An interesting take on revenue management and the last seat sold


Airline revenue management is, to say the least, quite complicated. I’m reasonably convinced that a decent amount of the work is performed with a Ouija board and perhaps offerings of rum and cigars to Jobu, but I’ve been promised by those actually doing the work that these figure in only rarely.

And yet there are still many who write about it and opine on the logic behind the decisions being made one way or the other from the airlines. Statements like, “but the flights are always full” as a justification for keeping underperforming routes are common and easy enough to dismiss – nearly any plane can be filled on a route if the fares are low enough – but there are some other statements which require a bit more thought and consideration. I don’t know that I’m right either, but I’d like to think I’ve got a fighting chance.

One of my favorites is the idea that the last seat sold on the plane represents the profits. The general theory is that cheapest fares show up right when the schedules open  and then tickets get more expensive over time until the day of departure where walk-up fares are sky-high to take advantage of business travelers with inflexible schedules. And those last seats sold are where the profits from operating the flight come from. Except that the pricing trend described has not actually been true for 20+ years now. And the idea of the last seat sold being all profit might not be so accurate either.

During today’s earnings call with investors and media JetBlue President Robin Hayes was tasked with answering the theoretically simple question of “Why not add more seats like other airlines are?” The question was mostly targeted at the carrier’s A320 aircraft which have 150 seats currently with 34″ pitch at a minimum and 38″ pitch in the “Even More” seats which carry an incremental charge to book. CEO David Barger chuckled when the question was asked and said, “Every quarter I ask Robin the same thing.” And then it was Mr. Hayes’s turn to actually provide the answer.

Every quarter the math is compelling.… It really is an economic decision.

Well, duh. I mean, if you’re going to claim that you’ve analyzed the market and made the decision then I’d hope it is based on the numbers. Of course, Mr. Hayes was not quite done.

Not only do we displace the “Even More” revenue that we sell but we need an extra flight attendant.

The key to the magic 150 number is that JetBlue can get away with staffing their aircraft with only 3 flight attendants per the FAA requirements. At a 50:1 ratio (based on the seats on board) the airline is maxed out without adding another crew member. And those costs add up in a hurry. There’s also additional fuel burn due to higher weights on board (seats, passengers and bags) and other considerations. But it isn’t only about the cost side of the equation.

And that’s where things got interesting. It basically flies in the face of the previously asserted “last seat” theories.

The fare that you’re getting for that last 6-12 seats is a fraction of the average fare; you’re chasing the low fare there.

So it turns out that selling out those last few seats actually is NOT driving the profit on the planes. At least not the way JetBlue does it.

Best I can figure (and this is all my speculation, not Hayes’s commentary) is that the company acknowledges there are going to be empty seats on most flights. And that makes sense given the average load factor numbers. The key is a combination of minimizing those empty seats while also not conditioning the market to assume that the “normal” price on a route is spectacularly low. If you routinely fill 140 seats on a plane then there is still the opportunity to sell those last 10 at full fare to an unexpected business traveler so you’re protected against losing the top-yielding customer. And adding 12 extra seats doesn’t actually change the number of those walk-up fares you’re likely to sell. All it does is open up more seats to unload during sales.

The flip side is that – assuming the costs are relatively similar – the extra revenue should simply augment the revenue stream without hurting the bottom line. But when the company is able to routinely sell Even More seats as a very nice premium and also keep their costs down it is easy to see why the numbers make sense in this regard.

As an additional data point remember that AA has blocked 4 seats on their new 737-800 layouts to drop the FA requirements down. Clearly JetBlue isn’t alone in this thought process.

I wonder when the rest of the world will catch up with them.

Never miss another post: Sign up for email alerts and get only the content you want direct to your inbox.


Seth Miller

I'm Seth, also known as the Wandering Aramean. I was bit by the travel bug 30 years ago and there's no sign of a cure. I fly ~200,000 miles annually; these are my stories. You can connect with me on Twitter, Facebook, and LinkedIn.

11 Comments

  1. Interesting, especially when juxtaposed with AA’s earnings call comments the other day when they said they’re adding seats to the 738’s, to either 162 or 164, while retaining MCE, but using slimmer seats. I suppose with an extra dozen or so seats, they feel they can absorb the additional FA cost and still increase revenue. Only time will tell, I suppose.

  2. One of my favorites is the idea that the last seat sold on the plane represents the profits. The general theory is that cheapest fares show up right when the schedules open and then tickets get more expensive over time until the day of departure where walk-up fares are sky-high to take advantage of business travelers with inflexible schedules. And those last seats sold are where the profits from operating the flight come from. Except that the pricing trend described has not actually been true for 20+ years now.

    The fare that you’re getting for that last 6-12 seats is a fraction of the average fare; you’re chasing the low fare there.

    You’re confusing terms here. Mr. Hayes is not saying that the last seats sold are lower than average fare, he’s saying seats added are added to the front end of inventory, in the sense that the added seats are available from the very beginning and will be sold at the cheap prices. Of course no one has ever argued that adding 10 seats means you can sell 10 more last minute, high price, full fare seats! That’s just a strawman.

    None of the words that Mr. Hayes spoke contradict the simple fact that the last seats sold on a flight usually have a far higher fare, thus far higher profit margin, than seats booked months in advance by leisure travelers. Perhaps there are some inaccuracies to the model you derisively disparaged in the portion I quoted of you above, but Mr. Hayes’ words do not contradict it.

  3. Why block seats when you can opt not to have them installed in the first place? Then you’re saving the weight of the seats and the need for the extra FA, and probably some negligible amount on the cost of the aircraft.

    1. In AA’s case it is middle seats in 2 rows which are blocked. That’s the best way to install 150 seats given their layout constraints. Otherwise the math doesn’t work and you end up with too many or too few on one side of the aisle.

      1. I found myself next to one of those seats on a recent flight. Not a bad deal at all paying coach. Now I understand why it was configured that way.

    2. Delta actually did this at one point several years ago (I don’t remember which narrowbody model, but I think it was the 738). They didn’t want to change the pitch of the remaining seats, or move the pax service units, so they just took out two rows of seats, leaving a small section in the back of the aircraft where the seats were missing. It definitely looked weird.

  4. First off, when I was an RM analyst, I did keep a Magic 8 Ball on my desk, so you’re not too far from the truth… 🙂

    That being said, I think you’re using two different definitions of the “last seat sold,” which is confusing the issue. One definition is based on time, and the other is based on fare. The literal last seat sold, the seat sold to a passenger close to departure, is indeed generally the most profitable seat on that flight, since it was likely sold at the highest fare. However, adding seats to an aircraft doesn’t give you any more “last seats” to sell as the highest fare. The “last seats” on a larger aircraft, looking at them from highest fare to lowest fare, are indeed sold at the lowest fare.

    Think of it this way. There are a limited number of passengers willing to pay a high fare for a flight, but lots willing to pay a low fare. The goal of revenue management is to save just enough seats for those willing to pay higher fares, and sell the remaining seats to those willing to pay the lower fare. Adding seats to the plane doesn’t increase the number of people willing to pay the high fare, so you end up filling the extra seats with low fare passengers.

    An example: for a given flight, you have 10 people willing to pay $100, 20 willing to pay $50, and 100 willing to pay $10. If I have an aircraft with 50 seats, I want to sell 10 $100 seats to all of the $100 people, 20 $50 seats to all of the $50 people, and 20 $10 seats to a subset of the $10 people. If I increase the size of the aircraft to 75 seats, I am simply selling the additional 25 seats at $10 each to more of the $10 people. This is was was meant on the JetBlue earnings call when they were talking about the below-average revenue associated with adding seats.

  5. I do understand the difference between the last passenger to buy a ticket and adding more seats to sell more of the “last seat” on the flight. But I also think that many miss the part where there are almost always “last seats” still for sale on many flights so simply adding more seats to be able to sell more of them isn’t rational math. That is a view which does get tossed around far too often and one which has diminishing returns. Mr. Hayes correctly called those out in a manner which I found quite illustrative and refreshing. Nothing more, nothing less.

    1. Indeed. It’s also important to remember that in RM, it is actually expected to end up with empty seats, since demand is probabilistic. Even if a high-fare passenger only shows up to buy that seat one time out of every three or four departures, it can still be more profitable to keep that seat empty, waiting for that high fare passenger to show up occasionally, than to sell it to somebody else for less money.

      1. Or, depending on the market, for the carrier to oversell the flight knowing that getting that last-minute full-fare pax generates more $ than the cost of issuing voluntary (or invol) denied boarding compensation vouchers. I seem to recall that these vouchers “cost” (in actual $ terms to the carrier) only ~30% (?) of the certificate value — some of these pax don’t end up using the vouchers, some pax end up taking an additional trip they wouldn’t have taken otherwise and end up spending incremental $ to bring along friends/family, etc.

  6. Great post. I think your missing one other thing specific to JetBlue. They have a large customer base that exclusively seeks them out, or at least high priority. They aren’t a discounter anymore for sure. One reason is their seat pitch.. They are for sure the anti Spirit. With IFE being added anywhere, this is a huge differentiator and if JetBlue added a few more seats, for only a little extra money after stewardesses, the risk of damaging the brand may end up being a money loser. Requires overall brand value thinking rather than strict revenue management bean counting, which JetBlue has always been strong in (the former).

Comments are closed.

BoardingArea