The notion of using an assistant to plan travel itineraries, hotels, flights, car hire, even currency exchange, has been around for decades. It’s such a popular solution that a few of the big travel brand outlets still exist on the high street. However, I recently discovered at the London WTM conference that new advances in AI could soon make booking travel with the aide of an online assistant a whole lot easier.
Chatbots have been around for a while. Originally designed as a cost-saving customer service and sales solution they are a good technical solution. However, simulating life-like person to person interaction has proven far harder to replicate than predicted.
When we chat we don’t want to talk to a machine, we want to have the kind of verbal interplay we usually receive from human interaction. Language is key to this fruitful relationship between the client and the seller. But realising this essential feature of the sales process is a mammoth task. Conversational interfaces are in development but are not yet sophisticated enough to satisfy our complex demands.
Pop up windows
Privacy laws have made a significant impact recently to pop up search solutions. Customer consent is now required to keep a record of any previous communication and personal data. Once given, this permission can be used to provide the answers to many travel questions via chat or text. However, the information provided is limited to the issues already covered in FAQ’s. An adequate solution for taking on some of the burdens of customer service that would usually have taken place over the phone, but still not quite the full service most of us dream of.
Alaska Air’s Jenn and Amtrak’s Julie are both examples of chat systems helping travellers with any information they can retrieve from their vast databases. Expedia is currently using a combination of chat services and Alexa Skill to provide personalised travel solutions. Both Alexa and google assistant are already capable of booking flights, hotels and car hire, flight status etc when instructed.
As the web floods with more information, the rise in online personal travel counsellors is evidence that the demand for a quick but comprehensive travel planning solution is needed. Natural language processing will undoubtedly be at the heart of any future travel assistant product. As an interpretation of intent, as well as the sophistication of understanding a customer’s needs and wants, are all required. It’s a sensitive game to play which as humans we are perfect at judging. Only when NLP is seamlessly integrated will the differences between a bot and a live customer experience be none existent. In the meantime, a fully computerised assistant that can properly discuss your travel demands is still a bit of a dream.