man explaining

Best conversation design strategies to get the most out of your chatbot or voice assistant

Let’s be honest, most chatbots and voice assistants kind of suck. They often don’t understand what you say and what you want to achieve. They are clumsy and robotic… silly at best. We have all been there. We have a question about a certain product and we decide to reach out to customer service. We stumble upon a chatbot and decide to give it a try. It often starts with a fun greeting.

quotes-jade-greenI’m company x virtual assistant! How may I help you?

But as soon as you enter your question it leaves you clueless.

quotes-jade-greenSorry, I didn’t get the question. Please try again.

It often replies. And so you try again… and again… and again… until you end up picking up the phone and calling customer service directly.

But some brands get it right. They know how to deliver human-centric interactions that help users get their job done when they reach out to the AI Assistant. So what do these brands have in common? If they all have access to the same technology, then why do some of them pull it off to create valuable interactions? Some leverage conversational AI technology to create natural conversations, but others leave users in frustration.

In this article, you will learn about five great conversation design strategies that you can implement to make your chatbots or voice assistant more helpful, natural, and human-centric when talking to your customers.

1. Bot personality is key to speaking in a consistent tone of voice

Personality goes a long way. One of the best strategies in conversation design is to make sure that your AI Assistant speaks in a consistent tone of voice. This makes it likeable and more trustworthy. The personality must make sense for your brand and the way it relates to your users and what they are trying to achieve. To create a bot persona you need to follow the steps in the bot personality worksheet. This worksheet can be downloaded from the Conversational Design Institute -Academy and it guides you through the process.

You will look at your brand values and the use-cases that your AI Assistant is going to handle. Then you look at the audience that is going to be interacting with your AI Assistant. This will help you develop a bot personality.

The questions to ask yourself are:

ic-checkWhich conversations are we going to have?
ic-checkWho are we going to have them with?
ic-checkAnd who is best to have these conversations on our behalf?

That last question is what defines your bot personality.

2. Use personality content to build a relationship with your users

Now that you have a bot personality that allows you to write in a consistent tone of voice, you can start with writing personality content. There is a misconception that personality content is just silly jokes and chit chat. That it serves no further purpose. But that is simply wrong.Want to learn more on building a great chatbot_-2

Personality content builds a relationship with your users. It makes it more fun and engaging. When people feel engaged and have a relationship with the AI Assistant, they become more forgiving when it doesn’t know the answer. That’s a big win for when you are starting your project, because at the beginning your AI assistant will be struggling a lot.

It is important to find the right balance between real content and personality content. You want to make sure it doesn’t become a gimmick, which could reflect poorly on your brand. That’s the last thing you want.

3. Framing throughout conversations

Framing takes the pressure off your natural language understanding. When you first go live, your AI Assistant will not know the answer to most questions. It often has to utter the dreadful words:

quotes-jade-greenI’m sorry, I don’t understand that question.

A trick to tackle this issue is by framing better. You see, when your AI Assistant says:

quotes-jade-greenHi, my name is Johnny and I know everything. What do you want to know?

You are inviting your users to ask any question he or she can think of and there is a big chance your AI Assistant won’t know the answer. So what’s a better way of doing this? Well, you guessed it, by framing better!

quotes-jade-greenHi, my name is Johnny. You can ask me anything about coffee. What would you like to know?

By framing the word coffee we are increasing the odds of our user asking us a question about coffee. Assuming we are properly trained to answer questions about coffee, we can increase the odds of the AI Assistant understanding the user and providing value.

4. Explicit confirmations to show empathy and gain trust

We often talk about empathy in design. Empathy is really important if you want users to engage with you. But what is it really about? Many believe that empathy has to do with feelings. That’s not the case. It’s about showing that you understand the user’s situation. By showing that you truly understand the context of your user, you are gaining trust and are increasing the completion rate of your dialogues.

An explicit confirmation is a simple trick to show empathy. It’s a type of confirmation where you let the user explicitly confirm that you properly understood it. It’s also often called a mirroring statement.

For example, your user says:

quotes-jade-greenI broke my phone and I want to know if I’m covered.

With an explicit confirmation, your AI Assistant sounds very empathetic.

quotes-jade-greenOK. So you broke your phone and you want to make sure it's covered by your insurance. Is that correct?

You are demonstrating that you truly understand your user and this builds tremendous trust. Your user will now follow along in the conversation because he believes you will guide him to the right solution. Sometimes the explicit confirmation seems a bit elaborate.

That’s why we only use it when:

ic-checkConfidence level is low
ic-checkIt’s about financials
ic-checkIt’s an emotional topic

5. Turn-taking to speed up the conversation

Last but certainly not least is turn-taking, which speeds up the conversation. Your AI Assistant will improve dramatically when you pay close attention to this concept. Turn-taking means that you want every message to end with a question.

The question signals the user that it’s his turn to speak. Does that make sense? Start the conversation with empathy and continue by asking simple, binary questions. The concept of turn-taking gives the user easy questions he’s happy to answer, since the explicit confirmation at the beginning leads him to believe that you are guiding him to the right solution.

Final thoughts on implementing conversation design best strategies

There you have it. These are 5 best strategies that you can implement today to get more value from your AI Assistant for both the user and the organization.

Make sure that you have a clear personality, write personality content, use proper framing throughout the conversations, use the explicit confirmation to demonstrate empathy and gain trust, and then use proper turn-taking to keep the conversation going.

If you want to learn more about the best strategies in conversation design, read the best tips for chatbot design from Conversation Design Experts in this article.