Battle of the Beds

Club World vs Upper Class – let’s use LOOP methods to find out which delivers the most sleep?

“Club World delivers the business man ready to do business”. So said one of BA’s most memorable marketing campaigns for the middle cabin. It showed a busy business executive well rested after his trans-Atlantic flight and ready for a robust meeting with his unsuspecting colleagues, who had hoped he would be hungry and tired instead, the next morning.

Virgin Atlantic once put demo Upper Class seats outside the BA lounge pavilion at Heathrow Terminal 4, allowing prospective passengers the chance to try the bed in practice. Equally powerful – perhaps even more so. Guerrilla marketing, airline style.

BA and Virgin Atlantic have quite different brands. BA is serious and Virgin Atlantic a little more funky. You might say that if Club World is a flying office, Upper Class is a flying night club.

But which gets the more likes on Facebook? Perhaps Migacore could tell us (see article). But whichever it is, whether passengers got a good night’s sleep on critical but short US east cost to Heathrow routes will make a big impact.

So which seat delivers the most sleep? I reckon that BA’s Club delivers up to 3.8% (7 minutes) more sleep per passenger than Virgin Atlantic’s Upper Class. All in all the two competitors seem closely matched. Read on to find out how I worked that out…

[Caution: I have not worked with these airlines so some “reasonable assumptions" are involved. The airlines and seat designers would have the real data or something close to it. Think of this as an entertaining back-of-the-envelope to begin the year showing the sort of things that my seat map optimisation tool LOOP (see article) can do …]

Model & maps

I took a look at BA and Richard Branson’s 2023 summer schedule as in effect today and focused on the week 11-Jun to 17-Jun. This is not the school holidays, is not a bank holiday in the UK and is mid-season for business travel. It should be quite representative of “normal” ops.

If you want to understand an airline, looking at it’s fleet and schedule is always a great first step. Despite everything you might read about low fares or plush beds, the schedule is the product and it is powered by the fleet. It is the first thing I look at when I do a LOOP study to figure out what seats or galleys will make the biggest possible revenue contribution to an airline.

Schedules are complex things. Most of the time for airlines like these they are relatively simple – planes leave a London hub and then come back. But sometimes they reach their destination and go on somewhere else before flying back home.

These are called fifth freedom sectors as the fifth “freedom of the air” is the right to fly between two foreign countries as part of a journey originating or terminating in the home country. Here is a map of BA’s fifth freedom flights in the Caribbean.

These Caribbean hops are all short puddle jumpers. I doubt anybody is getting any sleep on these so have excluded them from the analysis. BA also have a fifth freedom flight between Singapore and Sydney which I have treated as a completely separate flight from the London to Singapore sector.

Richard B also moves some planes around the Caribbean. Here is a map of his fifth freedom flights, also excluded from further analysis.

But Virgin Atlantic also have some long range flights out of Edinburgh and Manchester to a few key American cities (map below). People will certainly be sleeping on these flights.

When it comes to London Virgin Atlantic operates all it’s summer 23 services out of Heathrow (map below).

I was surprised by how inconsistent their JFK schedule is – they use all of their aircraft with many different seat models. I am sure that some reasonably regular Upper Class passengers will be confused when they do not get what they expect, which could reduce their likelihood of buying again and harm the airline’s revenue..

BA has a Gatwick operation serving some bucket and spade or fun-seeker routes plus New York JFK (map below). I have no idea why they choose to serve Doha from Gatwick.

BA’s whole network ex-London is much wider than Virgin Atlantic’s.

What is a night flight?

I defined a night flight as one that arrives at it’s destination after 4am the next day or departs before 4am on the same day that it arrives.

These examples are “obviously” night flights. They leave late and arrive early.

BA106 0125 DXB LHR 0615

VS138 2200 JFK LHR 1100+1

Some others are a little more ambiguous…

BA52 1335 SEA LHR 0650+1

BA282 1545 LAX LHR 1010+1

BA169 1225 LHR PVG 0755+1

VS250 1525 LHR PVG 0940+1

BA282 from LAX is perfectly timed for sleep if you only arrived a day or two ago from Europe and are now heading back. It’s 1545 departure is 2345 in London. It is also perfectly timed to do some work, have a meal and watch a movie before resting in your hotel the next day if you are on California time.

I took BA169 to Shanghai and the now terminated but similarly timed BA17 to Seoul in 2019 in First and Club World respectively. BA offered a breakfast service as the second meal on both flights but as far as I could see hardly anyone slept.

The reality is that flights like these will have people sleeping, but to a lesser extent than others. I will bear this in mind in the modelling later.

I took capacity data from aerolopa.com and assumed that BA will complete roll out of Club Suite on 777s but still use yin-yang Club World on other aircraft type by the time the summer 23 schedule flies.

Some high level results

The table below shows that BA’s network seems to be more calibrated for night flights than Virgin Atlantic’s and so sleep may be more important to the average Club passenger than Upper Class traveller.

Out of the total number of services, BA has 4.5% more night flights than Virgin Atlantic. Assuming an even 85% seat factor on every flight (a big assumption I know) 7.3% more of BA’s passengers are likely to be on an overnight flight.

But critically, because BA’s overnight flights are longer than Virgin Atlantic’s 10.1% more of BA’s capacity (ASM) and output (RPM) are on overnight flights compared to Virgin Atlantic.

Maybe Upper Class is not so much of a flying night club after all…

Modelling sleep

I also took the model a little further to think about how people might sleep on flights. As I showed above, a lunch time flight to Shanghai might well be overnight technically, but not many people will necessarily be getting a full night’s sleep. And some of BA’s monster flights like Singapore, which are not served by Virgin Atlantic, are so long that nobody will sleep all the way.

Meanwhile it is is not uncommon to get a few hours of shut eye on a westbound trans-Atlantic flight operating entirely in the hours of daylight.

For now I have excluded meals and loo visits for simplicity, but these could be added to a more comprehensive LOOP study with collaboration from an airline’s cabin service and catering departments. I imagine that these would be similar if not identical across both airlines.

Air New Zealand were quoted in Aircraft Interiors International magazine a short while ago on the subject of sleep. They did research showing that on overnight flights most people want to sleep for four hours but then they start to wake up.

Based on this I constructed this table of “reasonable assumptions” about how long people might sleep for on flights, depending on the time of day. The % of pax who sleep column is never zero because there will always be some people whose body clock is still in a different time zone or who just close their eyes when they are on a plane.

The wake rate per hour of sleep shows the number of people who I assume will wake up once they have had a certain amount of sleep, if the flight is long enough. On clearly overnight flights only a few people wake up before Air New Zealand’s four hours. On day flights I imagine that people will tend to sleep for only a few hours.

Some people take sleeping pills to get more than a full night’s sleep. But I limit sleep to ten hours. The model forces people to wake at 20 minutes to landing.

Based on these parameters the average BA Club World passenger will sleep only seven minutes longer (3.8%) more than a passenger in Virgin Atlantic’s Upper Class.

But is this close to the truth? Maybe Virgin Atlantic’s passengers are more likely to gather at the bar and party the night away. Perhaps BA’s passengers work the small hours on presentations. Or then again, perhaps not.  Until smart seats and fabrics start collecting data on sleeping passengers (see article), good people to ask would be the cabin crew.

The important thing is that both airlines have invested lots of money on giving their passengers a good night’s sleep from two quite different products. Based on my model the average passenger could get nearly three hours of sleep, even considering that many flights operate in the hours of daylight.

Based on these figures, it looks like both airlines have a close product. I hereby declare the battle of the beds a tie!

So why does this matter?

When airlines are deciding which seats to use or suppliers are offering their seats to a particular carrier, how the network is structured is relevant. Calculations like this help people understand how important sleep is to their passengers vs. competitors.

If it turns out that a competitor’s schedule is similarly calibrated for overnight ops., getting the best possible bed could be critical from the buying airline’s perspective. From the supplier perspective, the buying airline might have higher willingness to pay for better bed seats than some others might.

oliver AT ransonpricing DOT com