Chatbots With a High Emotional Quotient

Sure, there are a dime a dozen chatbots out there. Think of Siri, Google Assistant, Alexa, and their like.

However, what truly matters for businesses these days, with chatbots replacing customer service reps, is to have a chatbot that recognizes your customers' mood and offers reactions accordingly.

Consider this:

Recently, Amazon’s Alexa Prize, a multimillion-dollar bot powered by conversational AI, made one of the most significant goof-ups of this century.

When a curious little kid queried the Amazon bot about Santa’s whereabouts, this is what the bot replied:

“Santa Claus is the most elaborate lie ever told.”

The result: Amazon pulled the plug on this particular bot.

Why? Because the bot failed to sympathize with the kid's emotions.

But then, in hindsight, the question is: how can you find fault with a machine? Machines are programmed to offer automated replies, aren't they?

Put another way: Chatbots are not emotional beings and are not programmed to answer as per the emotional needs of humans. But given that they are driving customer service departments, customers expect them to have a twitch of human touch.

Since then, Amazon has come up with a challenge called the “Alexa Prize.”

As per the challenge developers need to create a bot using Alexa, that could talk to a human for 20 minutes without messing up. In return, you stand to win $1.5 million prizes, plus other grants and awards.

Long story short: The biggest AI challenge the chatbot developers are facing these days is, how to humanize bots so that they could hold real, emotional conversations with customers for 10-20 minutes at a stretch, without messing up?

Okay, before we start talking about the importance of developing a chatbot with high EQ quotient, let me first brief you about the current market scenario on chatbots:

One Side of the Coin: The Proliferation of Chatbots Across Industries

It’s being predicted that between 2016 and 2021, the chatbot market will grow at a compound rate of 35.2 percent. According to business insider, by 2020, more than 80 percent of businesses will have small to significant chatbot automation implemented.

Currently, live chatbots are being developed for platforms such as Facebook Messenger, Slack, Telegram, and more.

And with AI greasing the chatbots’ wheels, you can expect more intelligent chatbots to be rolled out in the future.

The Flip Side: Customers Prefer Humans Over Chatbots

As discussed above, developing a bot that genuinely understands the context and depth of a conversation and one that can hold a conversation for long has proven to be a challenging task for developers. Sure, you could engage a bot for 100 micro-conversations. But engage a bot a little longer, and it will be lost.

Not surprisingly, 71 percent of the U.S. consumers prefer to interact with humans over a chatbot. Sure, it’s bad news for 80 percent of the businesses that are already using chatbots or are planning to do so by 2020.

So, How to Embed the Elusive EQ in Chatbots?

While brands are trying hard to 'technify' their customer service operations, they need to ensure that their efforts don’t drown out the emotional element.

Simply put, the bots should be able to recognize customers' mood and respond accordingly. It should be programmed in a way that it can separate emotions into six different categories such as anger, sadness, happiness, fear, disgust, and surprise, and respond to their feelings appropriately.

And given that consumers favor humans over bots when it comes to customer service, several leading companies are working at providing engaging chatbot experiences to the customers to help them make a human connection with the brands.

Examples of Chatbots With High EQ Quotient

H&M

This multinational retail company has created a chatbot that acts as your sales guide. The bot helps mobile customers find suitable outfits based on their gender and style preferences. If the customer doesn’t like the recommendation, the bot will give another suggestion. In addition, the bot also throws up hipster vocab with text on an occasional basis. And when a customer has made up his mind on a particular outfit, the bot guides them through that process as well.

Replika

A lot of current AI chatbots such as Siri and Google Assistant might help you predict the weather, or for that matter, might help you schedule an appointment. However, Eugenia Kuyda’s chatbot Replika won’t do either of these things for you but would love to act like your best friend with whom you can confide personal matters.

Yes, you read that right! And nearly 2 million have already downloaded this app and are using it as their digital companion. Using a neural network, the chatbot holds a one-on-one conversation with the users and, over time, learns to speak just like them.

Here, I walk you through four must-haves to create emotional chatbots:

#1. Personalize

It could be in dash or dollops. You decide. But then personalize, no matter what. Why? Because cloning an existing bot won’t help your business cause in any way.

Your bot needs to stand out in its own unique way. If your bot repeats the same answers like other bots, you will quickly piss customers off. For example, if you're bot constantly prattles, “Sorry, I do not have an answer to this question," just like other bots do, the customers will feel disappointed, never to return. As per research, 1 out of 3 customers won’t do business with you even if they only had a single bad experience with you.

Which is why it’s vital to offer personalized experiences to customers. So give your bot a competitive edge by adding a USP to it.

#2. Great Customer experience

We cannot stress enough the importance of customer experience. And in case you are not aware, customers are more than happy to pay for better customer experience. In fact, 71 percent of the companies acknowledge that strong and positive customer experience is their safest bet. So, while developing a bot, a conversation designer should think of offering unique experience paths that could help the user reach their end goals.

#3. Respond Aptly

Think about every word and image that you intend to add it to your bots. Consider including emojis and GIFs to give your bot an extra edge. For example, if a person is looking sick and sad, your bot should be able to, at once, figure that out and send out an apt emoji.

#4. Old Data Drive Conversations

Bots connect the dots looking back. Simply put, by storing personal information from old customers chats, bots can help businesses offer enhanced customer experience. For example, offering something as simple as personalized recommendations based on previous shopping behavior could bring a smile on your customer's face.

Conclusion

Yes, it’s a big deal these days if your bot can chat with humans without goofing up. It’s a big deal if your chatbots can empathize with human emotions and react accordingly. So, if you think your chatbots are not human enough, it’s time to take a look at your chatbot strategies and see if you can humanize them.

If you've already integrated a human element in your bots, go ahead and share your experiences below.