ULA’s Customer Crisis (and Other Lessons)

ULA’s Customer Crisis (and Other Lessons)

This is the second part of a follow-up gap analysis of the United Launch Alliance. For readers interested in the first part, please go here

As promised in the first part, this article seeks to uncover lessons from the two gap analyses. It starts with the most obvious and somewhat predictable change and moves on to others that are more subtle. The good news for the industry from the start is that no one can point to ULA as having a monopoly on the U.S. launch market anymore.

The Unstoppable Monster: the Competition

The first and most apparent change is SpaceX's rise and dominance in the U.S. launch industry since 2015. Back in 2015, SpaceX’s activities and performance provided some hints about this future. The company was energetic and hungry for business. It was developing reusability. How it built its rockets went against everything legacy rocket manufacturers held dear. 

At the time, companies like ULA might have deemed the things that excited the public and some industry observers about SpaceX as weaknesses. After all, how can a company provide government customers (such as the DoD) mission assurance about a rocket undergoing constant “tweaking?” The Department of Defense's (DoD’s) decisions seemed to align with the company’s thinking. It didn’t initially award National Security Space Launches to SpaceX and, when sued, granted shares of NSSL contracts that favored ULA in launches and money.

Also, SpaceX just wasn’t launching that often then. In 2015, SpaceX had launched six times. During the same year, ULA had launched double that number–12. However, by 2021, the tables had not just been turned but flipped over. SpaceX launched 31 times that year to ULA’s five. And 2023 just worsened the disparity between the two companies, as ULA launched three while SpaceX conducted 96 orbital launches.

So, yes, SpaceX is performing significantly more launches even in 2024 than it did in 2015. ULA hasn’t kept up, and its launch cadence has declined. Its lack of a sense of urgency might seem puzzling at first until its primary customer, the DoD, gets some observation. 

By 2021 (when I conducted the first gap analysis), ULA had painted itself into a corner and seemed happy with the results. Maybe part of that contentment was because the U.S. Department of Defense still trusted ULA’s messaging that it would launch Vulcan way before the National Security Space Launch agreements that specified it kicked in. Again, the DoD had trusted ULA enough to award a larger portion of NSSL launches to ULA despite Vulcan’s status as a ghost rocket (vs. SpaceX’s very real Falcon 9) at the time.

The result of the DoD’s favoritism was that ULA remained complacent despite SpaceX’s evident industry inroads and launch increases. The company expected its past performance and reputation with the DoD to protect its place in contract allotments. That complacency can even be seen in its rocket development. There is nothing groundbreaking about it. It is merely a replacement for its Atlas V and Delta IV rockets, which were reliable and capable but expensive. 

Ignoring Changes?

Most of the launch industry changes that have happened since 2021 aren’t really surprises. Companies such as Arianespace and ULA knew target customers would continue to value rockets with excellent upmass capability. Blue Origin’s New Glenn and SpaceX’s Starship both up the ante for launch vehicle upmass. Even newcomer Rocket Lab has decided to build Neutron's reusable rocket to lift much more to orbit than its Electron.

Looking at rideshare trends. The chart below shows annual rideshare launches using medium-class launch vehicles or higher from 2018 through now. 

The number of spacecraft deployed from rideshare rockets is non-trivial (usually in the hundreds). Aside from the pandemic dip in 2020 and its impact (my assumption) on 2021, rideshare missions using medium-class+ launch vehicles seem relatively steady each year. That steadiness may indicate a need for more rideshare launches–a possible opportunity for Vulcan. 

As noted in earlier analyses, if an orbital rocket (like Vulcan) can launch more mass, its owners can use it more flexibly. That’s why India’s PSLV or Russia’s Soyuz are used to deploy more satellites in one launch. When using a reusable rocket, rideshare makes more sense. Rideshare is also why smallsats are growing in size. When selecting rideshare, operators' satellite launch costs (even for a larger one) are less than if they had gone with a 3U or 6U cubesat and launched with Electron. 

Because a PSLV lifts more mass than an Electron, its customers aren’t as constrained by the launch system's larger upmass capabilities and can share the launch costs with others. It’s a beneficial business model. However, based on Vulcan’s design, it's clear that ULA doesn’t think its primary customer is interested in that capability (despite the NRO’s latest 20-21 satellite rideshare on a Falcon 9). How Vulcan will deploy Kuiper satellites will be interesting to see, though.

ULA’s Vulcan is generally a good option for launches but can’t be used as flexibly for a broader set of missions as its SpaceX counterpart. Its performance fulfills customer expectations with its upmass capability (19,000kg to LEO and 8,400kg to GTO). The company didn’t bother working on a smallsat-capable orbital rocket as it made no sense to its bottom line. It didn’t look at reusability because it understood that its most important customer, the DoD, wouldn’t need ULA’s launch services often enough to justify developing that. Especially if that pesky SpaceX continued to hang about to snatch table scraps (and more).

The number of launch competitors the market can support is still a question. I believe it can support more than two large launch service providers, but the current U.S. market using large rockets totally relies on the Falcon 9. ULA has launched twice in 2024 so far (May 28, 2024), while SpaceX has launched 56 times. Considering that ULA’s target customer, the DoD, has not just maintained but increased launch demand (through the SDA), ULA’s nearly MIA market status in 2024 is notable.

Trust–Who Do Ya?

Maybe the most significant observation between the two gap analyses is the DoD’s apparent withdrawal of trust in ULA’s ability to conduct launches reliably. In this case, “reliably” refers to the company’s ability to launch when the government needs it to.

It’s the whole “US Government has doubts about supporting ULA” thing. And that loss of trust is a BIG DEAL for ULA. As noted earlier, the company relied on the DoD’s trust to win larger launch contracts, extract more money per launch, etc.--despite not having launched Vulcan once. ULA’s first Vulcan launch was conducted behind schedule (which put DoD launches behind schedule). The company still seems to make uncertain statements surrounding future launch cadence, likely because it doesn’t know if its engine supplier can keep up.

Worse for the DoD is the possibility of Blue Origin acquiring ULA. Blue Origin has yet to launch any orbital rocket. It’s the primary source of ULA’s schedule problems. The company acquiring ULA combines the worst of all worlds for the DoD–a launch operator with no established orbital launch history; a rocket engine manufacturer that can’t seem to manufacture enough engines for Vulcan and New Glenn; a launch operator with launch systems of unknown reliability; and a space company whose attention is divided (when it really should be proving its launch chops–as everything else hinges on that).

The circumstances have forced the DoD to make unusually clear statements regarding ULA. It has even gone so far as to fine the company, which–I believe–is unprecedented. These are the customer’s shots across ULA’s bow, which might indicate the customer’s willingness to start allocating more (if not all) NSSL launches to SpaceX, at least until ULA can demonstrate an ability to launch Vulcan more than once a year.

If ULA’s primary customer shows signs of skepticism and realism, imagine how a company less beholden to ULA, such as Amazon Kuiper, might respond. Amazon has already contracted a few launches with SpaceX. If ULA just can’t launch that often, no one will blame Amazon for shifting the bulk of its contracts to SpaceX (because the alternative–Arianespace–is about as bad off). No one will blame other possible customers for writing off ULA as a launch option altogether.

Yes, there’s more that could be guessed at between the analyses. It’s obvious to me that I should work more on these types of analyses. Looking at how I placed observations in the first one versus what I observed in the second makes me think I still need to learn more. 

I can’t help but think how ULA’s current predicament might have been driven by shortsighted decisions made over a decade earlier. It didn’t develop an engine when given direction and money to do so over ten years ago. The company, the DoD, and even Congress understood the consequences of using a Russian engine for National Security launches. 

Despite that knowledge, ULA influenced the DoD to continue using rockets with Russian engines. It convinced the DoD that Blue Origin’s non-operational BE-4 would be an excellent engine for Vulcan. To be clear, the DoD and ULA will always talk about how they make great partners for each other. However, the most recent activities indicate that one of those partners is questioning the other a bit more pointedly.

Ultimately, the latest gap analysis suggests that ULA is in a different place than it wants to be and is on a path that might lead away from its goal, missing market opportunities along the way. Its primary customer seems increasingly concerned about ULA's ability to get where it needs to be. Its launch activity declines, inability to field a fully operational Vulcan, and possible acquisition are causing trust challenges for its customers. That is the newest change for ULA and not a positive one.

If you liked this analysis (or any others from Ill-Defined Space), any donations are appreciated. For the subscribers who have donated—THANK YOU!!

I'm a Giver. Let me donate!