November 2010


Tonight I’ll be attending a function hosted by Edelman at their downtown Chicago office. As the invitation says, This is going to be a series of a few fast paced presentations followed by a group discussion on the latest trends in Social Media, focusing on Social Media Strategy. I’ll try to do a post event write up here in the next few days.

Presenters are

  • John Peters, VP/GM of Tripology/Rand McNally & Moderator of the Meetup
  • David Armano, SVP – Digital, Edelman Chicago
  • Abby Ross, VP Marketing – Always Be Social
  • Krista Parry, Director of Marketing & Communications – Park City Mountain Resort

If your job touches on social and you don’t know of Dave Armano’s blog then be sure to take a look, but the others on the roster for tonight are less well known to me.  Some of the topics they are promising to cover  are

  1. Influencing bloggers to tell your story
  2. Setting the right goals, resources and budget to get started properly and effectively
  3. Social Media and ROI
  4. How to align goals with metrics

And speaking of upcoming engagements, I’ve just accepted an invitation to speak at the Eye for Travel Customer Centric Strategies for Travel conference being held in Atlanta in late January. That one sounds like an interesting event, so take a look at the agenda and see if it is relevant for your job. I’ve come up with a good title for my presentation, so now I’ve just got to make sure the content is as interesting as the title! More on that in a few days.

Check out this incredible story I was reading over the Thanksgiving break. It comes from an Indonesian news source and is about the Indonesian airline Garuda.

A switch-over and data migration of the airlines’ Operational Control System (IOCS) that schedules aircraft utilization, crew assignments and passenger departures crashed on Sunday, November 21, 2010, leaving passengers waiting at airports across the nation staring at parked aircraft lacking the required flight and cabin crew. In the course of the data migration important data was lost, blamed by the airline’s information technology staff on a loose computer cable.

A loose computer cable sounds like a very difficult to believe excuse to me, but stranger things have happenend.

I’m not sure of the IT provider in the case above, but it was less than a week ago when Spirit had problems with their res system and the website could not take bookings. At the bottom of that Tnooz article there was an interesting comment from Colt Cooper.

I customize Navitaire’s New Skies for a number of airlines and it’s disturbing that rez system outages have hit some of Navitaire’s biggest clients like Ryanair, Virgin Blue and Spirit all in the span of a few months. Accenture-owned Navitaire has been running damage control since the Virgin Blue incident and I fear this can’t help their cause.

I was in a meeting with an airline customer recently where I emphasised the importance of execution and a proven track record with successful migrations when choosing a vendor, but sometimes it takes a major problem like that at Garuda to really bring home the importance of this point.

After four consecutive posts on AA, I promise this is the last for a while. And be warned, the screen shots below are around two months old. I had good reason for not publishing them for the first month, but no reason for the second month of delay! The good reason will become apparent soon, but where there is no excuse there can be no corresponding explanation!

First, take a look at this clever (some would use a very different word) way to induce upsell during the booking flow. Anyone in a hurry will immediately click the default button without realizing they have just added more to the cost of their ticket.

If you ever wondered why the US carriers have an advantage over their European rivals in the race to see who can get the most ancillary revenue, you just need to look at the relatively minor stuff that isn’t allowed in Europe and then compare it to the example above.

Now to the screen below – interesting use of the click for assistance panel on the right. Some people are predicting this will be a big trend on more airline websites in future, but I haven’t made up my mind yet. What I was really impressed with on this page was the way they try and get you to apply for their co-branded credit card. You may need to expand this screen shot to see it properly, and I’d love to see stats on card take up versus drop-out rate from this page, but without having the stats on hand (and that may change my opinion), I do like it.

But now let’s move to the strangest part of the flow, and the reason why I did not post this immediately. Look below and read the fine print. I was using a Spanish bank issued Visa card with my residential address on their records listing me as being in Madrid. So according to the fine print below I cannot buy on AA.com, at least if I use this credit card. So when I inserted the payment details and it asked for billing address, true to the fine print below, Spain was not one of the countries listed in the drop down box.

But I really wanted to put the tickets on this credit card, so I just used a fake UK address. I was expecting a rejection when the authorization was done as I know that address verification works in USA and UK (it is irrelevant in many other countries where Visa card are issued). Surprise, surprise, the auth went through and I could soon thereafter see via my online bank statement that the card in Spain had indeed been charged by American Airlines. The tickets were subsequently issued and I received my itinerary and confirmation via email.

Now you can see why I waited to recount this story. Even though it looked like everything had gone smoothly, it wasn’t until I had completed both legs of the trip and was sure that I was not going to end up with egg on my face by having boarding denied that I wanted to go public on this. In the end, the flights were taken without a hitch.

Why American Airlines would put such artificial limitations on the growth of their direct channel is beyond me. I know Nigerian issued Visa cards are not accepted by many merchants, but even a number of low fraud countries are excluded by AA. But then even more bizarre is that when you put up rules on acceptance they should at least be enforced; otherwise you are just turning away legitimate business in favor of the fraudulent or the creative – I’m putting myself in the latter of those two categories!

Sometimes airlines think they are doing something really innovative, but the most impressive part of all is overlooked for something more mundane but which is touted as the next big thing. I suspect AA’s new Search By Destination Ideas may fall into this category.

What I am about to write relies heavily on earlier thinking I have done on the future of travel search (see here and here) and also the recent discussions I have been having with various people in the industry about the major trend approaching that will see the merging of pre-shopping and shopping; and the significant implications this will have for airline websites in future. I don’t have enough time to go into that topic in detail here, but definitely ask me about if we meet. For now, let’s look at the concept of what is a “family destination.” In the past I was mildly critical of airlines using Family as an attribute in destination inspiration and American has also fallen into this trap. Despite all the cities listed in the screen on the right, I could not find any reason or justification for why these destinations were considered family friendly. But the bigger point is who cares. Why is anyone persisting with the notion that arbitrary classification using a one size fits all model is the way forward with travel inspiration? Time to repeat something I wrote months ago

To put it bluntly, today’s attributes have a long long way to go before we are at the end game. For example, today I can search by attributes listing museums as my preference. But are there really more than a handful of airports that don’t have at least one museum in the major city or town that they service?  What would be much more compelling would be to put in my maximum price, and have each destination returned with a flight, a hotel (that I could vary by star level using a slider) and one entry ticket to a temporary exhibition at a local museum.  As Stephen Joyce from Rezgo wrote about his hopes on sponsoring the PhoCusRight report:

We will also see large OTAs and distributors looking for ways to connect with suppliers in a big way. This will put pressure on application developers to build systems that easily connect to these distribution channels

So destination content is one part of the puzzle. Another part is more attributes that aren’t dependent upon human whim. I saw one attribute that could be selected was children. What makes one destination more child friendly than another without any knowledge of the ages of the children or what they like is utter guess work. Attributes like this lack credibility and take up valuable screen real estate for no apparent gain. Much better attributes would be things that could interact with a slider, and that also could be populated with data sources that were automatically updated rather than manually maintained. Average temperate at that time of the year, chance of rain, days of sunshine, water temperature – or maybe something really edgy like safety (aka muggings/murders per head of population) or even average age of other people that flew to this destination (difficult to be accurate in domestic). They are all off-the-top-of-the-head ideas without even giving it too much thought, but none are dependent upon human classification. Adding more attributes that follow these rules are potential game changers, and for airlines they have the added benefit of moving their presence to the left of the Bow Tie and into travel inspiration in a big way. But airlines will still be at a disadvantage relative to OTAs unless they can find a basis of competition that puts the ball back in their court – nothing I have written above is sufficiently unique to do that.

So which part of the new AA.com is the most interesting, and the part I alluded to in my opening paragraph. Take a look at the “family” results page below:

Forget the confusing order of fare families and price increments, or the color of the instant upgrade fare family being green when the cabin in the Tampa flight shows it as being First Class just like the fare family to the right, and forget that is says the default sort is by number of stops but the price order for different flights does not seem to adhere to this logic when you look at prices from top to bottom. Forget all that. The interesting part about this display is that it is taking us a step closer toward the merging of pre-shopping and shopping and any airline that doesn’t understand this needs to make sure they come to grips with it very quickly.

After starting with this idea of taking a deeper look at AA.com, I now think I’ve got a few topics to write on this week. Mostly just general observations regarding parts of the website.

When I was looking at the site the other day I thought it would be interesting to sign up to their email offers list. I clicked on the “email exclusives”  link that appears on the bottom of most pages, and it takes me to the page you can see in the top three quarters of this screenshot. Whilst not in the same league as Tiger and their impressive Stripes offer, you need to remember that American are a completely different model to Tiger so the same rules so not apply. In my personal case, I already have status enabling lounge access with American’s oneworld partner Iberia, so I don’t see any need to join AAdvantage. From what I saw, it looks like I am not allowed to opt in to the AA email distribution list without first joining their own frequent flier program. Not quite sure of the logic of that, but maybe if the offers really are exclusives and not just the same offers available to all then maybe it could make sense. If not, then it is just turning away potential customers, especially those already loyal to the alliance.

Hardly the most serious issue in the world (unlike the end of my last AA post) but sometimes it just helps to have an outsider (or semi-outsider in my case) point out things that may or may not have been realized already by an airline.

I’ve got a post on the AA.com booking flow that when I put it up later this week, I’m afraid it might already be outdated. The reason is that on November 15th American Airlines put out a press release announcing they had redesigned their site. Some of the improvements announced by American were written up by them in the following manner:

At the forefront of the redesign is the site’s homepage, which now provides simpler navigation. Access to all the tools and information customers need is just a click away, thanks to a navigation bar that has been relocated to the top of the screen, rather than the previous left side of the page.

Other features of the redesign include:

  • Upgrading the page width from 800 to 1,024 pixels, which enhances readability
  • A new header, footer and background image on all pages of the site
  • Added links to American’s social media sites in the footer of all pages
  • Centering the site in a customer’s browser

This will not stop me posting in a couple of days a booking flow review with screen shots from about 2 months ago, as it is hopefully still an interesting story. But back to the point of today’s post. Below is a comment from one customer of AA.com on Flyertalk (going by the name of Scion) soon after they put the new redesign into production.

I just had a disturbing phone call.

Another person who works in my same company called to report that while attempting to log in to the new AA.com, and after entering his username and password, he was presented with details of MY account … and evidently was able to navigate around, view details, and might, had he been so inclined, have been able to book tickets, use miles for other purposes, or who knows what.

I have never met this person [let’s call him Mr. X], and he works in a different city from me. Insofar as I know, we have never used the same computer for any purpose at any time.

After this experience, Mr. X called AA, who told him repeatedly that the problem must be at his/our end … presumbly with our comany’s Internet cache servers/firewall. They suggested that he determine whether the person whose account he was viewing worked for the same company. That turned out to be the case, he phoned me, I phoned the EP desk, and that’s where we stand.

This problem, if it is widespread, clearly represents a significant risk. Since I see no obvious reason to believe that my company’s infrastructure is unique in this regard, I suspect other users may be similarly vulnerable. Since the emergence of this problem coincides with the rollout of the new AA.com, it is difficult to imagine the two are not connected.

Very disturbing indeed if in fact it is true as told. About 10 years ago a very similar thing happened to me in a business I was managing at the time (pre Amadeus days), and for both me and the developers involved it was an extremely stressful period until we resolved the bug. Basically, the problem was that were were initializing a datastructure in memory on the server at site-startup time, rather than per-connection. The meant it was shared between subsequent connections; after one person finished the next person to connect would sometimes ‘inherit’ the data.

We fixed it by initializing the data for each connection. Our problem was related to the way Perl on Apache worked, but this sort of thing can be easy to do with languages and frameworks that use threading, such as Java and .Net.

I hope AA has managed to resolve the issue (it is extremely difficult to replicate such a problem on demand), but having been in a similar situation myself and knowing how difficult it was to diagnose, I feel for the development guys at AA who were probably pulling their hair out trying to understand what had happened.

Recently in the article titled When social, mobile, location and every other buzzword come together I was praising Vail ski resort for launching what is probably the most targeted social travel-related tool I have seen to date; but at Phocuswright in Arizona earlier this week I was sitting in a presentation by Pete Stein from Razorfish where he showed an example from Sweden that really impressed me – at least insomuch as it comes to targeted geo-gaming and using this as a way to get people engaged with a brand. Watch the short video below.

Seeing that ad I was reminded of writing back in April about geo-gaming being one of the five untapped opportunities for mobile and travel. I’m still waiting to see something as compelling as the MINI example from someone in the online travel space – now if I have seen that at the Travel Innovation Summit I might have even put my hand in my own pocket to back it!

And whilst talking of the Travel Innovation Summit, let’s try and tie that to Expedia in two ways.

Firstly, if you were there you might remember a company called TrustYou doing quite an interesting pitch on semantic search and trying to help their customers understand the intent of a reviewer rather than just focusing on keywords alone. So I was interested to see the following title to a story in Eye For Travel a couple of days later: Expedia to evaluate sentiments expressed in text collections.

The second tie in is that I was supposed to be having dinner with Mobiata CEO Ben Kazez and a group of others on the evening of the Travel Innovation Summit, but on the day before he pulled out saying he was no longer coming to Phocuswright. He told me something important had come up, but didn’t elaborate. Later that evening I was talking to Tim Hughes and we were both wondering what could possibly cause someone to pull out of such an important conference at short notice.

A couple of days later during the conference Expedia CEO Dara Khosrowshahi announced they were buying Mobiata! Congratulations to Ben and the team, as for a company that was only formed a couple of years ago, selling to Expedia is a phenomenal exit. And I also think that qualifies as a pretty justifiable excuse for pulling out of a dinner engagement with me at short notice.

Next Page »