Quicktake: Vertical integration is an overused term (Space industry case)
In line with SpaceX market dominance, vertical integration became to be used as a magic keyword for space industry companies - I'd argue it is not.
Here is my Thursday take on vertical integration in the space industry: In past years it has become a very popular term but imo a bit overused - many companies would be better off putting vertical integration plans on hold.
I think the baseline for the discussion stems from the fact that SpaceX became somewhat famous for its degree of vertical integration and industry commentators have begun to put it in a direct correlation with SpaceX’s success.
Yet, it might be a useful reminder that SpaceX relies on/works with third parties on at least four different levels.
On the component level, we know companies like Filtronic and formerly Pioneer Aerospace (more on that later) supplying to SpaceX. They just don’t shout it out too loud because SpaceX is not a particularly great customer - with pushing down margins to the floor, reliability, schedule requirements. Industry veterans (not me!) also have a lot of stories of Musk showing up at the Space Tech Expo Long Beach and searching for component suppliers there.
On the Starlink side, it is much more interesting
SpaceX works with satellite operators like SES to deliver a joint service [1].
SpaceX works with managed service providers like Marlink, Speedcast, Elcome, KVH to act as channel partners for its service [2]
SpaceX also works with retailers like Costco, BestBuy, Home Depot to sell their terminals, pretty much in direct opposition to what the direct-to-consumer vertically integrated approach would dictate.
In my understanding, this shows that a) SpaceX is broadly willing to work with third parties if they are able to adhere to their cost, reliability and schedule expectations and b) while for the aerospace component chain, it made sense to mostly vertically integrate due to the industry structure before, for the Starlink product/solution sales it suddenly make sense to work with third parties with SpaceX opportunistically does.
That sentiment is largely seconded in this hilarious quote by Abhishek Tripathi for TechCrunch [3]
“While SpaceX is famously known for insourcing components, Tripathi said he’s been in meetings with CEO Elon Musk who determines when to outsource based on two factors: that the supplier is not “a complete, incompetent idiot” (Tripathi was paraphrasing Musk here), and that SpaceX can trust that the supplier can deliver on schedule.”
So my point is that people overstate the benefits of vertical integration and even the pronounced leader in the field actually relies on third parties when it opportunistically fits. The main problem with vertical integration frankly is that it is expensive. For venture-backed companies, the goal is to a) demonstrate whatever minimum viable product with as little capital expenditure as possible and b) use access to capital to gain market dominance again while trying to limit the amount of cash used. And of course - acquiring other companies goes exactly against the doctrine.
Space domain is still scarce of successful exits. Interestingly, one of them, Apollo Fusion, very much relied on production capabilities intentionally not in-house to keep the process lean. See the thread [4].
In the end, I can think of 3 different situations where vertical integration is useful
In market where overall market expansion is not possible and eliminating margins in the value chain is the only or cheap way to improve financial outcomes
If the first targeted market is perceived as small, acquiring companies can help expand the TAM for investors “we don’t do only OTVs, but also Moon Landers and also rockets and propulsion”, because the market for all four is larger than any one in particular.
Because of cultural reasons - like strict push for fast-paced delivery or when simply such a component chain is not available.
That of course doesn’t say that targeted acquisitions, (like Nuview acquiring Asterea) are not a good idea - it’s just… let’s call them for what they are (strategic acquisitions - then discuss whether they are in fact strategic/helpful or not) instead of treating “vertical integration” as a magic kool-aid.
[1] https://spacenews.com/starlink-and-ses-join-forces-for-multi-orbit-cruise-connectivity/
[2] https://spacenews.com/starlink-may-account-for-up-to-40-of-spacexs-2023-revenues/