Ai Editorial: Doing away with “search” on a travel ecommerce site - is it the next holy grail?

First Published on 21st December, 2017

Ai Editorial: The “engine room” of search and booking may be tough to change from the outside-in. But this shouldn’t stop the industry from experimenting, looking at areas like recognizing a visitor, moments of inspiration, assessing the style of travel, deals, social context etc., writes Ai’s Ritesh Gupta


The gap between what and where to book to finally completing a transaction for a trip is rather a prolonged one, fragmented over sessions, devices etc. The tussle to overcome this challenge is an interesting one, and as we approach the end of 2017 certain interesting possibilities are cropping up.

One of them is related to doing away with the “search” functionality on, a meta-search site or

Targeting travellers early in the booking funnel and monetizing every user who ends up using an airline’s digital assets is increasingly gaining significance, but the fact remains that evaluating options remains a core issue. Why one really needs to share starting point and the end point of a journey to book an airline ticket? This aspect where one has to identify a destination and then further search for fares or offers around it is being questioned.   

The “search” problem

During the recently held, Ai’s MegaEvent in Palm Springs, California, this issue of making a customer search first and then book was discussed. Imagine, for first time traveller, who is trying to stay in Windsor, UK – how challenging is it for him to evaluate – how far a hotel is going to be and how to arrive from Gatwick, Luton, Heathrow or Stansted? So tabs featuring dates and location (from and to), the functionalities shouldn’t be about making users to do more – for instance, first think of what to do, then search, then choose from a selection of options etc. Planning a trip (be it for finalising a destination or a trip or selecting a flight or a room etc.) should be less tedious. 

“When system works a certain way, we try and make users work the same way. That’s natural way of thinking (sticking to the same pattern of search). But that doesn’t mean something novel or nifty can’t be built around it (a prime example of the latter is what meta-search engines have done around search). There is room for plenty of innovation that can be done if one of the variables is changed around how search is being it happens in the background probably wouldn’t change for next 10 years. But how the user interacts/ uses can be changed, the moments of inspiration, what the moments of transaction…can be altered,” pointed out Kevin O’Shaughnessy, CEO and co-founder,


Delving deeper, O’Shaughnessy recommends: “Think of this as if you’re designing another part of your service landscape. The first thing we do when booking as ask “from where”. It’s not the most engaging and, if your digital and marketing teams are going about this the right way, they’ll have guests visiting again and again, only to be greeted by the same question every time. Recognize a repeat visitor, whether direct, SEO or dumped mid-flow from meta-search and start to engage. This would apply from high frequency travellers, leisure or once-in-a-lifetime trips.” He added, “Passengers may not know their destination airport for example: in that case the traffic will start on search engines, or be referred by a someone else. Their destination may be flexible, but time inflexible: where can they go this weekend for their budget for a city break? Use progressive sign-up account creation for one-off trips: this means that you can build a profile for the guest, adding a little information over time to their account. This is particularly simple on web, almost expected by users on mobile.” 

“In a nutshell, start building loyalty and customer service before the booking even starts, not after. It will help differentiate from the get-go,” he said.   

There have been specialists that are focusing on recommendations (could be based on content, for instance, matches with a traveller loves to do or use their knowledge of a particular city and assess what’s similar in a new destination they would probably visit). It is about taking out uncertainty as soon as possible, and quickly build affiliation with the user by offering them what they could possibly like!

For instance, LikeWhere, an Irish company focuses on a specific niche of travel - personalising destination discovery. Their software analyses cities and breaks them down into neighbourhoods, and then enables airlines to match locations to the lifestyle preferences of their customers. The company validates all their data with a global network of local bloggers.  “We have found to deliver a nuanced recommendation, it is essential to humanise our data, but that certainly isn’t the case across all the verticals of travel. Once we establish certain parameters with a customer we use machine learning to add value, through informing more contextual recommendations. Our product enables airlines to begin their customer lifecycle earlier in the inspiration phase which positions them for the booking/ancillaries – that’s where the monetization is,” explained Matt Walker, Chief Storyteller, LikeWhere.    

Moving towards change

So how to bring in such a radical change? “Take loads of small steps and small experiments in a controlled way. Create a sandbox for quick experimentation, “fail fast, fail cheap”, work on feedback and evolve from there,” recommends O’Shaughnessy.

Do travellers really think of dates, destinations as they dream and plan or even book? To what extent the current way of searching is aligned with the functionalities offered? “Not necessarily (dates and location being always important). One of the traps one can fall into from the design perspective is to assume that everybody travels the same way. Think about the car sharing revolution in southern Europe – from France to Spain to Italy – according to Blablacar, this has created a net increase in mobility, rather than impacting flights and rail bookings. More travel to more destinations. Ask a travel-hungry centennial if they would prefer to travel less or more for the same budget. If one saves money, they can travel more. At the moment of transaction all flights are equal, but how you get to that transaction is as varied as people’s culture and personality,” said O’Shaughnessy.   

“Sometimes it seems that everyone travels differently and most believe they’re doing it right,” said O’Shaughnessy. “We'd like to think there was "one right way" we could use to build the perfect interactive system (or design our service landscape), but in reality, we need to keep touch-points broad, numerous and accessible for a range of different consumer behaviours.” 

“The more service you can leverage pre-booking, the better engagement you’ll have throughout the purchase occasion. As most digital teams may not be in direct control over the “deeper” parts of the PSS’s booking flow, and it may be resistant to innovation or experimentation, pre-booking engagement is an important factor in increasing conversion. The more ways you engage with search itself “weekend search”, “saved trips”, “travel preferences for pros” and more, increases the width of your funnel to allow for a higher quality of user engagement. This can happen earlier on in the booking process,” mentioned O’Shaughnessy.

“The "engine room" of search and booking may be tough to change from the outside-in. That doesn't mean that we can't experiment today with new layers which broaden the funnel and increase loyalty,” concluded O’Shaughnessy.

Follow Ai on Twitter: @Ai_Connects_Us