The first wave of chatbots gave us a glimpse of what the technology can do, but there was one thing consistently holding them back – conversational UI design. The travel industry has seen a lot of chatbots come and go over the last year with a number of tough lessons learnt along the way.

As machine learning advances and conversational UI design matures, the next wave of chatbots is fast approaching. And there’s much less room for mistakes this time around – so here are five conversational UI design lessons from the best travel bots we’ve seen so far.

#1: Reduce typing

A certain amount of typing is necessary for most conversations and that’s fine. However, reducing the amount of typing users need to do is one of your biggest concerns in conversational UI design.

In the video above, you’ll see a demo of a flight booking bot from MindMeld – an AI powered, machine learning chatbot platform.

In the demo, you’ll notice a number of UI elements reduce the need for typing, which keeps the conversation running smoothly. First of all, you’ll want to go with the most advanced voice recognition API you possibly can. Accuracy is important here. Next, you should use a mix of touch buttons and carousels when users need to choose between multiple predefined options.

Typing is always there as an option, of course. But users should rarely need to type anything if your conversation and UI elements are properly designed.

#2: Make date selection a breeze

Something else that works really well in the MindMeld video is date selection – and this is important. The thing is, if you’re not building your own bot from scratch, you’ll be confined to the options available on your platform of choice.

Messenger bots are a little more restricted than MindMeld but Expedia works within these confines:

Expedia travel chatbot - conversational ui example

By providing “tonight” and “tomorrow” options, users have the option to quickly select these choices. Users can also select “Another date” to bring up the default calendar UI or type in a date if they prefer – as the example above shows. Notice that Expedia’s bot gives examples of how to format their date choices and also sticks to using them in its own date options, which creates continuity.

#3: Make full use of device features

Yes, we’re talking about voice UIs here but there are plenty of other device features you can use to enhance the experience. We’ve already mentioned touch screens as one of these features and there are plenty of others; camera, location, contacts list, etc.

Make use of them.

For example, when the Skyscanner bot asks people where they want to fly from, it provides a one-tap option to provide their device location. This saves users typing or even speaking when their nearest airport is the preferred place to fly from. If it isn’t, nothing is lost.

Skyscanner chat bot

#4: Localise your information

Musement promises to help you find the best events and attractions in any given area – a handy tool if you’re visiting a new city for the first time. The conversation with its Messenger bot is nice and intuitive too. Once again, you can click to provide your location and choose between categories like “London for the kids” and “London on a budget”.

Musement conversational UL example

It’s just a shame that prices are displayed in USD by default and there’s no attempt to localise currencies when you’re specifically searching for things to do in a country that uses GBP.

#5: Think beyond the booking

Earlier this year, we looked at some of the best travel bots that were available at the time. One of the bots we featured was KLM’s Messenger bot which aims to make life easier for travellers after they’ve booked their tickets.

It’s a great example of using the technology as a customer service platform, which helps boost your brand image.

Here, KLM is making full use of smartphone features by allowing users to get their on-screen boarding pass, which can be scanned at the airport. Users also get updates on their flight status should there be any delays and they can also reach out for individual customer services needs (e.g. seat selection, meal preferences, etc).

Conversational UI design is maturing at a time when machine learning technology is rapidly advancing. The chatbots we were talking about at the beginning of this year are already looking outdated and the next generation of conversational UI highlights how much progress is being made.

Free travel seminar

If you’d like to learn more about designing conversational UIs and how to choose the right chatbot platforms, sign up for our free seminar and workshop: The future of travel search marketing

Save

Save