custom interactions - anything from checking an account balance, changing a seat on a flight or topping up a data plan. A chatbot solution reduces the volume of service tickets and creates a more efficient customer service process.

Chatbots help in lowering the wait times for your customers and creates more engaged agents. Your customers will love you if you could lower their support wait times. With chatbots deflecting common customer queries, your human agents have time to handle higher-value customer queries. Your agents will tackle the more complicated matters faster and easier than ever before.

Chatbots help you leverage your current self-service content – FAQs and Knowledge Base. As mentioned above, 70% of customer queries are usually the same repetitive questions that can be easily resolved by redirecting the customer to any self-service content like FAQs and support articles in your knowledge base.

Chatbots help overcome shopping cart abandonment. In today's world of short attention spans, a customer is at risk of forgetting to check out the items that they added to their shopping cart. A chatbot could automatically message customers who have abandoned their purchases.

Chatbots help you to convert more website traffic. By proactively assisting customers in need and by answering common queries, a chatbot can help you convert more of your inbound website traffic.

Chatbots help you generate more qualified leads. Landing pages have been in a state of stagnation for many years now. Adding a chatbot to a landing page is a great way to ensure that your potential leads can get their questions answered when they are evaluating your business. It’s a great way to take advantage of all the effort that you put into promoting your landing page.

Chatbots work 24/7. Chatbots can process customer inquiries during offline hours, assist shoppers from any time zone and continue to operate unassisted even when you are out of office.

Chatbots ultimately help improve cost savings and drive up revenue. They directly benefit your bottom-line. There is a cost-saving as your team’s time and energy are better utilized. A study published by Business Insider shows chatbots cut customer service costs up to 29% – 46%, depending on your business. At the same time, opportunities for revenue increase as your agents shift their focus from reactive to proactive and jumping on opportunities for upsells or special offers. Chatbots can also automate up-sell and targeted promotions. As customer satisfaction (CSAT) grows, the benefits of greater customer retention will also be seen on the bottom line.

How to Build a Chatbot?

The following instructions are designed to guide you to plan and build a customer service chatbot.

1. Determine the Goals of the Chatbot

Identify a list of questions that make up 80 percent of the volume of incoming inquiries. Determine whether you want your chatbot to interpret questions very narrowly – deflecting fewer questions from the contact center but being highly precise – or broadly, creating a higher rate of deflection at the risk of answering incorrectly.

2. Evaluate and Choose a Channel

Text-based chatbots can live on any communication channel. Whether that’s a traditional mobile carrier channel (SMS), a messaging app (Facebook Messenger, WeChat), social networks like Twitter, or a live chat widget embedded on a website or mobile app. Whichever channel you choose, the chatbot's capabilities are limited to what the channel offers. Sometimes the right channel of communication opens up new perspectives for better customer service.

3. Conversation Design

Chatbots are about a continuous flow of conversation that allows for any number of responses between the chatbot and the customer. When compared to the User Experience of mobile apps or websites, the messaging channel is story-based or flow-based, where all previous interactions are always visible to both parties.

This means the customer's queries and chatbot's responses can never be analyzed in isolation – they are always part of a larger conversation. What Information Design is to mobile apps and websites, Conversation Design is to designing chatbots.
You may want to design variations of the same message for frequently occurring queries. This is a technique called Randomized Prompts, where the chatbot uses a word or phrase variation to essentially say the same thing. This makes the whole experience feel less robotic and more human, which is something we should strive for in chatbot design.

4. Design Chatbot Integrations

Integrations open up endless possibilities for the chatbot. If you are already using a self-service or help desk platform (web or voice), you may want to integrate it with the chatbot.

For example, if your CRM (Customer Relationship Management) identifies your customer and provides order status information, then this information can add context to the customer's queries. Similarly, your chatbot can integrate with your help desk software to automatically create a ticket for unresolved queries.

Another popular chatbot integrations is with meeting scheduling. The chatbot can collect the necessary information and schedule a meeting with the customer. Calendar invitations are sent to the respective parties and kept in sync.


5. Building your own Chatbot

An interaction with a chatbot involves two stages:

1. Analysis of the user’s request – a computer program breaks down the incoming message to identify user intent and match it with relevant information or an action to be performed.

2. Respond to the input – based on the previous analysis, appropriate action is taken and a response message is delivered to the user.

Types of Chatbot

Based on the technology used, a chatbot may be one of two types.

1. Decision-tree based: The chatbot is a codified version of a decision-tree with numerous branches that can be traversed based on the user's input. The chatbot provides a visitor with a specific set of options to choose from. Choosing an option leads the user to the next set of functionality or information.

2. AI-powered: Using natural language processing to understand the customer's unique input and provide a response fetched from a predefined library. The most advanced chatbots use artificial intelligence and machine learning to imitate a real conversation.

Chatbot Development

If you do not have an existing data set to train your chatbot, you are better off with the decision-tree based approach. The latter also allows you to retain more control over how a question is interpreted, which matters in customer service, as you want to minimize the probability of giving out a wrong answer.

If you selected a platform based on Machine Learning (ML), you will provide this platform with your example sentences for every possible customer intent. The more examples you provide, the better the ML algorithm will learn, and the better it will learn how to distinguish between different customer intents. This learning can be a continuous process. As more customers use the chatbot, their inputs can be fed to the machine learning algorithm for further improvement.

    A chatbot's responses can take the form of:

  • A static text
  • Information retrieved from a knowledge base or other resource
  • Link or reference to a page where a person could learn more
  • Custom programmed actions. E.g. signing up for a newsletter, scheduling a meeting, etc.

6. Iterate and Refine


Work on the chatbot is never fully complete. To get the most out of a chatbot, it is important to monitor customer queries and refine the chatbot’s responses. Typical revisions include rewording certain responses as you review follow-up clarification questions from your customers that wouldn’t have been necessary if the bot’s answer had been clearer.

You may need to add new use cases if the designed use cases do not cover the majority of customer requests. To ensure continuous customer satisfaction it is important to view the chatbot design as an iterative process: Gather data, review it, and apply it to your chatbot’s design.

Key Factors to Consider before Choosing a Chatbot Platform

1. Chatbots Capabilities: Enterprises should look for the ability to deploy not only single-purpose chatbots, e.g. newsletter sign up, meeting scheduling, etc., but also multi-purpose chatbots that can communicate with multiple systems and complete a variety of tasks like lead generation, customer support, answering FAQs, etc.

2. Multiple Channel Support: Enterprises should evaluate chatbot platforms where the bots can be deployed across their websites, mobile apps, Slack, Facebook Messenger, Telegram, Skype and other potential customer touchpoints.

3. Natural Language Processing Support: Enterprises should look for the chatbot platforms that support Natural Language Processing (NLP) to provide the best results in understanding the user intent and responding with the relevant information.

4. A Ready-to-use Repository of Bots: The platform should provide a collection of ready-to-use chatbots for common functionalities without the need to create one from scratch.

5. Ability to Integrate with other Platform: The platform should offer the ability to integrate and exchange data with 3rd party services. This can enable fetching of useful information about the customer to serve them better. Also, any information collected during a bot interaction can be recorded to external services for future analysis.

6. Industry Experience and Domain Knowledge: Lastly, we need to ensure that the platform providers have considerable industry and domain knowledge. While there are several bot-building platforms out there offering a whole lot of features, enterprises need to identify which feature will matter to them. HappyFox is a leading Help Desk and Live Chat platform. HappyFox Help Desk has been rated as the best Help Desk platform for 4 years in a row. Talk to our product specialists to learn more about our chatbot service offering.

Key Takeaways from this Guide

  • Chatbots are software programs designed to proactively engage with website visitors.

  • Bots can help you achieve a wide range of business objectives like increasing sales, generating leads, streamlining customer support, overcoming the shopping cart abandonment and can even automate various business process activities.

  • The best customer service set-ups are a hybrid of automated solutions with personal human interactions. From the customer’s perspective, the journey is a smooth series of easy-to-use questions and responses with easy-to-access off-ramps to jump to a real person.

HappyFox Chatbot makes it incredibly easy to launch a fully custom chatbot solution personalized to your unique business needs. Talk to our Product Specialists to learn more about our unique product offering.

Still have questions?

Talk to someone smart, quickly.