As companies are exploring the use chatbots for customer service, one of the first candidates of "Content" that come to mind to automate is the current FAQs represented on corresponding sections on a website. It is tantalising to think that you just take these and change them into the form of a conversational UI since they seem to be in the form of answers to questions after all, right?
Tools that promise to just incorporate an FAQ section of a website and turn that into a conversational chatbot are out there and sound promising, but often fail in real practice.
So why doesn't an FAQ translate 1:1 into a chatbot?
It is in the nature of dialogues that people ask follow-up questions.
What is Apple Care? How can I contact the Apple Customer Service?What is the price per minute? It is thus crucial to maintaining context to be able to answer questions such as "How much is it?" or "What is it?".
Sometimes even the same question can yield different answers, depends on the flow of the conversation. Without it, users will read "I'm sorry, not sure what you just asked" a little too often a mistake frequently made with the early chatbot implementations we saw in 2016.
For not-so-frequently asked questions you're left alone or rather left to figure out how to get human help.A chatbot that just covers those items that are among the FAQs and doesn't provide an easy path to human help for those that aren't can easily get frustrating.
Frequently customers come with issues such as "Where is my order", or "I need to change my upcoming appointment".
FAQs do not answer these; they rather point to a place on the website where you can log in and get an answer to this question. If a chatbot does the same, rather than tell you where you order is or asking what day you would rather come in for your appointment, you produce friction and a break of a medium, which doesn't help the experience. Websites by nature are media-rich environments that allow for a high-fidelity display of information.
As part of the development of FAQ content into a chat interface, analyse your message and conversation UI. Formulations such as "Users can register here:" should be changed to second person singular:"You can register here:"; longer messages - and they cannot be avoided given the subject matter - must be split up into smaller pieces to fit into the constraints of Facebook Messenger. Lastly, it isn't sufficient to teach a chatbot how to understand the question if asked correctly as it is described in FAQs.
Questions in FAQ sections are intended as broad questions with all-encompassing answers, and they are worded in written language style. Where an FAQ might formulate "What are the conditions for the application of the adapter", a real person might ask "What do I need so I can use the adapter", or simply "How can I use it?".
FAQ sections on websites are meant as an additional source of help. A chatbot, say one that lives on Messenger or SMS, greets the user with a "Blinking cursor”. There is no way for you to guide the conversation, you have to respond to what they're sending, which might be a simple "Public domain" question, a CRM-type account question, or a complaint.
Pulling in FAQ content into your customer service chatbot per se is a good idea - the more information/data the chatbot has at its disposal, the better the conversational intelligence.
By bundling FAQs with the questions you are handling on other channels, say for example your IVR, combining it with your enterprise backend systems, and implementing the commands of the CUI, you should be able to realise the rewards quickly.