What should United Airlines learn from its latest fiasco?

What should WE, possible future fliers, learn?

And what should the other airlines learn?

The extra brutality is not the real issue. United, or XYZ Airlines, can always claim “it is not us; it is they, police or local authority security forces”. The real message to any passenger is:

“XYZ Airlines would, most probably, fly you to your destination, provided you have a valid ticket AND provided the airline does not have something more important to do. If XYZ would fail to fly you on time, they might compensate you, or maybe not.”

Should United and the other Airlines stick to the above message? Of course they should! Because, what can we, the passengers, do?  Do we know another airline with significantly better commitment to its passengers?

Why should United and the other airlines contemplate to change their basic paradigm of trying to exploit the seats of all their flights? This is absolutely in-line with TOC, isn’t it?  Some flights have the limited number of seats as a micro-constraint.  Overbooking is an exploitation scheme, because many times passengers don’t show up and then it is a waste not to sell the seat to a paying passenger.   Now, if there is a real need to fly several crew members, because otherwise another flight might be delayed or cancelled causing very high cost, then it makes perfect sense to nicely ask some passengers to give up their seat for some compensation the airlines think is fair.

The only question is whether WE, potential future passengers, agree to this widespread win-lose scheme of the airlines.

The scope of question is much wider than this specific case.  There are many other cases where we are, very politely, asked to change our plans because it is too expensive for the airline to fulfil their commitment. The Israeli airline, El-Al, has lately cancelled, at the very last minute, many flights because of lack of pilots.  Let me stress the point: cancelled at the last minute!  El-Al management blamed the pilots, and the pilots blamed the management.  It is nice to have somebody else to blame.  Do I care who is right?  Blaming does not solve anything.

I hope some airlines would see the opportunity to develop a decisive-competitive-edge policy of behavior and make it work. I do not claim it is easy – just that it is possible!

Remember: the real constraint of any business is the market demand! If you fail to subordinate to the demand then the future demand would go down, and then you won’t have an internal constraint as well.

It is still possible to have an internal constraint, but the constraint has to be exploited only to the level that still provides good overall subordination to the market.

Let me point out that having to add crew members to a full flight is a good example of interactive constraints! Another operational wisdom the airlines have to learn.


Published by

Eli Schragenheim

My love for challenges makes my life interesting. I'm concerned when I see organizations ignore uncertainty and I cannot understand people blindly following their leader.

15 thoughts on “What should United Airlines learn from its latest fiasco?”

  1. What if some airline decided to guarantee to carry you from point A to point B. What would such a service be worth? What would it cost?

    For the sake of an example, Let’s say each plane carries 200 people.

    The first rule is that ALL legs cost $200 regardless of how far in advance one books. The only reason to book early is to be sure to gain guaranteed travel.

    The second rule is to NEVER overbook flights at all. What does this cost? Nothing more than normal, unless the plane is fully booked. For fully booked planes, let’s assume that 5 people will not show up and they pay a small penalty for refund or rebooking, like $50. If I assume that 3 of these people were planing to travel in 2 legs and the rest for one leg, then the airline has lost $1,350/fully booked flight ($200/leg/person x 2 legs x 3 people + $200/leg/person x 1 leg x 2 people – $50/person x 5 people).

    Third rule, for NON-GUARANTEED travelers, the airline offers unused capacity (after it re-positions its own people) to allow STANDBYS for a flat price of $50 per leg. This is such an attractive offer that casual travelers queue up to buy unused seats. Therefore, the company gains another $250 for the unclaimed seats on each fully booked flight. Now, the cost of the guarantee is down to $1,100/fully booked flight.

    For lightly booked legs, the company garners an average of 10 last minute standbys, gaining $500/leg ($50/leg/person x 10 people) per lightly booked flight. Assuming there are 4 lightly booked flights for every fully booked flight, the standbys contribute $400/flight (($500/leg x 4 legs) / 5 flights). This gets the damage down to $700/fully booked flight ($1,100 – $400). Assume the average number of guaranteed passengers carried on non-fully booked flights is 175. Out of every 5 flights which include one fully booked flight with 195 guaranteed passengers and 4 lightly fully booked flights with 175 guaranteed passengers each, that makes 895 passengers. Recall the cost of not overbooking per fully booked flight is $700. Divide that by 895 passengers and it costs the airline about 78 cents per passenger per leg to offer the guarantee.

    Back to my first question of what the service would be worth. If I needed to be somewhere, I’d be willing to pay $10 per leg to fly with an airline that thinks of me before they think of themselves. Wouldn’t you? I bet there are many other customer-centric policies that are quite the opposite of what other airlines deploy. Hum, a premium airline without a premium cost structure. Nice for all.


    1. Henry, you outline a possible way to evaporate the basic business cloud: how to satisfy the customers without extra cost. What I miss in your analysis is the possible benefit of overall higher demand, on top of the extra $10. We should add the condition that the company succeeds to convince that the commitment to every customer is of utmost importance. I also miss what could happen if the airline continues to do what it does today, but another airline makes the change.

      You are right that for some travelers there is flexibility in their timing and tolerance, while for others timing is critical. There could be different terms to various customers, but within those terms there should be clear and visible commitment to do whatever-it-takes to meet the commitment.

      There cannot be a guaranty for arriving from A to B on time. There are true cases where weather or technical flaws make the commitment impossible. Problem is that the airlines use them to justify cancellations and very long delays, where the real reason is cost saving. It is a challenge to convince the customers of truly making all the efforts to meet all commitments, but it definitely can be done. There are some risks of going this way, but because of those risks not every airline will go this way, which means the decisive-competitive edge will be truly effective. This insight, Henry, I just learnt from you.


      1. Yes, Eli. You make great points I just inferred for the sake of brevity. I also agree that an airline with customer protective policies will pull enormous market share in an industry that enacts company protective policies (due to a long-term industry-wide track record of losing money and bankruptcies).

        I think there can be a guarantee for getting people from A to B. Remember, it does not have to be enough to repair the damage to the traveler, just enough to ensure the airline behaves in line with customers’ best interests. That is what swings the market.

        Heck, let’s buy an airline!!!!


        1. I would raise the question on whether systemically the airlines has imposed processes that encourage a constraint. In their case maybe they still believe that constraining the market is good for business long term. I have not done much of a study on this and I am leaving next week on vacation to Cancun and hope SW does not throw me off the plane. But Henry’s point about buying a airlines is probably being looked at by someone. I believe that business’s can only constrain the market for so long. After some period the market finds new paths to be satisfied.


          Liked by 1 person

  2. Another point of view:
    Define the goal, in the long term, to be our REPUTATION in the market, with the condition of not having negative profit. This definition implies that we must have always a protective capacity buffer. Yes, we accept not to squeeze to the last miserable cent. The capacity buffer are some few seats that we deliberately leave unsold, just for special cases as to add crew members sometimes. Many will see this as a heresy, because we live with a sick paradigm: make more money, make more money, make more money.
    If the company runs a big risk for leaving some five seats unsold, that means that the business is unviable and may need better kind of management or owners.
    What is the progress made by humanity in the last 50 years if we are not capable of having good protective buffers in our organizations? What real benefits have we obtained from technology, IT, comunications, etc in the last decades if we finish managing companies in childish ways?

    Liked by 1 person

  3. There are two possible solutions:

    1.The simplest solution that occurs to me is a combination of protective capacity buffers and stand by passengers. Keep everything else the same as currently but reserve 3-5 seats as buffer. Take stand by bookings against these 3-4 seats at 50% prices and take this decision with the understanding that they can be bumped off if a last minute crew member needs the seat. So only 3-4 seats experience uncertainty and they pay much lesser as a benefit.

    2. Sell all the seats as usual but when such a need arises, offer massive discounts to the volunteers or a kind of reverse auction. They could give away these as free as well ( full refund along with a free flight coupon). As the prices drop, there will be many willing passengers who will happily take the offers. The situation such occur as a lottery to people and each time they get onto a flight, they are actually hoping they get bumped off. The mathematics is simple. Such a situation would rarely arise. Let’s say 1 in a 100 flights. So the airline sold the last 4 seats at high prices for 99 flights and gave away free seats on 1 flight. Net net make full money on 98% occupancy.

    Based on the actual statistics of the frequency of overbooking and experience on how readily customers accept discounted tickets as compensation, a combination of the above 2 can be used as well.


    1. Thank you Julio and Anuj, your suggestions could be the start of a solution. A full solution has to be developed together with a management that sincerely like to explore the possibility of gaining new market through much better treatment of passengers.
      I like to highlight three points:
      1. Protective capacity for an airline is very expensive. This NBR does not rule out the idea, but it forces to do calculations, like Henry did and probably wider. The compensation for the protective capacity has to be more sales, either more demand or willingness to pay more.
      2. The real problem is not just the seats and overbooking. I even think that very careful overbooking and willingness to pay truly high compensation could work, because there is are segments in the market that would love it. The real problem is overall COMMITMENT and visibility to the passengers. The approach should encompass long delays and cancellation and make sure those are truly rare and that all efforts are taken not to let it happen. We all know that even with protective capacity and buffer management in some rare cases few orders are late and few shortages occur.
      3. Once such a solution is in place, the challenge is to make it seen and believed. This could be the more tricky part of the whole solution.


  4. when the airlines try to exploit every seat as micro-constraint, actually they put themselves in much less buffer to be more flexible when there comes up uncertainty. There has been succussful business model developed by Singapore airlines, Changi airport and the government collaborating together, who leverage the fundamental facility as critical resource to provide attractive service to the passengers. Meanwhile Singapore airlines embrace the competition to elaborate with smaller airlines as alliance sharing some of their short-distance routing so they can improve the capacity without investing on the aircraft, customer experience is always the most important value proposition they consistently create and deliver to different market segments.


  5. Foolish me always thought airlines already addressed this. The less the customer pats, the lesa flexibility. So if I pay $200, if I dobt show or want to change, tough! Thr airline gwts my money whether I show or not. And if I pay 3-4 x tgis price I cam change. But again, if they have got their prucing right they do not lose out in the long term, amd they can try and get a windfall bit of extra profit by selling last minute changes on standby.

    I think the problem is they got greedy and started to cut protective capacity and trued to predict the future and deliberately oversell tickets in the expectation of no shows.

    Maybe an offer for significant compensation if you are bumped might work? If not, then maybe thw rwgulatira should enforce this to prevent businesses continuously pushing the boundries of reasonableness and ethics!


  6. Eli,
    Delightful discussion as usual. My view is that the airline industry is one single system. If Sothwest discounts fares on one route, all airlines on same route do the same. They are masters at big data analysis and have excellent algorithims to determine how many oversolds on any plane on any route at any time. They constantly maximize utilization across their systems and will use competitor capacity with regularity to balance demand and supply. The true external constraint is airport gate availability. Unless more gates available, no more planes with spare seats.


  7. Very interesting follow on to a very thoughtful blog by Eli. The interesting thing is that this dilemma is no different than many of the business models today. Squeezing profit under the demands of growth and continuous improvement always take us to the same place. Which is that all is fine until Murphy strikes. In the case of UA Murphy struck.
    Don Myers


  8. I understand United stopped at $1,000. Perhaps at $2,000 somebody would have taken the offer?

    I believe that in the majority of cases, there has to be a good enough compensation that entices passengers to volunteer to take a next flight. If it is good enough then it’s a win-win. Isn’t this a basic assumption of how overbooking should work?

    Give the front managers authority to keep increasing the compensation when overbooking happens. In the long run, it is much cheaper than the hundred million PR disaster that United has been living in the last days.


  9. I understand Ray’s claim that all the airlines are one single system, because they imitate each other so heavily. Together they are like one monopole taking advantage of the customers. I assume many airlines are happy with the fiasco of United, after all every airline looks to gain more on the expense of the others. Within the big system there is a constant fight between smaller systems.

    What would happen is a large enough airline would “change the rules of the game” and offer win-win service? Such a move could make a dramatic change.

    My main complaint is not so much overbooking, because I agree that when the airline is ready to pay more they would find the volunteers and then it is a good enough win-win. This policy was not good enough for United. My complaint is that the airlines offer win-lose as a basic policy. They do not always offer fair compensation when they abruptly change the plans of their passengers. I wonder what would the airlines learn from the United case and whether they use Machine Learning to understand how their efforts to exploit the micro-constraint are changing the customers behavior in the long run.

    Kevin Cai mentioned Singapore Airlines, which has a much more favorable reputation than any other airline. The problem is that Singapore Airline offers very limited competition, so for the rest of the world we still face a monopole. But, the one that will break the monopole would gain a lot.

    I heard that terminals are the overall constraint. Are they properly utilized?


  10. Eli, you state, “the real constraint of any business is the market demand!” Isn’t the real constraint of any business in a market economy the capacity of their new product development process to meet customer needs, and thus grow the business? By what mechanism does “the market” act to limit the Throughput of a business? And when a business goes out of business, is the root cause “the market”?


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s