With Hipmunk and Room 77 the current darlings of the Valley crowd, it is perhaps time to revisit the user experience of travel sites.
So-called GDS "green screens" so beloved by travel agents everywhere, and completely dismissed by just about everyone else, have a great degree of control over the interaction between data sources and the user.
And this is a requirement because the GDS systems designed in the 1950s were born in times when getting any form of functionality in front of a user was a tough thing.
In those days, programmers ruled and users cowered. Functionality was defined by capability rather than by ease of use.
Mainframe-based systems were the order of the day, where scarce computing horsepower and disk space was the constraining factors. Fast forward to today and the world is different.
We have every form of communications interface. Human, virtual human, mobile, character-based, pixel-based etc etc. But what has this done to the core systems?
At the back of everything in air and hotel technologies are some remarkably robust and functional systems. These are rock solid – work 7 x 24 (well, okay, not always in hotels!) and rarely go down.
If and when they do it is usually in a dramatic fashion and in a manner that is immediately seized on by all concerned as a portend of the end of the world.
Yet at the same time all of us experience poor reliable applications on our PCs (Windows or Macs) along with, frankly, appallingly badly written apps on our smartphones that do not work well with the terribly unreliable mobile data connections.
We are perfectly happy to trade off poor quality mobile services for the value of being mobile.
So enter the new user experience. As several of my fellow Tnoozers have written (particularly Martin Collings and Tim Hughes), travel ideation and the interface are key elements of the process of travel.
The user experience for the past 15 years in online travel has changed very little. So there is a real need to revisit the subject area.
In the 1980s and 1990s we used to get excited about GUIs. Even before that the watch word was the acronym WYSIWYG – literally standing for What You See Is What You Get.
I have often spoken about the need for reform with the user interface - about the poor user experience and the need for a new model. In fact, the process of an interface should be as unobtrusive as possible.
Indeed, we should rid the user from the need of an explicit interface, unless they want it.
Hipmunk and Room 77 do a nice job of creating ooohs and aaahs in front of the digerati - however to an old clunker like me, the interface I can see in both of them is not going to improve my personal experience.
I freely admit here that I am not your typical user. I travel more than 250,000 miles a year and am constantly looking at all forms of travel products for both personal and professional reasons.
Consequently I give all the tools a good workout. Neither of these tools do it for me, yet. And they won't for a while.
The amount of work necessary for Hipmunk to work across a wide range of flights is probably okay for a once in a while occasional flyer/stayer and daydreamer.
For a frequent traveller and a business traveller it lacks a significant level of functionality - not least of which is the element of control.
Room 77 is a product that probably belongs right up there with Second Life. I don’t buy the concept (again yet) of augmented reality and this frankly doesn’t do it for me.
For a very simple set of reasons. The ability to do anything with the information is very limited. Am I going to insist on room 4004 over room 4002? If so, how will I communicate that message to the hotel?
Just imagine the dialogue with the reservation desk when booking the hotel by the phone. Do I hear a lot of sniggering in the background?
Some might draw the conclusion that Room 77's mission in life is to keep hotel call centres in business!
If I look at the way it works – there are few controls. Counting the ones on the query and the results pages of the new Hipmunk Hotel app, you also have very few choices:
- Input – date, place
- Modification – price, amenities and its info overlays
Hardly earth shattering.
The problem is about the level of nuance. This is the progression of objective definitive statement - eg. the lowest possible price to the next level lowest logical price and then, of course, the highly subjective and unique personal service which is the lowest desirable price.
ITA Software, for example, claims there are hundreds - even thousands - of possible controls, but abstracting that to work is very hard. That statement is based on the assumption that the data elements are discrete and defined and raw.
Hmmm, not the case.
What is missing is an appreciation that the infrastructure powering the current generation of systems has been pushed to its limits.
Indeed I would probably say we may have pushed it too far. We are now suffering from such ailments as poor caching, higher results failures and booking failures. Before anything major can occur we need to look hard at the back-end no matter how nice the front end looks.
These products, such as even the current generation of high-end UEX-led sites, cannot work to their potential without better infrastructure. I liken this to trying to observe a scene using a telescope backwards.
While I may be harsh in my assessment of both these products, I am not completely turned off. They are transitional products on the way to an implicit application, as opposed to an explicit interface.
Thus, if we are going to proceed with fixing some of the infrastructure let’s do it in a smart way whereby we look at the whole end to end workflow.
If we have to go through this step then so be it. I would much prefer to go down the path that products like Portaga attempted to blaze but sadly didn’t get there.
In other words: we need to do a better job of aligning the content acquisition with infrastructure, processing and final the elements of search-based commerce.
We should be thinking about the travel interface and, in particular, the booking process as much as we should be thinking about character spacing on a word processer. Yup not at all.
NB: Here is a presentation to go alongside the article.