When Scott Kirby left American to take over as President of United, he noted that United needed to do a better job of attracting connecting traffic. This week, United put out a big summer expansion plan, and guess what? It’s all about connections.
You may still be scratching your head as to why an airline would ever want to focus on connections. We’ve heard for years that travelers flying nonstop are much better for the bottom line. They pay a premium to fly nonstop while at the same time cost less because there’s only one flight instead of two or more. It is still true that airlines want to serve those high-dollar business travelers, but a lot has changed since those days.
Most notably, the rise of low cost carriers focusing on big markets has changed the game to some extent. Now you may have all kinds of low fare competition flying from, say, Philly to Ft Lauderdale (to use an American Basic Economy market), but you have almost no competition flying from State College to Ft Lauderdale via Philly. I took some random days this summer where availability was good, and the cheapest fare from Philly to Ft Lauderdale was $162.40 roundtrip (regular coach is $40 more than that Basic Economy fare). But if you add State College on either end and have the same exact flights between Philly and Ft Lauderdale? The roundtrip price is $413.59. Southwest and Spirit may be able to get you from Philly, but they can’t touch State College. And those fares are good money.
You might be wondering why I’m giving an American example in a post about United. It’s because American (and really, US Airways before the merger) was a pro at this. Look at Charlotte. There isn’t much local demand there, but there are so many smaller cities with limited demand that American can feed them all through Charlotte and turn it into a powerhouse. Did you know American serves 16 cities in the Carolinas alone? It’s insane.
And of course, Scott Kirby knew this all too well. So when he got to United, he found an airline that wasn’t looking at things the same way. It did have something working like this to some extent in its Denver hub, and Denver, it turns out, was the most profitable hub in the system. Outside Denver it was a different story. Look at a place like Chicago where American has much greater depth. It serves the Champaigns, Dubuques, and Bloomingtons of the world while United doesn’t… or didn’t.
So it was no surprise to see United announce it would add or upgrade service in 31 markets earlier this week, and most of them are focused on connections. Here’s how the markets that get new service break down.
Chicago Goes Small
I mentioned how United didn’t serve some of the same markets as American from Chicago, but it’s building those up. You’ll now see nonstops from Chicago to Charlottesville (Virginia), Champaign (Illinois), Columbia (Missouri), Reno/Tahoe, Rochester (Minnesota), and Tucson (goes year-round from winter-only). What do all of these have in common? American flies every single one. United is also adding Spokane, a market American doesn’t fly.
Chicago also gets more frequencies to Ft Lauderdale and Ft Myers, undoubtedly well-timed to connect to these and other flights.
San Francisco Goes Broad
San Francisco is seeing a different kind of build-up. It’s getting nonstop service to larger cities that don’t have service today, but these are markets that don’t work on their own. They only work because of connections. Look for new nonstops to Cincinnati, Detroit, Spokane, Hartford, Munich (summer only), and Santa Rosa (California). There will also be an extension of winter flights to New Orleans and Kalispell to be year-round.
These are almost all decently-sized markets, but they aren’t ones that work without a ton of connecting traffic. Munich, of course, is large but there’s no way SFO can support two daily nonstops (Lufthansa has another) without a ton of traffic feeding both ends.
The one outlier here is Santa Rosa. That’s a market that American just started from Phoenix. I was quite surprised to see United start it, however, because it’s a mere 1.5 hour drive from there to SFO. There will not be a single local traveler on these flights yet they’re operating 3 times a day. But guess what? United thinks there’s opportunity to connect a bunch of people from there via SFO to the world. As an added bonus, this will allow United to now choose between Monterey and Santa Rosa when it starts canceling flights when the fog rolls in. So congratulations on that, Monterey.
Denver Gets Tweaks
As I mentioned, Denver is the most-profitable and most connection-dependent hub in the United system. That means it’s where Kirby wants it to be, so he’s just extending his reach a bit. There will be new nonstops to Columbia (Missouri) and San Luis Obispo. There will also be more frequency to Albuquerque and Florida.
Washington/Dulles Gets Frequency
There honestly isn’t a ton happening at Dulles, but we are seeing some of the smaller markets get increased frequencies to build-up the hub a bit more. That includes more flights to Jacksonville, Providence, Portland (Maine), and Roanoke (Virginia). The Ft Lauderdale flight which used to be winter-only now goes year-round.
Newark Goes West
Newark admittedly isn’t seeing much of a change, but if there is a trend, it’s that flights are going west. Newark gets a new nonstop to Sacramento and a second one to Portland (Oregon). Its winter-only Salt Lake flight goes year-round. It also gets more flights to Atlanta and Detroit.
Houston/Intercontinental Gets Little
Ok, so Houston really isn’t getting much at all. It gains a nonstop to Springfield (Missouri) and added frequencies to Atlanta and Charleston (South Carolina). But that’s it for now.
Los Angeles Gets Nothing
Los Angeles is noticeably absent on this list. Why? Because LAX is not a connecting hub. I’m in no way surprised to see LAX left off completely.
These changes are substantial, and it shows the direction United is looking to go.
57 comments on “United’s Big Summer Expansion Confirms Its Focus on Connecting Traffic”
Not much local demand in Charlotte? Metro population almost 3 million now….guess you mean in relation to larger cities
Google says Charlotte’s metro area population is 2.38 million. That’s about the same as Cincinnati, and compare the operations at Delta’s Cincinnati hub to those of American in Charlotte…
I think Brett’s point is that compared to almost all the other true hub cities in the US, Charlotte is among the smallest in terms of local demand and/or population.
There would also be especially little local demand for the many flights within the Carolinas and other under 300-ish mile flights where almost everyone would just drive to save time and money.
Great point. Similarly, there is relatively little demand to places like Columbia, SC from even major airports in the US (see: David Samson & United’s PANY/Newark scandal), and vice versa, but running pax through the Charlotte hub makes things work.
Charlotte is peanuts. Sorry.
Pedro – It’s not a big market. There may be a decent number of people there, but there isn’t the kind of base to sustain anything near what the airport has today. And that’s why there is such a high percentage of connecting traffic there.
Seems like a lot of additions – are the added fights being offset by cuts? higher utilization? new airplanes? I would think if you want a more hub / banked schedule that would also increase ground time.
Noahkimmel – I haven’t heard about that. But the re-banking work they’re doing could free up some additional aircraft time. Also, on the 50 seat end, I’m not entirely clear what’s going on because Air Wisconsin just came back to United and will start picking up more. Presumably this is mostly to replace existing aircraft operated by others, but I don’t know for sure.
True, but these additions are more imminent I don’t think ZW comes back online until 2018 (?) at UA.
I suspect 50 seater utilization is on the lower end and they can easily absorb the new additions.
This is admittedly reading way, way between the lines regarding Dulles, but a few things add up here:
IAD’s strength is a decent O/D base, particularly for international traffic. UA does a lot of connecting there, but my understanding is that it’s not nearly as high of a connecting percentage as, say, CLT.
The airport has a lot of room to grow, unlike EWR.
United signed a new lease at IAD; they’re working with the airport authority to lower costs; and UA is doing more flying there recently. One way to bring down the fixed costs is to grow traffic and spread those costs out over more flights and more passengers.
And there’s this tidbit from an Oscar Munoz visit to IAD to speak with employees: https://skift.com/2017/02/01/united-ceo-flying-with-subpar-product-and-subpar-schedules-doesnt-work/
“United has historically used Dulles as a trans-Atlantic gateway, rather than a domestic connecting hub, though that could change, Munoz said.”
So… that *could* change.
Maybe United and IAD could finally replace the decrepit C/D concourse.
I would love to hear from Cranky building on this, possibly even in a stand-alone post. It seems like IAD has a lot of the same potential CLT had before AA/US built it up, or JFK did before JetBlue exploited it. I know its geography is not quite as advantageous as ATL or CLT, and it has the problem of BWI and DCA pulling away local fliers, but the DC metro has a lot more population than CLT and the equivalent of ATL. What would keep UA from turning IAD into a “best of both worlds” hub with domestic connections similar to CLT and ATL, but also international connections like EWR or PHL, but with a lot less congestion than those two?
United previously grew DOM connections at IAD (mainly north-south along the East Coast). Once in competitive response to USAirways growth at IAD (1998/99ish) and once to respond to Independence Air (2006?). If I recall, they had at least 4 daily R/T to even the smallest markets with higher frequency to markets like ORF, LGA, BOS. The four main banks were approx 8am, 12pm, 5pm and 10pm (this included new INTL flights to S. America, Mid East and additioanl frequencies to LHR/FRA/CDG).
Tory – There are a lot of issues with Dulles. First and foremost it’s an extremely expensive airport. So that hurts the ability to serve a lot of these smaller markets. Second, there’s a geography problem. Dulles may not seem all that far from Charlotte, but a lot of intra-South connectivity (where Charlotte really makes its money) doesn’t get served very well through Dulles. It adds a lot of time and isn’t competitive. Third, the facility isn’t adequate. United’s “temporary midfield concourse” has been there for 30 years with no funding to replace it.
Ah, thanks. That explains a lot. I didn’t realize the money was really intra-South. Obviously longer-distance or inter-regional connections would have a lot more competition with lower margins. I’m assuming that’s the same reason Denver makes so much money for UA is the intra-Mountain West flying where other hubs just don’t work.
Cranky, the day United builds up Dulles is the day Amtrak becomes profitable. Dulles’ geography problem is that it is less than 200 flying miles from EWR. No way that the Continental hub is de-peaked for Dulles. None. It’s one or the other and if Dulles were not the national capital, it would have been closed long ago.
Definitely not arguing Dulles over EWR. EWR is a premier hub, but it is for all intents and purposes maxed out, and focused on international connections more than domestic. Given the congestion, I think most people would avoid it for domestic connections at all costs. IAD would be more of a domestic hub connecting the midwest and northeast with the south, although with some international connections thrown in to support the DC market. For comparison, AA has PHL+CLT, DL has JFK+ATL, and UA has EWR+IAD, each performing complimentary roles.
1) Good luck driving SFO – STS in under two hours at rush hour or on weekends. It can be done, but there is a lot of traffic on 101. But, then again, I would be very leery of booking this flight due to a high probability of cancellation since SFO can’t handle weather.
2) I was surprised that UA never served CMI since I was a student in there in the 1990’s. They used to serve DNV, but it was AA, NW, and TW serving CMI when I lived there. Now, it’s just AA since DL pulled out twice.
19th Avenue alone is the reason these flights will be a success…and the CASM doesn’t need a huge load factor – they could probably survive 50% with the yields they’ll most likely get out of a market like Sonoma/Marin. Maybe SMART will help with the airport overall.
The airport lobbied for DEN but the aircraft weren’t available: http://sonomacountyairport.org/wp-content/uploads/aviationcommminutes_11-17-16.pdf
This route will also be the shortest in UA’s network, beating out ORD-MKE.
> This route will also be the shortest in UA’s network, beating out ORD-MKE.
Looks like it’s shorter by a mile. Wow.
Does that figure include “shortest in the network,” or “shortest by a mainline plane” (excluding regionals)? I know when I flew ORD-MKE the captain mentioned something about it being the shortest route, but I can’t remember if he was referring to just mainline metal or all flights operated by the airline.
Network
Southbay – Doesn’t matter if it’s 2 hours or 3 hours. There is absolutely nobody who is going to fly that locally.
Reminds me back in the mileage run days to get the minimum miles needed for an award when I worked at TWA. Would have people fly the BUR-LGB portion of the winter STL-BUR-LGB-BUR-STL flight….LOL
Wow, I never knew there was such a flight… MD-80?
I flew DCA-IAD once (and stayed on the plane, which was going to Calif.)…
Yes MD80.
Back in the day there used to be a lot of short hops between SFO/OAK, SFO/SJC, OAK/SJC going to/from other places on jets
CF – I know that. I was just surprised to see such an optimistic driving time between those cities. The SF Chronicle had an article that was shocked that UA would offer SFO – STS for $400 round trip. It’s obviously for connecting traffic. Same with ORD – CMI since that can be driven in less than 3 hours.
I don’t think anybody is arguing that their will be local traffic except the local media/ Press Democrat, who, even after the skeds were published, had a headline that read something to the effect that, according to Reddit, UA might start service to SFO.
It’s also interesting to mention that Jon Stout, the airport manager at STS, is lobbying AS for STS-BUR. Based on past history, since he managed a P2P like STS-SNA, I think it will happen. I think that after that, that will be all that STS will get for awhile. Interesting to point out is that G4 hasn’t taken Santa Rosa-Phoenix/Mesa off of their route map or out of their booking engine yet.
Air Wisconsin entered the CMI market in 1985 when it acquired Mississippi Valley Airlines (MVA). October 1986 marked systemwide branding as a United Express carrier and the airline withdrew ORD-CMI on February 10, 1993.
United Express more recently served BMI (2000-2008), but CMI makes more sense due to greater distance from existing stations PIA and SPI.
First, United started fixing its excessive use of costly 50 seat regional jets several years ago by acquiring more 2 class RJs and using mainline aircraft in longer and more highly competitive markets.
Second, Denver has improved for UA because Frontier is much less of a factor than it was years ago and fares have risen in what was once a 3-way battle between UA, WN and F9. UA used small RJs in many long markets from DEN so the profitability was bound to improve when lower unit cost larger aircraft began to replace 50 seat RJs.
With all due respect, US built Charlotte into the hub it is by looking at what Delta has at Atlanta and duplicating the hubs and connecting traffic. Even though CLT has a smaller local market, the airport has low operating costs and for now, has relatively little low cost carrier presence which helps keep fares in the local market high. CLT still has one of the highest percentages of connecting traffic of any major US hub and AA@CLT is heavily dependent on regional carriers; the cost of regional carrier capacity is increasing due to the pilot shortage so simply adding regional carrier flights might not be a viable long-term means to build a hub for either AA in CLT or UA at ORD. Further, AA is finally rationalizing a lot of internal network competition between its multiple east coast hubs so UA’s addition of new capacity at ORD is, in some cases simply adding back industry capacity.
Most of United’s additional capacity is on 50 seat regional jets or with mainline aircraft operating on overnight flights, which does not generally result in the best revenue esp. in off-peak periods. Red-eye flights work as part of a robust schedule that includes many daytime flights.
Finally, Kirby is hyper competitive and this list of market additions proves that a large part of UA’s strategy involves adding flights in markets where other carriers are stronger, something UA has already done in large ORD markets. While UA might gain more share in some of those markets, other carriers are adding capacity in UA’s major markets. It is far from clear that UA won’t just end up with lower share in more of its largest markets even if it gains a larger share in markets where other carriers are the dominant carrier.
I was going to mention something like your last point as well. I noticed that many of the larger markets being added just happen to be competitive routes with DL. I always thought it was strange that UA didn’t serve SFO-DTW…you’d really think they’d have enough local+connecting feed to give that a shot, rather than sending everyone to DL. Same with CVG.
Then, you also see that EWR is adding year-round SLC and getting more ATL and DTW. Again, the DL pattern. DL flies EWR-SLC non-stop all year (granted, just once daily). UA must feel like they can take advantage of connections and stop giving away passengers to DL.
I can’t think of a lot of traveling experiences that are worse than flying united and making connections. From being stranded at Dulles all day waiting for my 1 hour flight to Richmond or Raleigh to their surly and mean staff. I just say no to United.
The paradigm shift between local and connecting traffic is a lot like what many carriers around the world have seen with respect to domestic vs international traffic. The popular notion (especially in the US) was that domestic routes were a loss leader to feed the international moneymakers. Now in the post-consolidation, Open Skies, 787/350 world, it’s international routes facing the fiercest competition while domestic markets have become the cash cows for carriers lucky enough to have large ones. At least in places like the US, Australia, and to a lesser extent Japan.
Unrelated, I had read somewhere that after a certain point, there’s a correlation between route length and competitive pressure and that UA was at a structural revenue disadvantage since their average domestic route lengths were longer than AA or DL. It’ll be interesting to see if these moves put that theory to the test.
Thanks for the news on United’s plans for expansion of service. Too bad there is nothing there indicating any sort of improvement of service to Monterey, my local airport. Ever since the daily flight to Denver was dropped, getting back east in a timely manner has become a bit more of a challenge. Maybe something will be forthcoming somewhere down the line. I much enjoy your blog – thanks for the great detail and service.
Patrick Dunne
With the similar demographics and similar distance from SFO it is strange that STS gets 3x daily when MRY has only has 2x daily, although this is the winter season and maybe STS will go down to 2x daily after September.
MRY also has flights to LAX.
Good point
Excited to see 3 new UA flights to/from COU, my hometown airport. AA is the only airline serving our airport now and the flights are always packed. They obviously studied the load factors and realized there is room for some competition.
Here is my question. Why doesn’t United have any flights out of MDW? They were actually founded there. I know they have been in and out in the past. They could, I believe, profitably fly to/from MDW to their other main hubs, and then offer international service from there. I live in Chicago and don’t understand why United doesn’t have at least some flights out of MDW. I sometimes see United charters at the airport. By the way, Southwest RULES at that airport, and that is the airline I fly since I prefer that airport. Any help?
The “TED” subsidiary tried it around ten years ago but didn’t last. I’m no means an expert on UA but I would assume that it not only cannabalized/competed too much with their existing ORD routes but also MDW is not a premium pax market (which is why WN dominates it) which airlines like UA survive on. Plus there aren’t but one or two available gates at MDW….at least there weren’t when F9 was doing the ILG/TTN routes. Others will know much more than I.
you are right and the two reasons for UA and AA not flying to MDW are related. WN loves to operate its hubs/focus cities from airports it can dominate and where there are few gates for competitors. No carrier wants to start service in another carrier’s strength market and not be able to defend it if necessary. WN has plenty of gates it can use to run off a competitor.
DL has been the most aggressive at serving WN hubs and it does so by serving at most its largest hubs. In the process, Delta becomes the largest carrier from the city as a whole to the hubs it serves from both airports… ie Delta serves both Chicago airports to ATL, DTW and MSP and is the largest carrier from Chicago to those cities and also gets a revenue premium because of it. Note also that Delta doesn’t have a hub in Chicago or Dallas or Houston where it uses the same strategy while AA and UA do have hubs in those cities. DL’s largest hubs are in single airport cities unlike AA and UA.
Very informative observations. Thanks much.
Charcoal22 – Because if you live in Chicago and like United, you’re going to fly out of O’Hare. As others mentioned, United has tried it on occasion. It did Ted service there most recently, but it’s been awhile.
The reality is that it’s a money-loser for them. The business travelers who like United are going to drive to O’Hare and fly nonstop wherever they’re going. The Midway service ends up being low dollar stuff.
Please read
Good to see the daytime PDX-EWR restored and daily too. It was (I think) 5x weekly until early 2015 when it was pulled.
That is a huge advantage for us UA fliers going to India as it gives us a one stop to DEL/BOM, arriving at 9pm-ish, as opposed to flying Lufthansa and arriving at stupid o’clock in the early morning. Otherwise the Delta partnership with Jet Airways would have swept up the PDX-DEL/BOM market via AMS, as well as TLV** and connections to the tier 2 European cities that United serves from EWR.
** – yes we can get to TLV via SFO but EWR gives greater choice.
Another addition that I think is interesting: DEN-KOA goes from Saturday only to daily this summer on a pmCO 752. If my memory is correct, I don’t think UA has ever flown anything DEN-Hawaii daily except for DEN-HNL. But doing DEN-KOA daily definitely gives a broader reach into the interior of the mainland better than the traditional Hawaii gateways (SFO/LAX) do. Interesting to see how this enables things like COU-KOA to be done with only one stop.
Funny you should mention that. I am looking to book a COU-OGG or COU-KOA this winter :)
this O&D seems like a marginal move. how much margin? imo, very little.
Just wondering how much of a waste (cost, use of air space, etc.) our national air system is by what you are talking about here. State College is a good example–traffic is split up 4 ways–SkyWest/Delta via DTW; Piedmont/AA via PHL; CommutAir/UA via IAD; and ExpressJet/UA via ORD. Why should there be 4 operators out one place like this, and the same for every one of the many State Colleges across this faif land?
Do people decide/particularly care what/how to fly from the “State Colleges” of the country–the spoke operator and its reputation and its aircraft/seats, distance to a hub, the hub itself, flight frequency, the marketing airline, the total travel experience, or simply total price?
I really don’t know, but having 4 “DBAs”, 4 hubs, to get the State College traffic seems such a waste for the national air traffic system. Isn’t there a simpler system of one carrier from State College to one, maybe two natural airport connecters, like PIT, maybe PHL, where every,EVERY US airline would compete/do its own pricing d(State College-to connector point-to destination) which would increase competition and dramatically drop the fares from State College to everyplace? It seems to me, the current system simply leads to inefficiency and higher fares.
that’s EXACTLY what deregulation was intended to end.
I can assure you that four operators at a small airport don’t matter in the scope of the national ATC system.
more importantly, four operators add competition which many small cities want.
Jaybru – I don’t think I’m following this at all. There are three different marketing carriers that fly to State College (the airline that actually operates it is pretty unimportant). Are you saying that people don’t care about competition in these markets and they should just be forced to go through one hub because of airspace issues? I don’t see that argument at all.
To be honest, I am looking down the road as to where the 3-legacy-hub and spoke system is taking us. I see fewer and fewer hubs, longer and longer spokes, which I question whether this will benefit the local communities as State College, service-wise, price-wise. The State Colleges are never(?) going to be able to support point-to-point service, but as hubs contract and spokes lengthen, can the locals ever expect to see a level of competition producing better service, better fares unless things change?
Of course, things change, and there are smarter people than me out there. Back in my time, getting local service Allegheny to put service up there into Phillisburg for the cow college seemed like an impossibility! But…!
I can’t say I follow. If only one carrier flies from State College to a “connector”, then chances are that they’d be the only hub and spoke airline that flies to other spokes from there, like AA at PHL. So you’d really have pretty much no competition for connecting traffic unless you expect people to ticket-split. Or are you talking about dismantling the hub and spoke system and just have everyone compete on point-to-point routes?
Not really. There’s a certain amount of daily demand out of a place like State College – what does it matter if those seats are spread over 4 carriers or 1 if it’s the same total number of seats? In fact, 4 is better because there is price competition!
Back in 2003, I flew from CHO to CMI for a reunion (Go Illini!). CHO-CVG-CLE-DTW-CMI. On the return, CMI-DTW-CVG-CHO. The other option was connecting at IAD and ORD, but that was not practical either. Now, there are more one-stop options, like via ORD or (CVG when Delta had it) Delta used to the only carrier for it, but now, American. Soon, United. Too bad it is unlikely that I will fly that route ever again.
Since I reside in Springfield MO, the new flights from SGF to IAH are very important to me. AA has been adding flights at SGF while UA had cut back until now.
Kirby addressed in an employee letter why LAX is not getting any growth.
We are not adding any additional flying at LAX at this time because we need more gates before we can embark on a growth journey there. We are committed to LAX for the long game – exemplified through our recently remodeled and reopened United Club — and we’re working hard to secure additional facilities that will give us the ability to compete vigorously. I believe in the future of LAX and its critical role as one of our hubs, and, as we are able to get commitments for more gates, I’m looking forward to growth in LAX.