Article

Can chatbots give travellers personalised customer service like your staff would?

Almost, yet they work 24-7 and don't complain
Letitia Stevenson
May 8, 2019

Can chatbots give travellers personalised customer service like your staff would?

The short answer is - yes they can. However, be aware of the many chatbots that can’t.  Chatbots come in different shapes and forms, including some that advertise you can build yourself.  However, it’s the sophisticated natural language chatbots driven by clever code which can provide amazing personalised service, driving direct bookings and customer engagement for accommodation and activity providers all while saving staff time.

What does personalisation mean?

“the action of designing or producing something to meet someone's individual requirements.”

Typically when we get great customer service it’s because we’ve felt as though the agent understands what we’re talking about and we don’t have to spell out our request.  They just seamlessly understand you.  You can probably remember experiences when the opposite occurred - maybe ringing a help desk and finding yourself trying to teach the person on the other end of the line, that person might be new or just doesn’t know what they’re talking about.

In conversation, the basic premise of good customer service is a customer needs to be able to ask a question and get a response which matches their question and the context given in their question.  Even better, to receive relevant information you didn’t know to ask but which is of personal interest to you.


What it takes to deliver great personalisation with tourism chatbots

Let’s first look at what it takes for a human to provide quality personalised responses.  There are a few basic premises:

  • Responses match the question.  That means vague questions need follow up questions to provide an accurate response.  Detailed questions need the correct detailed response immediately.
  • Provide more than an answer, offer related information based on what you are understanding your customer thinks/likes.
  • Remember context given in the first question, such as a date, number of people and age of children.


When we’re describing personalising chatbots then the best tourism chatbots follow the same principles.  Here are a few technical features which allow this:

  • Natural Language Artificially Intelligent chatbots.  This means it understands what you ask, you don’t have to answer a list of questions to get what you need
  • It needs to have a wide knowledge base - many questions built, a lot of data to train it.
  • Smart logic to give contextual responses.
  • It remembers context and can respond appropriately based on that context (like age, number of people, etc.)

In the same way it takes time to train and educate staff to provide great customer service, tourism chatbots also take time to develop to a good level.  

Make sure that you are clear on the goals you are trying to achieve with a chatbot and that you are prepared to go through a training period, depending on what data you are starting with.


Examples of great personalised messaging

Welcome/Greetings

Chatbots can give your potential customers a warm welcome based on whether they are a new customer or returning for another visit.

This welcome can be timed to pop up at specific times or only when people look for chat, this is entirely up to you.

Examples:

“Welcome to Yonder, let us know if we can help you with anything”

“Welcome back Steve, are you still interested in the Sea view suite? There’s only a few left for your dates”


Get tasks done

There’s nothing more annoying than being put on hold to get someone else in a different department to help you.  The same can happen with chatbots if they’re not built with the full range of actions required to complete a task.  Tourism businesses have many of these tasks, such as taking a newbooking, making booking changes and cancellations.

To enable these tasks the correct conversational flows need to be built and integrated with databases and operational systems such as channel managers, PMS’s, Booking systems, CRM’s.  It’s not trivial to build these integrations, however built correctly these flows will result in customers having an immediate seamless customer experience all through chat.


Web page context

We frequently see people asking questions without giving the full context, for example “How much is this?”. In person, we do this all the time - we glean clues from the situation to decide how to respond, such as looking at what they are pointing at. Providing this sort of sophisticated response in a chatbot is smart stuff, there’s not many tools which can do this. The tourism specific chat tool Yonder is able to deduce what end users are talking about by which webpage they’re on when they ask the question.    This leads to a much better customer experience.


Understanding conversational context

Basic chatbots will give a generic answer to questions.  Let’s take an example like “How much are your seaview  twin rooms?“ A basic answer might be . “Our prices range from $X - $Y depending on the type of room.”. This isn’t a very satisfactory answer to a specific question.  However, it works well if your business only has 1-2 products..

Smart tourism chatbots will recognise the product or the type of product  and give a relevant and accurate answer . Using the same example as above, then an accurate answer might be eg “our twin rooms are currently $X per night.  

Smarter tourism chatbots will continue the conversation and ask further questions to help provide a better answer, and lead them to discover a product and book. For example, if a traveler asked about price, the chatbot will ask about which nights to stay and lead the customer to finding availability.  Then with some crafty wording describe why they need to book now and how easy it is to book online, all like a great sales person.  “And can I tempt you to an upgrade?”  Yes, you can build in criteria to help upsell.  

This feature is important as it replicates human conversations.  Building this into a tourism chatbot requires detailed planning and knowledge of customers to help build engaging and intuitive conversations.


Names

Chatbots can ask users for their name and then use this during future interaction. This personalises the chatbots' response helping to develop a relationship.

Note that some users will not like this intrusion and barrier to discussion and would rather not enter their name. An attraction of chat is that people can stay anonymous if they choose. You need to weigh up the value of this versus the (micro-) barrier it creates for customers to get the answer to their question.    


Seamless handover to customer service staff

Chatbots which give the best experience are the ones that help customers to the best of their ability and handover seamlessly to the customer support team when required.  A chatbot can not be designed to resolve 100% of enquiries, it needs to gracefully fetch customer service to help or gather customer details for customer service team to respond later.

Customers get frustrated with chatbots when they can’t get the answer they need and conversations go around in circles with the frustrating words “I’m sorry I don’t understand, can you rephrase that?”.  Moreover, it is important that chatbots are not designed to block a customer from getting help from customer service team, rather it’s designed as an assistant that works 24-7 and answers a large number of questions and holds many conversations on its own.


Conclusion

Tourism chatbots can give high quality and personalised conversation.   If you are looking into tourism chatbots then ensure you take time to understand the differences in chatbot offerings and the customer experience they will provide.

Yonder has developed a large tourism data set and integrations with tourism booking systems which allows it to build smart chatbots fast and cost effectively. It also have extensive knowledge of tourism businesses and are industry leaders designing conversational flows that lead to people booking direct . Yonder aims to build relationships with businesses to really understand how to design and embed effectively and help navigate this new communication channel.

Time is needed to achieve the best results. Yonder supports businesses through this learning curve and continues to train your chatbot until it becomes "an indispensable staff member" !  These our the words of our customers, check out our case studies to learn why.

Recent posts