Orbital Sciences: You probably shouldn’t outsource rocket science if you’re selling rockets

Choosing to outsource part of your business is a strategic decision – but how much is too much outsourcing? From its beginning, Orbital Sciences has sought to combine its mastery of systems engineering with extreme “make vs. buy” decisions to remain competitive in the space industry.

In February this year, Orbital Sciences Corporation merged with industry peer ATK to form Orbital ATK. The deal was hailed as a win-win for both as “ATK looks to bolster its aerospace business and Orbital Sciences hopes to boost the scale of its existing operations as well as gain a foothold in the defense sector”1. More so, the merger will more than likely help realign Orbital’s business with its roots and save it from becoming irrelevant.

The Beginnings…

Orbital Sciences was founded in 1982 by three Harvard Business School classmates – David W. Thompson, Bruce Ferguson and Scott Webster1. Its business was and continues to be developing, building and selling aerospace systems to the government. They secured their first contract in 1985, with NASA, for delivery of a several Transfer Orbital Stage (TOS) vehicles3. The low-cost TOS was designed to sit inside the Space Shuttle Payload Bay where it would be removed in space by astronauts and then launched with a payload (see images below).

mainLeft: TOS being erected from Shuttle Payload Bay4; Right: TOS after jettison from Shuttle Payload Bay5

Orbital only sold two TOS systems to the government. Success or failure of TOS project aside, the impact that first endeavor had on the foundations of Orbital’s operating model in the decades to follow was profound. TOS was largely developed out-of-house by Martin Marietta6, whom had at the time just successfully emerged from a bloody hostile takeover battle and was looking for growth7. The Orbital value-add in the creation of TOS would largely be systems engineering – the Orbital bread and butter. Specifically, Orbital would take the pieces developed by Martin Marietta and other subsystem vendors and ensure, when all assembled, the sum of the parts worked.

Fast Forward…

After TOS, Orbital Sciences would go on to produce satellites and Pegasus, its first rocket (which used rocket motors designed by Hercules Aerospace, now ATK)8, and go public on the NASDAQ stock exchange9. This rocket would be the basis of several more to come – the latest of which won a $1.9B contract to ferry supplies to the International Space Station10. The Orbital Sciences rocket family, mainly small- and medium-lift rockets, would operate virtually competition-free for two decades.

main2Above: Orbital Profit Margin vs. other aerospace companies11

Even with its space and launch units comprising most of the firm’s revenue, it achieved margins similar to more vertically integrated and better diversified firms like ATK, Boeing and Lockheed Martin (see chart above). The Orbital Sciences model – leveraging its mastery of systems engineering with extreme “make vs. buy” decisions – was a sure winner.


Orbital Sciences leveraged its ability in systems engineering successfully for quite a while. In the early 2000’s, NASA funded an effort to develop a new rocket system which would ferry supplies to the International Space Station10. Orbital Sciences undoubtedly saw this as an opportunity to expand its foothold in the launch industry. They would build off the heritage from previous rockets but utilize a liquid rocket engine – something they had only limited experience with13. This liquid-fuel based engine, named RD-180, would be sourced from the Ukraine14. During the Cold War, Soviet Russia produced hundreds of these engines but never used them. The plan was that Orbital Sciences would buy these left-over imported engines from a U.S. subsidiary and integrated them into their rocket – all at a cost much less than its closest domestic capability15. Orbital Sciences executed this approach and flew three rockets to the ISS. On its fourth flight to the ISS, the rocket blew up about two hundred feet off the launch pad.

An extensive investigation by both Orbital and NASA would later show that the liquid engine failed and the root cause was either a design or manufacturing flaw17. Orbital Sciences, with a contract obligation to deliver supplies to the ISS on a timetable, would be forced to buy rockets from its much more expensive competitor ULA at retail prices. Details of the order with ULA were never made public, but Orbital Sciences 2015 profits would decrease by over $138M (-41%) year-over-year12.

“Why do we fall? So we can learn to pick ourselves up.”

As it worked to reestablish its capability and continue to serve its contract with NASA, Orbital Sciences would need to find and integrate an alternative engine. Again, they chose to source this new engine from the same Ukrainian manufacture18. Many questioned this strategy given the risk the same issues may arise and considerable geopolitical pressure associated with the Russian-Ukraine conflict – including ATK whom Orbital was likely in early discussions with at the time regarding the merger19. The merger earlier this year with U.S.-based ATK may be a strategic move back to a more sustainable mode. ATK for over four decades has been a world leader in the design and manufacturing of solid rocket motors. Given Orbital Sciences extensive experience integrating solid motors, the merger makes sense and might provide a good footing for the company’s rocket business going into the future. Owning the the capability in-house is surely a better arrangement for Orbital, and their intentions to shift to more in-house R&D12 will likely come to represent more of the operating model in the years to come.

Can they adapt to a model focused on more R&D so they compete in a marketplace they’ve historical not had to compete in? If they do, will this new model give them the edge to compete with the likes of companies like SpaceX who, unlike Orbital, sell to both government and private entities? Only time will tell.


1 ATK to merge with Orbital Sciences in $5 billion deal, spin off sports division, Washington Post (web link)

2 http://www.fundinguniverse.com/company-histories/orbital-sciences-corporation-history/

3 https://en.wikipedia.org/wiki/Orbital_Sciences_Corporation

4 https://en.wikipedia.org/wiki/STS-51#/media/File:ACTS_deployment_from_STS-51.jpg

5 https://en.wikipedia.org/wiki/Transfer_Orbit_Stage#/media/File:ACTS_Deploy_3.jpg

6 http://www.nytimes.com/1992/09/26/us/us-launches-a-spacecraft-on-a-mars-trip.html

7 https://en.wikipedia.org/wiki/Martin_Marietta

8 https://www.orbitalatk.com/news-room/feature-stories/Pegasus-25th/default.aspx

9 https://www.crunchbase.com/ipo/82cb0f6b018a605b933fd19b8cb57762

10 http://spaceflightnow.com/antares/orb1/140109launch/#.VmYcbuLKPVI

11 Data source for development of graphic: amigobulls.com

12 Company 2015 10-K

13 http://www.parabolicarc.com/2013/08/19/closer-orbital-sciences-stable-launch-vehicles/

14 https://en.wikipedia.org/wiki/RD-180

15 http://survincity.com/2011/06/in-the-u-s-the-struggle-for-the-rocket-engine-from-2/

16 video source: https://www.youtube.com/watch?v=3jCystkiIBs

17 http://spacenews.com/42502turbopump-in-aj-26-engine-implicated-in-antares-failure/

18 http://spacenews.com/orbital-sciences-orders-rd-181-engines-for-antares-rocket/

19 http://www.spaceflightinsider.com/organizations/ula/atk-offers-solid-rocket-alternative-russian-made-rd-180/

20 featured image: http://blog.flickr.net/2015/11/04/nasa-releases-the-images-of-the-antares-rocket-launch-failure/


Off-Grid Electric’s Mobile Power


Virtue in Vice Media – The “Time Warner of the streets” and its plan to occupy…

1 thought on “Orbital Sciences: You probably shouldn’t outsource rocket science if you’re selling rockets

  1. Couldn’t agree more with this write up – great post! We’re seeing this in the submarine industry as well, with countries like India buying submarines from Russia. I can’t imagine what it is like to use a complex piece of equipment such as a rocket that was designed in a country that doesn’t exist anymore (USSR) with a different language/design philosophy/system of measurement. But, given the fact that these engines are laying around, are there any other uses for them, if we don’t trust them for actual launch use? Could Orbital Sciences keep the same operating model and just shift their line of business by finding creative uses for the engines? (Education/training/testing parts/testing fuels…)

    Also, I had no idea about the Orbital Sciences/HBS connection. Very interesting!

Leave a comment