Conversational design: Why customers don’t like chatbots, and what you can do to change that

For businesses, chat interfaces, such as chatbots, are extremely useful. They can help customers 24/7, solve issues fast, and reduce the workload for employees. However, businesses that want to implement chatbots also face a big challenge: customers don’t particularly like chatbots. In the following, we look at the issues behind human-bot-interaction, and explore why conversational design is the key to solving them.  

Chat interfaces are on the rise! From their very humble beginnings in the 1960s, chatbots since have come a long way to become useful tools for companies. Chatbots are used in customer service, they can help clients to buy products or services, and smart bots can even help companies to gather data and generate leads. Overall, they can attend the clients’ needs faster, cut costs, and reduce the workload for employees.

It is thus not surprising that more and more companies are using chatbots. On Facebook alone, the number of bots grew from 11,000 in 2016 to 300,000 in 2019.

However, when thinking about implementing chat interfaces, companies face a huge challenge: customers don’t like them very much. In the following, we look at the reasons behind this reluctance, and how conversational design can be the key to a successful human-bot-interaction.

Awkward conversations: Why customers don’t like chatbots

Many businesses assume that their customers don’t like chatbots because they prefer to interact with humans. However, research shows that most users are in fact quite open to bots. According to a study by customer support company Pidas, almost 89 percent of customers are interested in digital conversations with chatbots. Provided they are helpful.


And that’s exactly where the problem lies! Many chatbots have been programmed to solve specific tasks or to lead customers on a certain conversational path, but they have not been designed to resemble an authentic human conversation. Therefore, the experiences many customers have with chatbots are very awkward, and, to be frank, the opposite of helpful.

Not understanding the customer

If chatbots are not equipped to follow human conversations, they will also have a hard time understanding the customers’ requests. For example, if a chatbot is only programmed to answer the question “Can you tell me more about the status of my delivery”, but a user asks, “Where is my package?”, it might take the bot a while to understand what the customer really wants.

Transferring customers to agents

Another issue in human-bot-interaction is that a customer is not transferred at the right time to a human agent. If the re-routing happens to early, the employees end up still having too much work. If it happens to late, customers might have already had a frustrating experience with the bots, and feel that they haven’t been helped.

Chatbots can help your team, but they cannot replace the employees, so it’s crucial for companies to figure out at what point customers are being referred to a human representative, and also how that is communicated. Otherwise, the users’ chatbot experience can be quite rocky.

Customers develop negative bias around chatbots

This is not only frustrating, it also is not particularly helpful, as it might take customers a lot of time to receive the assistance they need. After a few negative experiences, users start forming a negative bias against chatbots. A recent study from the Technical University of Darmstadt and the Copenhagen Business School has found that once users have had several negative chatbot interactions, they are more aggressive and less compliant with bots, to a point where they refuse to follow a chatbot’s instructions to solve their problems.

For businesses that want to use chatbots, but are faced with hesitant users, this is obviously a considerable hurdle. The good news is: there is a solution—conversational design.

🤖Tipp: Read more about how chatbots can improve your customer service.  

What is conversational design?

Conversational design refers to designing the flow of conversation between a chatbot and a human user. It combines:

  • artificial intelligence (AI)
  • natural language processing (NLP)
  • user experience design (UX), and
  • writing and linguistics.

Depending on the type of bot, voice, or motion design can also be a part of it. The idea of implementing conversational design into chatbots is to make the interaction between bot and human as natural and as smooth as possible. There are four key features that contribute to a successful conversational design.

1. Understanding

In order for a chatbot to have a successful conversation with a human, understanding is very important. However, understanding goes beyond simply knowing the meaning of a word or a sentence. It requires understanding their implications. In linguistics this is called pragmatics, and refers to the understanding of the social, cultural and situational context.

For instance, if a person tells another person of their household that they are out of milk, it is implied that someone will have to buy more milk. Consequently, a chatbot that uses conversational design, will also not simply say “It’s too bad that you are out of milk.”, but also add: “Would you like me to put it on your digital shopping list?”

2. Interaction

The more chatbots mirror human conversations, the higher their acceptance with useres will be. To a certain extent, even if customers know they are talking to a bot, they expect a human-like presence on the other end. And this means: interacting!

After all, just like in a regular conversation with another person, nobody wants to listen to a long monologue. The same goes for talking to chatbots. If they just present customers with a long message and no interaction, it can be off-putting. This means that conversational design has to focus on the interactional aspect of the conversation between humans and bots. It can include for example shorter chat messages from a bot that give the customer a chance to respond and ask questions.

3. Context

Understanding the context of the customer is another important aspect in conversational design. Chatbots need to understand where the user is coming from.

If somebody writes a message from a smarpthone, the response might have to be shorter than if they are sent from a desktop computer. The chatbot should also be aware of the fact that mobile users are possibly on the go, and might not be able to access as much information as desktop users at home.

4. Natural language

Users respond better to chatbots if they sound more like humans. This includes following the general norms of a conversation, such as saying “hello” or “goodbye”, or introducing themselves with a name and using personal pronouns.

It is also critical for bots to show empathy, just like a human agent would. If, for example, a customer has technical issues, the chatbot should use phrases like “I am so sorry that you have problems, let me see how we can solve them as quickly as possible.” It will make the customer understood, and also more willing to follow instructions.

💡 Find out more about the difference between rule-based and AI bots

Chatbots for businesses: benefits of conversational design

The benefits of conversational designs are quite simple: the smoother the conversation between bot and human is, the better the customer experience. Consequently, once you have been able to implement high-quality chatbots in your company, you will also benefit more from them.

You will be able to:

  • 👍… improve your brand’s image
  • 🦘… reduce bounce rates
  • 🤝… increase brand loyalty
  • 👜… drive conversions

What type of chatbot do I need for my business?

Before you start building a chatbot for your company, there are a few things you should consider first.

  1. Know your customers, so you can tailor the conversational design around them.
  2. Think about the persona you want to develop for your bot. It should not only be built around the customer, but also represent your brand.
  3. Determine what you want to achieve with your bot. Should it answer FAQs, direct customers towards the right department, or solve complex problems?
  4. At what point do you want the bot to transfer customers to human agents? This process has to be smooth and also help your team.

Chatbots need to be tailored to the customers.

So, determining what the “right” chatbot is for you really depends on your businesses’ goals. Our “Messenger Communication Platform” therefore offers you different solutions. If you are looking for a quick and efficient chatbot for simple tasks, the Chatbot Builder by MessengerPeople is a great place to start.

For an AI-based chatbot, the deep integration with Chatlayer by Sinch allows you to build virtual assistants for all of your channels. The conversational design will allow you to have a higher personalization in your dialogues, and more natural conversations.

Chatlayer by Sinch Website screenshot

👉 Tipp: Do you want to start your own AI chatbot project? Just schedule a live demonstration with Chatlayer. The team will guide you via web demonstration through the product, give you expert advice, and answer all of your questions.

More articles on this topic


Have a great idea for customer service via messaging apps and want to talk about it? Message us!