When life gives you lemons, make a lemon slushy in Iceland — a tale of airport operational failures and hope
As a consultant, I am sadly, no stranger to extensive travel, and the inevitable hassles and pains that come with it. Nor, as a resident 24 years in Central Illinois, 10 in Boston, 5 in Detroit, and 6 in Chicago, am I a novice to how to handle bitter winter conditions. However, recently, my wife and I shared such a bad travel experience with 4,000-5,000 other people that it compelled me to not only rage about it while it was happening, but to reflect more deeply upon it as an ops and process improvement professional.
In this article, I will not only take you through horrible travel experience, but also look at it from a critical ops professional point of view. Hopefully, Hiller Associates can not only provide some advice to a foreign airline and airport, but that we all can consider in designing and building organizations to be fault tolerant and resilient to surges and shocks.
My wife and I were returning from Europe to the US several weeks ago, after succumbing to this newish first-world fad, a “babymoon”. It had been a whirlwind of education in history, culture, religion, art etc. in the winter in Paris, Belgium, and the Netherlands – even combined with a business meeting (who says multitasking doesn’t work?). After our last three nights, in Belgium, where an old friend kindly hosted us in his home with his family, we were flying out of Brussels. Always cost conscious, I had our flight going from Brussels (BRU) to Keflavik (KEF, Iceland) back to Chicago (ORD) on Icelandair (IA), which seemed to be saving a fair amount vs. other options.
Enter winter weather, which is to be expected in January, and was a massive shock to the system (in this case Keflavik airport and Iceland Air). As we will see, such shocks are sometimes unavoidable, but what matters is how you have (1) prepared for and (2) react to that surge / shock condition.
Feel free to follow along with what I am about to tell you with a birds-eye process flow view using Figure 1.
Click picture to enlarge!
We started spending 4-5 hours comfortably seated in Brussels airport, as we received furious texts and emails from Iceland Air delaying the flight, repeatedly. However, I do not remember getting any notification that a voucher for food was available, until I talked to a nice Belgium employee of Iceland air… 30 minutes before we had to board. Unfortunately, Iceland Air was being frustratingly obtuse on whether our connection to Chicago would be waiting when we landed for several hours, or not. We were supposed to leave at 1 pm CET from Brussels. It was now 5:00 pm.
Three and a half hours later, we were circling Keflavik, and that is when the problems got much worse. We ended up circling for 15-30 minutes. We had to complement the pilot for getting us down safely, as it may not have been a safe idea to have let us take off in the first place, given the winds and blizzard conditions. Iceland claims it is warmer than NYC in the winter, but the conditions were very bad this night.
We’re there! Wrong. We were about a football field from the airport… but we have to wait for over 2 hours. Eventually Iceland Air (who does NOT serve meals on international flights unless one [over]pays) gives everyone a little water and a sad granola bar. This is a place is called ICEland, but they seem to have no plows to clear a path to the runway, and for some reason, we cannot just creep slowly for a few minutes through the snow. When we finally do taxi to the airport, there is not even a jetway. We have to exit down an icy set of stairs, out into the blizzard, then carry bags upstairs to the airport. We thought to ourselves, “couldn’t we have just walked the football field 2-3 hours ago?” My hopes of making the connection are now dashed, but at least the ordeal is over, right? Wrong.
~10:00 pm UK time – upon entering the airport, my recovering engineer and process geek spidey-senses were going off immediately. The scene had a refugee camp malaise and chaos to it, with people looking utterly confused (including Iceland Air employees). We were eventually funneled into a long line of passengers from multiple flights waiting to speak with only two people from Iceland Air to help us figure out what was going on. In the US, I would have called the airline, but it was clear that strategy would be futile with this national carrier. I send my 7 months pregnant wife out for food immediately, fearing the worst, as I held position. She returned 10-15 minutes later shocked that there is only one market / newstand open for food. She reported the sort of scene I remember from 60 Minutes segments in the 80’s, inside a supermarket in the USSR, as frantic passengers swarming like locusts over rapidly emptying shelves with not nearly enough food and drink. She spent over $40 for two sandwiches, a couple bags of potato chips, a water, and a strange chia seed fruit dessert cup thing. The only protein my wife can find in the middle of the night is cold deli meat on the sandwiches, which she is nervous to eat, being pregnant. We fuel up for what is going to be bad night.
After making friends with people in line, and waiting for almost 2 hours, suddenly everyone in our line and the entire airport is told to get their luggage. But, it’s unclear to everyone what to do? Wait in line? Go to baggage? Given the cycle time of Iceland air to help each person seems to be 10-15 minutes, I decide this was sunk cost, and we headed to baggage. However, the airport had now sent several thousand people through a very small pipe (one escalator) at once. It took half an hour to move several hundred feet. Getting the luggage was not that bad, considering, but then we were directed to ANOTHER line. At least this time there were about 6 Iceland Air desks open. However, there were 2-3 times that amount that COULD have been open. Why are they closed? My wife, who is pretty low-maintenance and much more patient than I am, beseeches the Iceland Air lady who is herding us. She says she’s pregnant and we need something more than the airport floor tonight. The women shrugs, and it is clear that she could not care less. After 2 more hours of miserable standing, they announce there are no more hotels (at least in Keflavik).
While waiting we noticed an army of heavily geared-up search and rescue guys, looking like they wanted to help around the airport, but not being used in the least. The airport also announced that the Iceland Red Cross was setting up a “shelter” with some water and cots, because there are no hotels. It’s now 1 am. Children are asleep on suitcases; the disabled are dozing in wheelchairs. One lady passenger kindly offers us a blanket she stole from the plane. My wife, typically a trouper, is done; she bursts into tears. I am out of patience, too. I’ve been noticing failure after failure in the operations of this place, and I am professionally and personally offended.
Hailing a search and rescue guy, I explain the situation. He is the only person in the airport we have met that is (a) empathic and (b) had an action bias. He says not to worry and he’s on it. After 15-20 more minutes, we do have vouchers to a hotel in Reykjavik (it is 45 minutes by car, but at least my wife is not riding on a donkey while I seek a stable and manger). But then we find out that apparently there was an accident on the road and other plowing problems for which they were unprepared (in a country called ICEland). Busses are delayed; cabs are delayed; it is 2 in the morning. A group of 10 of us have to wait about 30 minutes outside in the blizzard, as one cab arrives every 5 minutes. Finally, we get a very nice driver who speedily spirits us away to the hotel.
The next morning (after a good hotel and breakfast), we feel a little better, but things are not better at the airport. It takes 4x the time it should to check-in, because Icelandair’s one kiosks and systems are confused that we have all been transferred to new fights, so the system forces everyone to the human desks (now 12 are open thankfully). After more lines at security, we go to the gate, which is chaos as well. There is no seating, and people are jammed up in the hall blocking other people who are trying to get through, while other passengers pack themselves on the stairs from our level up to the jetway, creating a fire hazard on the stairs. No one from the airport or Iceland Air takes charge to get the people off the stairs and have them line up in an orderly fashion along the hallway wall (I regret that I did not show the initiative to do so, myself).
But, eventually we do board and, mostly, uneventfully get home to Chicago.
What went wrong and how to improve the operational system
So, that is the story. It could have been a lot worse, i.e. as far as I know, no one was serious hurt in the ordeal, but the suffering of thousands of passengers (and many airport and Iceland Air employees) in that airport that night was clear. The weather was so severe it was unlikely that most people would have ever gotten out that night, but the experience could have been much better, if things were handled differently.
Figure 2 shows the various critical elements of an organization (Culture, Strategy, Process, Team, and Tools / Infrastructure), as well as KEF’s assets, some of the failures that we observed, and ways these failures could have been eliminated, or at least, significantly mitigated.
While we go through this, keep in your minds that iconic scene in Apollo 13, where ingenious NASA engineers look at the problem from the standpoint of what they DO have to work with, dumping out a box of miscellaneous items that are on the spacecraft. I don’t know all the assets that KEF and IA had; it is probably far greater than what I observe here, but we did observe some powerful tools that could have been leverage much more.
Click picture to enlarge!
Let’s start with culture, because as a famous quote probably mis-attributed to Peter Drucker said, “Organizational culture eats strategy for breakfast, lunch and dinner so don’t leave it unattended.” The first thing that was apparent to me was there seemed to be a lack of strong and directive leadership. Different styles of leadership tend to work better or worse depending on the situation, but in a crisis, somebody needs to be in charge. Perhaps this was influenced by culture. For example, according to Wikipedia, “… an important key to understanding Icelanders and their culture… is the high importance they place on the traits of independence and self-sufficiency. In the June 2005 European Commission Eurobarometer public opinion analysis, over 85% of Icelanders found independence to be “very important” contrasted with the EU25 average of 53%, 47% for the Norwegians, and 49% for the Danish.” Some of the other Americans at KEF were telling us this similar observation about Iceland after having spent a couple weeks there, as well. Often our greatest strength is also a weakness at times. The lack of leadership caused a palpable uncertainty that we could feel from the small things (no one telling us to get out of the middle hallway at our boarding gate and line up along the wall) to the process (no one was parsing passengers into needs-based groups that could be handled en masse) to the holistic (KEF and IA seemed to be using the same processes they would use for a normal load scenario in this surge scenario).
No leader seemed to be stepping up and saying, “Stop! This is not working. Let’s do this instead…” (and deploying resources accordingly). Even the visible presence of an executive leader (maybe even dressed differently to show their rank) walking around the airport controlling things would have given certainty to passengers. We recommend that KEF and IA work on a command structure with clear roles and authority for the future. One asset that could likely help is Iceland’s tight-knit, homogenous population. They only have ~360,000 people (the capital is about half of that). It is much easier to mobilize and impact with a small and ideologically-aligned group (e.g. the Navy Seals) than a lumbering mass of people. So, KEF should have taken advantage of this (as we were told they did after 9/11 when planes were grounded).
The second major area to improve is Strategy. The short-term strategic failure was allowing so many flights to take off, knowing that (a) there was a high probability no connections could leave and (b) the airport did not have the capacity to handle the mass of passenger who would be stranded. This strategic error caused the crisis. Considering the expected low volume of the passengers who could fly out (likely none that day), KEF should have grounded / delay all incoming flights beyond the capacity of KEF to house efficiently. These passengers could have been housed at their origin airports, spreading the load. We understand that often you want to get people “part way home” at least, but there is a limit.
The long-term strategic error was not building surge capacity to begin with. I am not sure KEF and IA even were aware of what the max capacity was (or maybe they severely overestimated it?). KEF and IA needed to have tight understanding of capacity and a surge / crisis plan, beyond a Red Cross shelter.
Process was a big problem in the airport. There were many problems I observed, but let us focus on three that caused a lot of problems. First, KEF and IA seemed to make no attempt to parse the passengers into needs-based groups, such as:
- Special needs passengers with small children, pregnant women, the elderly, the disabled, who needed special care and more immediate attention
- Passengers who have been re-booked already and only need a hotel and food for the night
- Passenger who need to be re-booked
- Sub-groups that are going out early in the day vs. late.
If they had, they could have directed them into different lines where each like group could be handled en masse with much faster cycle times. And, special cases could be handled appropriately with more time and care (Lane Strategy).
There also seemed to be little preparation for this. Clearly, it took several hours to fly to KEF and almost double that time to get to the gate. IA could have prepared transport strategy, vouchers, bags of food to tie people over until the hotel, etc. This would have given comfort and certainty to the passengers and greatly sped up the process for the IA and KEF team.
The team also seemed unaware of what the capacity of hotels was, given that they had only gotten through half the final queue we were in, when they announced the KEF area hotels were full. There should have been a process to know the capacity of hotels in KEF, Reykjavik, etc.
It was clear that the team was severely understaffed. Hearkening back to Culture (leadership) and Strategy (building capacity), IA should have activated more resources to handle the crush of passengers. And, not all jobs require the same skill. KEF and IA could have built surge / temporary resources that have a basic level of training. These auxiliaries would allow the full-time employees to focus on the more sophisticated tasks. Here KEF had a big asset that was underutilized: The Red Cross / EMS. If I get lost in the wild, Iceland might not be a bad place to be, because it’s clear that Iceland had created a small army of EMS, loaded for bear (maybe literally?). They were geared up and all over the airport baggage, and ticketing, looking eager to help. But… no one was using them (except me — they were our saviors that night). Clearly, emergency care is their primary vocation, but in these situations, it’s likely there will be few traditional medical emergencies. A large number of the EMS folks could have been cross-trained to be auxiliaries for KEF and the airlines, hopefully with generous donations or direct payment to support the EMS organizations.
One note on training of the team: my wife’s experience with the IA employee was really poor, in that she did not seem to recognize or care that my wife was very pregnant and had been trapped at KEF for over 6 hours. I hope IA invests in some been training with an eye to customer service.
Tools & Infrastructure
Often companies and other organizations make the mistake of thinking that more or fancier tools can cure deficient culture, strategy, process, and team. They cannot. Tools and infrastructure are only force multipliers and enablers of culture, strategy, process, and team. However, the converse is not true. If you don’t have enough of them, it can frustrate the good progress you make in culture, strategy, process, and team. In KEF that night, this was clearly the case.
Let us begin with the lack of food options for thousand of people and the fact the food was so expensive. KEF and the airlines should have had agreements with the restaurants for emergency catering. It would have been easy to have tables of free or inexpensive fruit, water, juice, snacks, etc. Alternatively, lunch bags could have been hastily prepared for people to grab or handed out after they had been parsed into needs-based groups. (EMS or other auxiliaries could have set all this up).
The seating situation was terrible. There was not enough seating within security and virtually none outside of it. People were standing, literally, for hours. KEF had an asset here: it was a fairly large airport. Foldable chairs could have been set up by EMS / auxiliaries. This would have worked well in conjunction with dividing people into needs-based groups. People could get their luggage and sit comfortably with their sorted group while they waited for attention. For example, the airport team could have call 10-20 people at a time to go to an agent to help them, or they could address a block of 20-50 seats at a time with instructions, vouchers, etc.
The cab queue being outdoors in a blizzard was ridiculous. However, many winter airports in the US (e.g. O’hare) also have this problem. However, KEF had an asset here – cabs could drive right up to the terminal. Once again, some simple blocks of folding chairs would have allowed exhausted passengers to sit comfortably in the warmth and the cabbies could come get people in order as they arrived. Once again, parsing people by groups would have helped as well. If people were parsed by hotel area, buses, vans, and cars could have been efficiently filled. For example, my wife and I were the only people in a very nice 8 passenger van (75% of capacity was wasted).
We were told that there were problems both with getting a path plowed from tarmac to airport and on the roads to the cities of Keflavik and Reykjavik. This seemed bizarre in a place called ICEland. This was extreme weather, apparently for Iceland, but once again there was no surge capacity in the system. KEF and airlines could work with city and state to subsidize more equipment that the airport could acces for use inside the airport and on the arteries to it.
Finally, we have a sad tale of woe about an excellent asset underutilized when it was needed. Iceland Air had a very good automatic text and email system for updates. This could have been used to pre-sort people into needs-based groups and communicate instructions. Imagine this survey:
Hi! Welcome to Keflavik airport. Regrettably, we are having bad weather tonight, but we’ll keep you warm and fed while we sort it out. To help us do that, please answer the following questions:
- Are you or are you traveling with someone with special needs (with small children, pregnant, disabled, elderly, etc.) [Eric: Yes!]
- Sadly, all the connections to your destination for tonight have left or were canceled. According to our system, you have been re-booked on IA flight XYZ to Chicago (ORD) at #:## tomorrow. Does that sound right? [Eric: Yes!]
- We have the following options for you for accommodations. Which would you prefer?:
- Stay in the airport (we’ll give you blankets and pillows and food – no bathing options available).
- Stay on cots in our nearby Red Cross shelter (we’ll give you blankets and pillows and food – no bathing options available).
- Stay in a hotel that we will provide (we cannot guarantee whether this will be in Keflavik which is 15 minutes away or Reykjavik which is 45 minutes, but we will provide transport)[Eric: C, please!]
Thanks! When you arrive in the terminal:
- Please go to gate X, where we’ll give you some food and drink to go.
- Then proceed to baggage claim and collect your bags.
- Once you have your luggage, please proceed to the XYZ area (follow the signs), where we will give you vouchers and further instructions.
If you have any questions, please see the personal dressed in XYZ uniform. Thanks!
The other tech failure was the next morning in the check-in system. Although we had received our boarding passes the night before, we had to re-check bags. However, the system was not designed to handle this common exception case, so we (and everyone else) were sent to live desks of people to check bags. This idled ~25 very expensive looking kiosks that could have reduced the load on the desks by 50-80%.
A vision for the future
Looking at Figure 3 & 4, you can see designs of how the system COULD have worked to make a much smoother experience that would also increase throughput and capacity of the system. It would also result in much higher customer satisfaction, and maybe even the same or less cost.
Click picture to enlarge!
Click picture to enlarge!
In the grand scheme of our lives, this was a minor hiccup, but a very painful one for everyone stuck in it for 8-10 hours. Our thanks to the team at Keflavik airport and Icelandair who did keep us safe, and eventually put us on a plane home. Hopefully, they will find this operational analysis useful in the future.
And, hopefully, this little case study is also something to ponder for your own organization. Do you know what your maximum capacity is? Do you know where your bottle-necks are? Do you have a plan for surge and shock to your systems? Do you need to build capacity? All good things to conside.
p.s. If you like nature (or Bjork), I hear Iceland is a lovely place to visit!