How to make a stupid chatbot seem like a smart assistant

Developing a brand used to mean creating a logo and choosing color palettes, but those days are numbered. Thanks to chatbots and intelligent hardware devices such as Amazon Alexa and Google Home, visual brand elements are starting to take a backseat to brand elements that are completely focused on what people will read or hear.

Companies building products on chatbot platforms aren’t always able to control the appearance of what they’re communicating to their users — it’s all about the words. This applies to an even greater extent with smart speakers because the words are spoken and you have nothing to show to your users.

Product designers who are focused on these platforms are going to have to learn how to shift from a visual mindset to a world where picking the right words and communication style is what will help them to stand out from the competition. These trends are already starting to lead to the rise of a new job description: conversation designer. This individual is a mix of a writer and a UX designer, but instead of living in the Adobe Creative Cloud applications, they live in a text editor.

When chatbots and other smart assistants are doing their jobs — whether it’s scheduling a meeting, booking travel arrangements, ordering something, or whatever — efficiency is key, so there may not be many opportunities for your brand to shine through. But when they get something wrong, get confused, or are just having small talk with the user during the quest to get something accomplished, then the work of the conversation designer can come to the forefront.

Designing the conversation

To define your brand personality, you have to decide the answers to some basic questions. Does the bot have a sense of humor, or is it strictly business? Does it feel like a friend or simply a personal assistant? Is it supposed to be identified as being from a specific country because of the way that it spells or speaks? You can easily understand why the personality of a bot would differ based on the business that it’s working for; a bot that represents a large bank would need to have a completely different communication style than a bot for an independent record store.

Let’s take a look at some fictional examples to see the contrast between different communication styles. First of all, notice the difference in how you can decide to start a conversation between a user and your bot.

1. “Hey, how’s it going? I’m Trevor, and I’m the best thing that ever happened to you. Well, maybe not, but I can still help you out. What can I do for you on this fine day?”
2. “This is our automated support system. Please enter your request.”

The first example is personal, informal, and humorous. Since you’re communicating with a bot that has a name, the interaction feels more conversational and natural. In the second example, all of the personality is erased and no effort is made to make the assistant feel human.

No matter how intelligent your bot is, there will be times when it doesn’t understand something that was said. Let’s now see how a response could be designed when the bot is in this situation.

1. “Sorry, I just had one of those moments where I temporarily forget how to understand English (don’t tell anyone). Can you try saying that again in a different way? Thx.”
2. “Sorry, that wasn’t understood. Please try again.”

The first example uses humor to try to relieve any frustration that the user might feel about the fact that the bot didn’t understand their request. The second example gets straight to the point.

In our last set of examples, notice two different ways that you could decide to conclude a bot’s conversation with a user.

1. “If you need any help or suggestions in the future, then you know where to find me. I think you and I have a good thing going.”
2. “Thank you for using our automated support system. Please visit our website to get additional information about our products.”

The first example encourages an ongoing relationship in a personal way, while the second example tries to get the user to explore the website on their own and stop using the bot. Who would be excited to have regular communication with something described as an automated support system, anyway?

The first examples in these three scenarios might not be the right fit for your brand, and that’s OK. An informal and humorous conversation style isn’t right for every audience. In fact, your users might even hate it. Design what works for your brand.

Making your bot famous

Startups may be forced to create personalities for their bots from scratch, but larger companies have the option of hiring celebrities to be the voices for their brands. GPS companies such as Waze, Garmin, and TomTom have made it possible for users to get turn-by-turn voice directions from stars such as Morgan Freeman and John Cleese and even fictional characters like Homer Simpson and Yoda. This is the kind of thing that can turn a boring drive into an interesting and entertaining experience, giving you a positive impression of the brand.

When words and voices are doing all of the talking, the appearance of the person behind the hired voice for a brand doesn’t even matter. Instead of finding someone appropriately good-looking, companies will focus much more on choosing someone who’s known for having interesting things to say and is instantly recognizable by their word choice, which is a complete shift from the way brands typically pick celebrity representatives.

Whether you choose to use a celebrity as the voice of your brand or create a personality for a bot from scratch, it’s time to start asking yourself an important question: If I had a conversation with my company or product, what would it say and how would it say it? Before you know it, your brand might have to start speaking for itself — literally.

Matty Mariansky is a cofounder of Meekan by Doodle and an expert on conversational design.

Above: VB Profiles Chatbots Landscape. (Disclosure: VB Profiles is a cooperative effort between VentureBeat and Spoke Intelligence.) This article is part of our bots landscape series. You can download a high-resolution version of the landscape featuring 197 companies here.