Michael Lovegrove
July 9, 2017

The Chatbot review — # 1: Skyscanner

Welcome to the Chatbot review! Every week, we will be reviewing a leading chat bot, so that you can learn what they’re doing, why they’re doing it and how effective it is.

We will be reviewing the following:

  1. Chatbot objective
  2. Specific features
  3. Natural language processing (answering questions)
  4. Conversational design
  5. Overall effectiveness

With that being said, let’s begin!

In the first episode of the Chatbot review series, we will be reviewing The Skyscanner Facebook Messenger chatbot.

For those who are unaware about what Skyscanner is — it’s an online platform that finds the cheapest flights. Essentially, the Trivago for air travel.

To view the chatbot in action, click here.

Chatbot objective

The point of Skyscanner is to find a cheap flight to a particular destination, however this assumes that a given customer already knows where they want to go. What about when a customer actually has no idea where they would like to go? Other than actually wanting to head abroad?

This is where their chatbot comes in…

Even when a customer has no idea where they would like to go, the Skyscanner chatbot will give said customer a recommendation based on their location:

Even though providing the same functionality on a web application is possible, it’s also significantly less likely to be used by consumers, because it’s harder to find. If you want to read more on consumer digital tendencies, read about it here.

Specific features

Skyscanner have managed to pack in a number of other features into their chatbot. If designed incorrectly, then some of these features can be hard to find. However, the Skyscanner chatbot does a great job at showcasing further functionality after its primary objective.

  • Find flight deals: Tell the bot where you would like to go and it will find the cheapest flight.
  • Explore destinations: Skycanner gives the user a taste of a few popular destinations to provoke interest, before asking to perform a search for a flight (the feature above)
  • Travel tips: How and when to travel
  • Flight price alerts: Alerts when pricing change

Natural language processing

The Skyscanner bot can answer a selection of topics, and variations around a given topic. As a user, simply tell the Bot where you would like to fly to, and the bot will do the rest!

However, a bot typically needs to ask a few questions before providing an accurate answer. If you interrupt such questions, then expect the bot to be thrown out of whack…

As you can see above, Skyscanner understood my basic request, until I threw in a request in the middle of it’s function. Not exactly a problem with the bot, but with it’s user (me!)

Conversational design

Conversational design is similar to UX design. In both instances, we aim to provide a simplistic solution that is easy to use.

With that being said, a chatbot must get the the point, quick. At the same time, the ‘point’ must be relevant to the user/customer.

Skyscanners bot actually does a really good job at this. From the first message, users/customers understand what its ability is. Within a handful of interactions, a user can have a flight offer on their screen.

PS — this is the first time I have heard about Montego Bay… Looks like a nice place though!

Overall effectiveness

Skyscanner is probably one of the most effective bots currently available. It provides utility in a more simple manner than a website, meaning it can be used by more people.

If you are looking to develop your own bot for your business, then applying some of the methodologies that are discussed above will help you to create a successful bot.

About the Author

Michael Lovegrove

Michael has a background in innovative technology, public speaking and technical consulting. Now as the Co-Founder and CEO of botbuild.ai, he is a leader in human-conversation driven design.

Related articles