Today, 44% of consumers expect quick answers to their questions. Imagine what would you feel if you had to call a helpline and wait for hours to talk with a consultant, not even knowing whether he’ll be able to help you. Step into the shoes of a client who sent a complaint email to a company and has to wait for the response longer than he’d expected. There are many more such examples, just like there are many more potential Avenger movie endings. So let me show you how to make a chatbot that will improve your customer service and become your best employee of the year 🙂
Effective client communication is a real challenge for many businesses. Especially if we consider the huge number of channels clients use to contact companies. So there’s no wonder why customer service is one of the most important chatbot functionalities.
Let’s start with some basics to illustrate how the chatbot customer service environment looks like. According to statistics Facebook proudly shares, people exchange 20 billion messages with brands per month and 58% of users feel more comfortable writing messages than calling. Customer service is a major part of user-chatbot interactions – it appears in (depending on different studies) 50% of all types of conversations. Businesses no longer see chatbots as “gadgets”. They see them as a growing source of customer satisfaction and as an effective tool for saving money. Need some proof? Here you go! 🙂
Are you convinced? Think about adding a chatbot to your customer service if:
The company’s Facebook fan page is usually the first point where customers interact with a brand. If we’re designing a chatbot to automate customer service processes, we need to consider several important aspects.
Messenger is one of the most popular messaging apps in the world, so it’s a good idea to prepare chatbot user paths in different languages. Of course, the choice of languages depends on the countries where the brand sells its products or services. If we give users the option to choose a language, we’re using e.g. a language variable. However, there may be words that have the same translation in both languages. Remember about this when you design interactions for different conversation flows.
This happened to the Wroclaw Airport chatbot. There were numerous words like “boarding” or “status” that have the same meaning in both Polish and English. What’s more, we also have to consider situations where the user starts writing his message but hasn’t yet chosen a language. In this case, we have to pick a default language or write the chatbot messages in both languages.
But if users enter a foreign word, the easiest way is to prepare a response message, which will direct them straight to a moderator. There’s nothing more annoying than a chatbot that keeps responding in a foreign language you don’t know. It’s a much better idea to inform users that the chatbot is currently available only in an English language version, and that they’ll soon be contacted by a moderator.
Apart from languages, there are also other challenges chatbot designers have to face. Especially when it comes to customer service. For example, when the company (e.g. a hotel or restaurant chain) has many business locations which offer different services. If we’re talking about a gym network, users can ask about the time at which they can go to Zumba classes. In this case, they’re interested in a specific gym location. If each and every gym has a separate fan page, then a multi-chatbot is a perfect solution. Because it allows every fan page to adjust the chatbot content to specific gym locations. You can read more about multi-chatbots in our article on how to manage many brand fan pages in one place.
When designing chatbots for customer service, we have to determine the accuracy with which the chatbots will provide information and solve user problems. We should also examine how broadly we can integrate our chatbot with the client’s system.
The solution we recommend is, of course, to connect the chatbot with the client’s IT system. Thanks to this, the chatbot will be able to forward information about e.g. orders without having to engage consultants. This, of course, allows to safely send information about order statuses. But you can also accept user complaints and even send them directly to the client’s system.
More about chatbot integrations coming soon 🙂
Chatbots that aren’t integrated with the client’s IT system are great at performing repetitive and general tasks – the most tedious and time-consuming type of work for many people. They also help to gather information for consultants to review specific user cases. In the classic model, the user:
Chatbots can speed this process up significantly. Whenever a user clicks on a specific button or enters a phrase that suggests, for example, that he wants to ask about his order status, we can automatically ask him questions. Questions, which will help the moderator decide whom he should contact. Then, the moderator sends the user’s message to an appropriate consultant who is able to solve the problem.
Despite the enormous amount of work we put into designing user paths, we can’t predict everything. Users can contact a Messenger chatbot that won’t be able to help them. That’s why it should enable them to talk with a consultant. After all, our number one goal is customer satisfaction; to provide clients with a comfortable tool that will allow them to quickly solve their problems.
Technology can be amazing if we design it thinking about its potential users – on both sides. Imagine a customer service assistant seeing as the number of messages is growing at an alarming rate. I see a person who’s under a lot of pressure – after all, on the other side, there are people who are waiting for a quick reply. But you can automate the entire process and allow your team to focus on more important things. With chatbots, you can take customer service to a higher level.