Whether Chatting with a Bot or Person, Language Matters
February 7, 2017
People are either loving or hating chatbots. Taco Bell, Facebook and even CNN are using chatbots to help with transactions and customer support. But love ‘em or hate ‘em, to be effective, chatbots still have to speak their customer's language particularly if they’re helping with customer service.
Chatbots are great for many things just like quick machine translations such as Google Translate works for some communications. It seems chatbots have been really effective at helping customers who want multi-channel support immediately, but as with all things, they leave a few gaps in service. What, for example, happens when the customer has a question or concern the little bot can’t answer with an algorithm or blanket response?
Many businesses are asking the same question as they gather data about bot use and customer services outcomes. Customer Think, one of the leading customer service publications, wrote a great piece in May 2016 about how bots can kill customer experience and user experience. At least right now. As Salesforce CEO Marc Benioff was told Customer Think, we don’t know where the AI technology that fuels bots will go. It could be great, but for now, it’s not enough to replace models of customer service that we currently know.
One thing we do know, is that a key component to customer service is being able to communicate with customers in the language they speak. It’s possible to use emoticons for some basic bot communications. Emoticons are somewhat universal although I must confess, my own misunderstanding of emoticons has opened some interesting conversations. That being said, even if you communicate well with emoticons, they can’t explain everything. They certainly can’t solve every problem a customer has. At some point, the customer is going to need a response in their language.
When it comes to communicating with people, the proper use of language is key. As the author of the Customer Think piece notes, bots need to excel at basic natural language detection. This means that for now, some iteration of customer service as we understand it will continue to exist. That means that companies will need to continue providing customer services in the languages their customers speak.
If you’d like more information on how to streamline accurate, multilingual customer support across multiple channels, please contact us here. We can help.