News

City Insider: Human expertise is the key to winning the tech arms race

In the second instalment of his regular feature this month, David Stevenson remarks how little travel has been truly changed by technology – but could that be about to change?

Yesterday I highlighted the danger of companies reaching the Gorging Point in travel, referencing some of the practices of piling surcharging on customers in the cruise sector.


One other observation this month is centred on how exploiting technology is one way in which Gorging Points can be overcome.


Smart new outfits are those that come up with clever ideas that kill the old charging model dead.


The cruise sector in particular has some pretty chunky barriers to entry – ships cost a fortune – but the travel sector generally is highly vulnerable to big changes in tech.


And yet I’m always astonished by how little progress automation has actually made into the practical working of the travel sector.


Sure, the internet has changed everything. But the first wave of technology has fundamentally been transactional based, with the main innovation giving the end customer access to a reservations database.


But in reality this technology is pretty dumb. Every time I use B2C travel platforms I’m astonished by:



  • How little they really know about me;
  • How often I have to keep repeating information;
  • How pointless most of the ‘automated’ suggestions are.

In fact I’d go so far as to say that the average dynamic packager is displaying almost no situational intelligence about me as a customer, with knowledge-based rules almost non-existent.


All this talk of artificial intelligence is, in my humble experience, complete junk. The best travel IT platforms are essentially dumb.


Yet there is a quiet revolution going out there in the real world around machine-based intelligence.


It’s not the high concept world of AI (artificial intelligence), more the ability of machine codes to use algorithms to parse the right information and then shortcut the whole booking process by using tech-based cut-throughs.


As one expert in the industry recently put it; research shows that travellers visit 20 or more sites when planning a trip. The winner will be the one that can do it in two or three steps.


To understand the opportunities that await, consider a simple piece of technology such as the use of Siri or in my case OK Google (on android devices).


I’ve only just started to learn how to use this amazing function on my phone – my teenage kids think I’m positively Victorian – and by god it works.


Obviously its takes some getting used to and I’m sure there’s plenty of training going on in the background but this is machine intelligence at work in your pocket.


Absolutely none of this amazing technology has any meaningful impact on the travel sector. All the talk of replacing travel agents with “personal travel concierges,” able to sift through mountains of information to give you tailored advice on how to plan your trip, is complete pie in the sky at the moment.


All the trendy ideas about cognitive computers learning like a human may as well be on another planet when it comes to the travel sector.


In reality most travel-based techs are still struggling with simpler issues. A recent note by one bunch of machine intelligence experts who won a competition for Expedia explains the mundane challenge.


This group of coders observed how they searched for hotels in Manhattan for four days and were given 363 results.


From this point they were offered filters by rating and price (5 options each), neighbourhood (18 options), accommodation type (5 options), preferences (16 options), and accessibility (7 options).


Each change of filter requires a change in the results list.  But the $64 million question is how to make the most important result change – the hotel at the top of the list. The one listed first is most likely to be clicked on and ultimately purchased.


If that number one slot – or even numbers 1 to 4 slot – aren’t quite right, most customers tend to vanish within seconds.


The challenge for most travel techs is to simply getting a search result right, let alone worrying about things like personalisation. But again I think we may be near another business sector turning point, though thankfully not one involving gorging the poor old customer.


This year and next should see the emergence of a whole new wave of apps and cloud based products hit the markets including HelloGbye.


This platforms aims to use all the data available in the search journey to improve customer outcomes. Another start up to watch is Utrip whose predictive web app spits out fully-fledged travel itineraries.


The process is quick – the average traveller takes less than a minute adjusting Utrip’s 16 “levers,” chief executive Gilad Berenstein told one news service. Utrip generates 10,000 iterations of a relevant itinerary.


Perhaps the most interesting business in the field though is called WayBlazer. This is building a whole suite of tech tricks on top of a robust platform – IBM’s massive Watson machine.


This monster piece of cloud based technology is as close as one can get to the Terminator’s SkyNet AI – a vast hub of artificial intelligence that can be rented out to business partners to solve problems.


The problems tend to be around making sense of the vast amounts of data out there and then tailoring the results for the end user.


WayBlazer’s products are not yet on the market and even when it is ready it won’t be selling its services direct to the customer – its partners are major travel businesses.


But its core product set is fascinating, including services for language processing, framing questions to more easily extract knowledge from large knowledge bases, language translation, user modelling and message tailoring based on those models.


All this cleverness depends on a high volume of user interactions. You can see a first iteration of this technology product via the Austin Convention and Visitors Bureau (ACVB) which has a live WayBlazer Discovery Tool that helps visitors discover the best of Austin. Services include figuring out things to do, places to eat, places to stay.


My own sense is that we are only at the very beginnings of a much bigger revolution in travel tech.


There’s a very good reason why so many people continue to use travel agents and that’s because the sheer amount of online choice is too vast and consumers frankly can’t be bothered to shop around.


But machine based intelligences at work could make a huge impact, narrowing down choice and personalising product options.


And here’s the central insight. This machine based intelligence doesn’t tend to sit comfortably on its own – it’s best used alongside an actual human expert who can make the results come alive.


And so, in this tech arms race the real winners could be the giant travel experts on the high street using all this AI to produce a more customised travel experience. 

Share article

View Comments

Jacobs Media Group is honoured to be the recipient of the 2020 Queen's Award for Enterprise.

The highest official awards for UK businesses since being established by royal warrant in 1965. Read more.