Best practices

Everything you need to know about a chatbot!

Dec 12, 2021 7 min
Afbeelding van mens met robot hoofd breed

Everything you need to know about a chatbot!

Leestijd: 7 minuten

Do you remember Chatman? The Hi chatbot, that worked super fast with MSN? If you don’t remember the chatbot, maybe you still remember the jingle from the commercial. The Hi marketing campaign in 2005 unlocked a new era. “The buddy-bot is a piece of software that can communicate naturally, as if it’s a living person.” But Chatman wasn’t the first chatbot on the scene. Do you remember Clippy? The little paperclip that answered all your Microsoft Word questions?

We’ve seen a ton of different kinds of chatbots emerge over the last several years. You’ve probably interacted with a chatbot at least once, and you probably know at a high level what a chatbot is and does. But do you know all the ways you can use chatbot for customer service? This article explains the different types of chatbots, what kind of goals a chatbot might have, and we’ll share some chatbot trends that we’re expecting to see play out in 2022.

What is a chatbot?

The meaning of the word chatbot comes from a combination of the words ‘chat’ and ‘robot’ – it’s actually an automated conversation partner. A chatbot can understand a question and associate it with the correct answer. You can implement a chatbot on a website and on social media for a variety of different reasons. Companies often use chatbots to answer frequently asked questions, or to generate leads.

Generally speaking, there are three different types of chatbots: a button-based bot, an AI-chatbot, or a hybrid model. Which type of chatbot will be most useful to your customer service team depends on the complexity of the questions your customers are asking, the way your customers behave, and the time and resources you have available to implement a chatbot.

A button-based chatbot

A button-based chatbot is the most simple way to get started with a chatbot. The customer follows a specific conversation path with the chatbot; the chatbot gives the customer a number of options to choose from, and the chatbot provides the associated answer. This type of chatbot is most useful when you get a lot of the same questions all the time and want to answer these automatically and as quickly as possible.

Simply put, a button-based bot has a fixed conversation flow that a customer or lead follows. The person chatting can’t ask any open questions, and the chatbot can’t learn along the way as more conversations are had. A button-based bot is ideal for identifying customers, for example. The chatbot can ask for details like name, email address and phone number. Customers respond with these details, and the chatbot then transfers the conversation to an employee.

An AI-chatbot

An AI-chatbot is powered by artificial intelligence, also known as AI. This software receives data, and determines the best answer based on the data, without human intervention. This software is able to self-learn and make decisions, based on information it’s already learned.

Another term that’s relevant to understanding how an AI-chatbot works is Natural Language Processing (NLP). NLP is used to transform human language into something that a computer can understand. You teach the chatbot how to answer a number of frequently asked questions by adding what’s known as question variations, meaning different ways in which a person might ask the same question.

This gives the AI-chatbot a foundational set of questions it can answer. As the chatbot has more conversations, the chatbot learns to understand more variations on the same question and can consistently improve its ability to answer questions. If the chatbot doesn’t understand a question, the question is registered as a mismatch. It’s then up to you to determine if the mismatch needs to be added to the knowledge base, or if another answer should be selected as the correct response.

The difference

The main difference between a button-based chatbot and an AI-chatbot is that an AI-chatbot can answer an open question, while a button-based chatbot answers based on which button is pressed. The AI-chatbot is more intelligent than a button-based bot, and can (over time) use NLP to answer questions independently.

A hybrid chatbot

A hybrid chatbot is a model where the chatbot works closely with a live service agent. The chatbot asks for the necessary details and may provide some preliminary information, and then transfers the conversation to the appropriate agent. This collaboration between the chatbot and the agent ensures your customer service team can function even more efficiently. The customer service agents now have more time to focus on complex questions while the chatbot is busy gathering details.

Blog afbeelding vragen product

What are the different goals for chatbots?

You can use a chatbot for a number of purposes, including but not limited to customer service. We’ll share a number of options with you!

Customer service goals

The first type of chatbot you thought of was probably a chatbot for customer service. You’ve probably chatted with a bot from a customer service department before – big companies like Bol.com, KPN or A.S.R. insurance have already been using a chatbot for years. The customer service chatbot is most often implemented to direct you to specific information, or offer you alternatives for specific products or services. A customer service chatbot can be used in every industry, including healthcare, financial services, e-Commerce, education or government agencies.

Marketing and sales goals

A chatbot can be a digital employee for more than just the customer service team, but can also help out the sales team by generating leads! The chatbot can suggest products or services, or ask for customer details and send these to the sales team. Imagine the following situation: a potential customer is looking at a particular product or service on your website. The chatbot proactively begins a discussion and asks “Can I help you with anything?”

The potential customer indicates what they’re looking for, whereafter the chatbot provides them with a link to the product page that has when they need. If that’s still not quite it, then the chatbot transfer the conversation to a human agent! This approach allows you to offer fast, extra helpful service and keep potential customers on your website for a longer amount of time.

Internal communication

You can also use a chatbot to support your organisation’s internal communications. An HR chatbot can support the HR department with questions about your company, questions from potential job candidates, or questions from your existing team.

Let’s walk through a concrete example. Imagine you’re getting a lot of questions about an open position, but these questions are coming in outside of office hours and you risk losing a few good candidates. By using a chatbot on the web page where the job is posted, the chatbot can give potential candidates more information about the job, but also ask for their information and request an upload of their CV or cover letter. When the recruiter shows up for work the next morning, that perfect candidate can get a call back right away!

Blog afbeelding chatbot bouwen

What are the top chatbot trends?

Chatbots have exploded onto the scene this year. Research shows that 1 out of 5 consumers have already made a purchase via a chatbot. In 2022, a number of sources predict that 70% of customer interaction will be managed by AI. It might not surprise you, then, that the technology behind chatbots has evolved to meet this growing demand for performance.

Trend 1: More payments via chatbot

In 2022, it’s predicted that the chatbot will continue to play a role across various digital contact touchpoints in a customer’s journey. Until today, chatbots mostly contribute to the post-sales process in answering questions. This year, however, we expect a shift to take place such that the chatbot takes a more active role at the beginning of the journey. This has been termed ‘conversational selling’, or making a purchase during a conversation.

Consumers have gained trust in chatbots, and appreciate the help of the digital employee when making a purchase. Research reveals that 47% of users would pay via a link provided by a chatbot. Moreover, paying via chatbot is considered even more user-friendly. It’s easy and fast, there aren’t any shifty pop-up windows or new tabs opened. Plus, chatbots now have more integrations with payment systems including Mollie, Moneybird or Apple Pay.

Trend 2: Chatbots with a ‘human touch’

The personality of a chatbot is an important aspect to the digital colleague. The chatbot’s persona is what makes it authentic and communicates a part of the company. Research shows that 74% of consumers want the chatbot to introduce itself at the beginning of the conversation. About 53% of consumers felt more positively about the company when the chatbot responded in a fun, humorous way.

We can thus rightfully assume that more ‘human’ chatbots will be making their way to the main stage. The bots with a ‘human touch’ ensure the customer remains engaged in the conversation and is less likely to abandon the chat. Artificial intelligence chatbots ensure that customers are better understood. AI-chatbot will recognise behaviour patterns in consumers, and thereby be able to help them even more quickly.

A great example of a personal chatbot is Chatbot Max from the AMP Group. Chatbot Max a copy of Max, the Business Consultant. The chatbot makes the same choice of words as Max, making the chatbot even more personal. Maybe too personal: “We can see from the responses of some customers that they’re not always aware that they’re chatting with a bot. Even I might be fooled”, says the real life Max.

Trend 3: Better collaboration…

There are even more chatbot trends for 2022 that we’d love to share with you! Do you want to read the rest of the trends for 2022? Read our blog: “7 chatbot trends that will steal the show in 2022”.

Want to know more?

Request a free demo and find out more about Watermelon.

request demo