Rockets of Cooperation

Vertical Integration

High-profile, hyper-vertically integrated firms such as SpaceX, Tesla, Apple and Peloton have revived conversations about strategic synergy, after decades of commerce and management scholars pooh-poohing the practice. Bloat, overcomplexity, loss of employee morale and ultimately unprofitability were supposed to be the great dangers of straying too far from ‘core competencies.’ How have these firms combined organizational units inside the enterprise and beyond without falling into the old traps?

Millions around the world find themselves glued to their screens whenever SpaceX, the first private company to travel to space, has its latest launch. We are captivated by the gob-smacking achievement of figuring out how to land a rocket back on Earth instead of it splashing into the ocean, the occasional spectacles of explosions upon launch or landing and the heartwarming return of America launching astronauts to space a decade after the mothballing of the Space Shuttle program.

But the real breakthrough by the firm, the one that doesn’t attract all the likes and livestream eyeballs or even headlines, but which is what allows them to achieve all this, is SpaceX’s radical business model innovation that upended decades of received wisdom regarding corporate synergy, in particular vertical integration.

Since the 1980s, commerce and management scholars have frequently warned of the significant risks involved with vertical integration and of overdoing ‘corporate synergy.’ The threat of lowered employee morale and runaway investment costs led decision-makers to embrace the mantra of the experts that they rigidly stick to their ‘core competencies’ and outsource the rest.

Yet it is precisely through in-house production of most of its components that SpaceX has been able to radically undercut its main competitor, the Boeing-Lockheed Martin joint partnership United Space Alliance. Each of the 1,200 suppliers in the latter’s outsourced supply chain, spread out across the US, adds their own profit margin to the manufacture of a component, inflating the cost per launch to USD 460 million. Spurning such conventional business wisdom, SpaceX manufactures in excess of 70% of its main products in-house, allowing it to charge its clients just USD 62 million per launch. And even more eye-popping, according to CEO Elon Musk, the firm’s marginal cost for launch of a re-used Falcon 9 booster is only USD 15 million.

As SpaceX steadily perfects reusability and rockets are reused dozens of times, this figure will only fall, as the more times a rocket is used, the more launch costs decline. The firm is not just integrating ‘backward’ (or ‘upstream’) toward what would be its own suppliers, and ‘forward’ (or ‘downstream’) toward its customers, but is even tossing out the pretty uncontroversial business maxim that one at least does not compete with one’s customers. SpaceX’s Starlink project aims to deploy some 42 thousand satellites to provide high-speed internet access anywhere in the world – going up against the satellite operators who purchase its services to get their devices into low-earth orbit 1 2.

Although it is true that raw materials arrive at the SpaceX factory and rockets come out the other end, one of the only parts of the supply chain SpaceX doesn’t appear to have a finger in is the actual extraction of raw materials. But that may not be true for long for sister company Tesla, which has likewise embraced hyper-integration of its supply chain. In a 2020 letter to investors, CEO Elon Musk described his radical ‘in-sourcing’ strategy as key to its rapid growth. A great deal of the components they need simply don’t exist. ‘So we made the machine that made the machine that made the machine,’ Musk bombastically declares, aware that he’s tearing up the business rule book. When Tesla announced that it was to set up its own ‘gigafactory’ in Nevada to build its own batteries, the move was ridiculed as an obvious business unit that should be outsourced. One company can’t specialize in everything. Let the people who really know batteries make Tesla’s batteries 3 4.

Such sneers have since disappeared entirely as the firm achieves substantial profit margins on its luxury vehicles sought after by wealthy consumers regardless of their personal environmental ethics, while most auto manufacturers struggle to even make a profit on electric vehicles at all.

Were the business scholars of decades past wrong? Is there nothing to learn any more with respect to their anxiety about complexity, employee morale and costs? Or are high-tech businesses like SpaceX, Tesla and Apple special cases, and the old lessons still hold for most other enterprises? Or is the story more nuanced than any of this?

To answer these questions, we need to return to the time of New Wave music, shoulder pads and corporate raiders. That’s right, we’re going back to the 80s.

In 1985, the head of car manufacturer General Motors’ European Opel division, cost-cutting evangelist Jose Ignacio Lopez, revolutionized the sector with his ‘Global Sourcing’ strategy. This strategy in effect stripped operations down to GM’s main areas of specialization – which turned out to be not manufacturing cars after all, but instead car design, production of essential parts and marketing – and outsourcing the rest. At a time of weakened trade unions, management convinced many groups of workers to spin out their expertise into their own small companies, which could theoretically supply the larger enterprise with parts. Entranced with the idea of being their own boss and promises of substantial contracts, many found the idea attractive.

Meanwhile, GM could pick and choose amongst the suppliers, which were now set against each other. The competition between them could drive down cost much more easily than the whip-hand of any manager with respect to an internal team locked into a contract. The firm’s successes would steadily be emulated by the rest of the industry.

Easing the transition, China’s workforce was in the process of opening up to the world. And only four years later, the Soviet bloc fell, delivering another billion, often high-skilled workers to the global workforce. Labor costs cratered and the industry enjoyed a period of spectacular growth worldwide.

In addition to these carrots, there was also the stick of corporate raiders in this period – made famous by Michael Douglas’s character Gordon Gekko in that ur-80s film, Wall Street – buying up sufficient stock of firms with apparently undervalued and ‘over-integrated’ assets to force restructuring, downsizing or outsourcing in order to goose stock value.

These real-world results were only reinforced by data-driven commerce and managerial science research that would develop a theory of why this worked.

A 1983 paper by commerce scholar Robert Buzzell in the Harvard Business Review interrogating whether vertical integration was profitable was seminal in this regard, and it may even have encouraged Ignacio Lopez’s restructuring strategy. One has to remember that the 80s was the first time that fast and relatively cheap search and analysis of databases was made feasible by computerization.

What made Buzzell’s name was his research team’s development of the Profit Impact of Marketing Strategies (PIMS) database, in which various aspects of business strategy were entered in order to quantitatively identify and measure what drove profits. It was a game-changer compared to decades’ worth of ‘just-so stories’ of corporate storytelling about ‘how to succeed in business’ 5.

At the time, there was a debate over whether vertical integration was crucial to a business’s survival or whether excessive integration was what was responsible for a series of high-profile corporate failures. Chemicals giant Dupont had not long before purchased oil producer Conoco for a then record-breaking USD 7.3 billion. It was supposed to deliver to Dupont ‘a captive hydrocarbon feedstock source’ and ‘reduce the exposure of the combined companies to fluctuations in the price of energy and hydrocarbons.’ Troubled from the start, the deeply unprofitable merger would be unwound 17 years later, with analysts concluding that Dupont would have made more money simply investing the purchase price in the stock market.

In the 70s, a series of integrated circuits firms and consumer electronics manufacturers vertically integrated forward and back into each other’s markets. The most famous of which was Texas Instruments’ branching out into watches, calculators and, famous to children of the 80s, the Speak n’ Spell speech synthesizer. But the then-president of Commodore, another producer of calculators and personal computers, argued that integration was a disastrous strategy. It was worth swallowing the increased cost of externally produced microchips to be agile enough to ‘get into and out of a technology when you want to’ 6.

Who was right? Buzzell could now dig deep into thousands of businesses and compare. What he found was not always encouraging for the partisans of synergy and integration.

He concluded that there was a ‘V-shaped’ relationship to integration: either a very low level or very high level of integration yielded above average returns on investment, but middling levels of integration produced the lowest earnings.

The idea of synergy, of which vertical integration is just one expression, is that the combination of or cooperation between two or more entities adds up to an output that is more than the sum of the parts. Duplication is eliminated. The cost of transactions between two parties is likewise removed, or at least reduced (there is, for example, no need for Unit A of a company to market to Unit B, eliminating those substantial associated costs). Supply of critical materials for a certain stage of production is assured because they are already in house. For the same reason, scheduling between stages is better coordinated.

But there are downsides, too. To set up or purchase new units upstream or downstream in the supply chain requires capital. The PIMS database showed that unless the cost savings of integration substantially outweighed the investment requirements, then the strategy would be counterproductive.

Unbalanced throughput added to the problem. In order to achieve costs cheaper than an otherwise outsourced competitor, production of a particular product within the supply chain may need to be at a very high volume. Thus, the firm may be producing that intermediate product at a much greater and more wasteful scale than required by a subsequent stage of production. To make this stage worthwhile, there may need to be consumers of that stage’s product outside the firm, returning once thought eliminated transaction costs to the process.

Related to this problem is reduced flexibility. If a stage of production brought in-house becomes obsolete, the firm is burdened with albatross of the materiel and the labor associated with that stage. If outsourced instead, then as soon as obsolescence is identified, the supplier can be dropped. Buzzell gave the example of a women’s clothing producer that had bought a double-knit textile mill and which was stuck with an unprofitable stage of production when double-knits went out of fashion.

But the gravest danger was over-complexity and loss of specialization – two sides of the same coin.

There are wildly differing labor practices and thus distinct managerial approaches at various stages of production or distribution. The transition from direct ownership of gas stations by oil companies to a franchise relationship with gas stations was prompted by the recognition that the managerial approach to manufacturing was very different to that required in the wholesaling or retail space.

More importantly, or so the argument went, it is very hard for executives and managers to be specialists in multiple different sectors. One of the key social transformations permitting the industrial revolution was the transition from craft manufacturing, in which a single artisan was responsible for all stages of production, to a break-down of stages of production, a division of labor allowing a worker to become very, very good at one particular task. The principle supposedly held for businesses, too. Greater complexity derived from having to be good at many tasks was no less inefficient. Instead of spending capital on developing or acquiring other units, that capital could be spent on getting better and better at that core competency task 7.

By the end of the decade, these sorts of findings were translated, in bestselling business strategy books like Competing for The Future by C. K. Prahalad and Gary Hamel, into the maxim that market leaders achieve their success by being able to identify and stick to these core competencies. This in turn permits a constant reinvestment in those specializations, allowing them to extend their lead over the competition still further.

Fixing the Synergy Biases

But this was something of a misreading of what the data had said. Buzzell had never concluded that vertical integration was not profitable. Instead, he argued that it simply depended on the circumstances of the particular firm. Sometimes it was profitable, sometimes it wasn’t. He had found the type of businesses and sectors, and at which stage of their development, where integration made the most sense. He also briefly mentioned some alternative forms of strategic synergy that offered many of the benefits of integration while avoiding some of the pitfalls, such as long-term contracts with suppliers.

And integration never really went away, anyway. Mergers and acquisitions, whether horizontally (expanding into or taking over competitors at the same position in the supply chain rather than moving upstream or downstream) or vertically continued right up to and beyond the boundaries of anti-trust law.

As early as 1998, another seminal piece of analysis regarding corporate synergy, this time from corporate synergy researchers Michael Goold and Andrew Campbell, detailed their findings after many years of studying the subject. Every business, even every enterprise has units, tasks or individuals that are specialized, and thus a need to knit them together. And so both the benefits and challenges of navigating such complexity are not something only vertically integrated firms experience, but are common to all endeavors.

The paper, titled ‘Desperately Seeking Synergy’ (another 80s reference, this time to the 1985 rom-com Desperately Seeking Susan), defined the term in business usage as the ability of two or more units or companies to attain greater value by working together than they could working apart. The authors noted how frequently managers’ initial keen pursuit of synergy fell short of expectations, with quick bursts of energy before petering out. They found that synergy drives distracted managers’ attention from the nuts and bolts of their businesses – the famous core competencies. In some cases, the synergy programs actually backfired, undermining employee morale, eroding customer relationships or damaging their brand 8 . And Goold and Campbell were supporters of synergy efforts!

What they found in their investigations was that it was a mistake to just assume that by bringing people, units or companies together, synergistic benefits would naturally materialize by dint of increased cooperation. Instead, synergy required rigorous, even skeptical evaluation at every stage.

Goold and Campbell had found that that synergy tended to come in six flavors: shared know-how, shared resources, coordinated strategies, pooled negotiating power, combined business creation and the aforementioned vertical integration.

Shared know-how is perhaps the most obvious form cooperation can take: units often benefit from learning the knowledge or skills about a process, function or geographic area possessed by other units. Why should each unit reinvent the wheel?

Units can save money by sharing physical assets or other resources. From a shared photocopier to a shared research lab or manufacturing facility, avoiding duplication is one of the main sources of economies of scale for an enterprise.

A similar logic lies behind the pooling of negotiating power when different units combine their purchases. They can gain greater leverage over suppliers and thus reduce costs or even improve the quality of goods they can buy. A university buying printer paper for all departments will enjoy dramatic cost savings over each department purchasing paper themselves.

Coordinated strategies involve aligning plans of action across different units such as a common branding across different markets. This, in principle, is a powerful way to reduce interunit competition, counter competitive threats and again to avoid duplication and slash costs. But in reality, they can be difficult to achieve and are not uncommonly a source of morale deterioration as business-unit autonomy is seemingly undermined.

Going beyond shared approaches to what already exists, combined business creation, either by bringing two groups together into a new unit or establishing joint ventures or alliances, are about the synergies involved in creating something new. And vertical integration we have already explored.

Goold and Campbell noted four managerial biases common to them all. They placed special emphasis on the role of executives, because they found that executives frequently blamed employee recalcitrance or incompetence for the failure of synergy efforts. They discovered that instead it was more often the fault of executives who suffered from one or more of four types of bias: what they called the synergy bias, the parenting bias, the skills bias and the upside bias.

The first, synergy bias, involves an overestimation by executives of the benefits of synergy and an underestimation of its costs. It reflects, the authors believe, an executive’s anxiety that if they do not promote coordination, standardization and other links, that there is no role for them. That’s what they do. This can lead to poor decisions or investments.

The example they give is of an unnamed global food manufacturer that enjoyed a presence in many countries, but whose national units operated entirely autonomously, barely engaged in any cross-border sharing of know-how, assets or purchasing power. A successful UK biscuit brand was disastrously launched in the US, land of cookies rather than biscuits, at considerable expense. Despite the popular belief that these are two different words for the same thing, the brand failed to acknowledge that the two are actually quite different commodities.

Likewise, a German pasta rolled out in Italy and Spain flopped, as did attempts to standardize ingredients across Europe in order to achieve economies of scale in purchasing and manufacturing. Consumers had become accustomed to a particular flavor and recoiled from the new common-denominator reformulation. And because all this had been directed by a new class of manager tasked specifically with these high-profile synergy wheezes without any local assessment of their viability, rather than encouraging inter-group cooperation, it actively hampered what cooperation there was, as national units became ever more dug in and convinced of the uniqueness of their local markets.

Related to this is the parenting bias. On the assumption that unit managers and workers in a unit are inherently resistant to cooperation, executives decide that the only way to achieve synergy is through, as Goold and Campbell put it, ‘intervention of the parent.’ The parent in this case being the executive, the corporate HQ, the head of a division or any other entities that oversee more than a single business unit.

With the above case, morale collapsed as national managers and workers felt undermined, overlooked and presumed to have no expertise. The argument is not that executives should never actively encourage synergy but rather that they should work with their subordinates to identify where these might be. They probably have a better idea as they know their unit best of anyone.

The skills bias arises from an executive who, upon deciding a synergy strategy is necessary, assumes that they are the one to carry out the strategy. But synergy has its own set of skills. The given executive may not have strengths in this area. They may lack the operating knowledge, personal relationships, skills in facilitation, patience or character needed to achieve the desired cooperation. They may be correct in identifying a synergy opportunity. Are they, however, the right person for overseeing that job?

Finally, the upside bias comes from a focus only on the upsides of synergy. Such efforts can enhance or undermine a sense amongst employees of their personal accountability for the performance of the enterprise. It can drive or inhibit organizational change. It can boost or crater employee motivation. It can improve or diminish managers’ conception of their roles. It can cut costs or grow them. Goold and Campbell stress the need of executives and managers to consider both the upsides and the downsides, and to do so in an objective, preferably quantitative way.

They gave the example of a consulting firm that decided to form a joint unit bringing together its IT consultants and strategic consultants. One night when working well into the evening, the strategy consultants ordered pizza and charged it to the client. The IT consultants were shocked that their colleagues were allowed to do this. This in turn led to a discussion about other differences between the units, including a realization that the strategy team not only enjoyed a raft of fringe benefits denied to the IT folks, but were being paid 50% higher, despite, as the joint venture showed, engaged in similar sort of work. What became known as the ‘pizza problem’ in the company only got worse as one team grew jealous of the other, driving friction between the units.

Efforts to resolve the problem failed and some of the company’s best IT consultants left in frustration. The synergy effort was a disaster all around. One might argue that this might never have happened had the IT workers been appropriately recompensed and that this was not really a synergy problem, but it was certainly revealed by the synergy effort. It is precisely these sorts of downsides that need to be considered.

Goold and Campbell argued that the best antidote against all these biases, as with all critical thinking, was awareness and discipline. Simply by being aware that these biases are ever-present dangers, executives, managers and workers are primed to spot such errors and ask hard questions at every step. In addition, the pair recommend being as descriptive as possible.

Synergy efforts repeatedly suffer from extremely vague formulations such as ‘sharing best practices’ or ‘cross-fertilizing ideas.’ These are not wrong, but how exactly do you know whether the team is achieving these goals? How do the team members know what needs to be done to implement them? The most important discipline in sound synergy strategy is instead a hard-nosed clarification of the objectives, benefits and potential risks – what Goold and Campbell call ‘sizing the prize.’ Be as precise as possible, disaggregate overarching goals into discrete objectives, and subject them to rigorous financial analysis.

Revisiting the subject at the turn of the millennium after a decade wherein many companies de-emphasised synergy, the two authors explored the importance of understanding where knowledge can exist in an enterprise regarding genuine synergy opportunities.As an example, they note that if most marketing managers in different countries believe there could be a synergistic win from sharing best practice in advertising, then it is probably a good idea for the corporate marketing director above them to pay close attention. Why? Because the ‘gut feel and intuitive judgement of experienced managers should carry considerable weight.’

There is a great body of tacit knowledge at these lower levels, especially after years of experience, that is not always filtered upward. Likewise, if the corporate marketing director is super keen on such an effort but the national marketing directors are reluctant, there is likely a good reason for this.

Related to this is the need for employees at all levels of a firm to feel that information can flow freely and they can express their views without penalty. Goold and Campbell note that in some enterprises, managers from different units might not meet together often enough to share views or knowledge, or they may even feel discouraged from doing so.

Thus, encouraging the setting up of synergy efforts in order to increase the sharing of knowledge is not the only direction in which this process occurs. An environment where the sharing of knowledge is the norm can be a precondition for effective synergy. Concretely, a structured approach to drawing out such tacit information can be helpful here, whether in the form of questionnaires, focus groups or other styles of systematic interviewing processes.

The Innovation Question

There doesn’t seem to be any evidence, however, that the primary reason organizations the likes of SpaceX, Tesla, Apple and Peloton have re-embraced the long-spurned corporate strategy of vertical integration is because they learned these lessons of how to do synergy well. Perhaps they did, but all we know for sure from executives reporting to us as to why they opted for such strategies is that any of the traditional potential drawbacks of strategic synergy efforts were easily overshadowed by the enormous gains resulting from the one undeniable major benefit barely or never touched upon by the 80s and 90s theorists of synergy, whether for or against: innovation. SpaceX and Tesla had no choice but to invent what they needed, because it just didn’t exist.

Further, Tesla wanted produce not only the best electric vehicle in the world, but one of the best vehicles in the world, period. This was necessary to avoid the trap of marketing electric vehicles only to those concerned enough about climate change to let it influence what car they bought – a very small sliver of the car-buying public.

Tesla’s vertical integration didn’t just make sense from a profit perspective, but also from a climate perspective. If we are going to solve climate change, everyone needs to be driving an electric car, not just those who care about the environment. And to make such a car, one that would wow even the biggest gearheads, would require abandoning the notion of optimizing one part on its own, but optimizing all parts in concert with each other.

The head of Peloton, John Foley, describes an almost identical experience. To produce his world-beating, multi-billion-dollar exercise equipment and media operation, Foley and his team initially thought they could develop their own software and electronics and then basically strap this onto someone else’s bike and yet another firm’s tablet computer 9.

Very quickly they realized that none of these parts if externally sourced would be remotely appropriate for what they were trying to do. So they designed their own bike to go along with the proprietary software and electronics, and integrated back into hardware manufacturing. They put together their own production studio and trained their own instructors rather than depending on existing retail spinning class outfits. And finally, they developed their own ‘white glove’ delivery and installation service and opened their own shops and e-commerce channel rather than depending on third-party stores.

As the 80s and 90s theorists of core competencies had warned, all this engorged investment requirements. But by delivering the Tesla of fitness bikes, a product and service whose every consumer touch point is superior to rivals, Peloton was able to raise almost a billion in private capital. As Leonard Sherman of Columbia Business School notes in his telling of the Peloton story, last year the company saw USD 1.8 billion in revenues, and enjoyed higher hardware gross profit margins, customer satisfaction and customer retention rates than Tesla or Apple 10.

The key lesson here is that such integration and synergy allow firms to innovate much faster than what occurs through outsourcing, as the different elements of production are better synchronized, tested and reformulated in service of a common goal thanks to in-house cooperation’s reduction of friction between those elements.

None of this is to endorse vertical integration or deep synergy as a strategy for every enterprise, or all units within an enterprise. Even Tesla does not make its own tires. But the enhancement of innovation delivered by such integration, and the advantage this offers, is certainly something enterprises need to keep in mind.

Returning back to the auto giants that thought they had captured the goose that laid the golden eggs with their embrace of outsourcing almost everything, we see in retrospect what a gamble this was. While unprecedented revenues were enjoyed for years as a result of the practice, the car giants today really don’t make cars at all. It can be argued that the delivery of all their know-how to their suppliers has left them unable now to know what to specify from them or to ask for technical improvements. The tail, uninterested in innovation, is wagging the dog. This has left these firms struggling to deal with the demands of a global economy shifting as fast as it can away from the internal combustion engine amid the climate emergency.

Given how the hubris of the evangelists of core competencies in the 80s has come a cropper, it behooves us in 2021 to be wary of being similarly over-confident with respect to integration and synergy. Perhaps the single greatest lesson from the last 40 years of such investigations is that there is still so much to learn about the science of human interaction within enterprises. By the time SpaceX, or whichever of the space-faring firms supporting NASA with reusable rockets, help get humanity to Mars, we can be sure only that much of the understanding of both synergy and core competencies, integration and outsourcing, at that point will be as radically different to ours as ours is to that of the 80s.

Picture of Leigh Phillips

Leigh Phillips

Leigh Phillips is a science writer and political journalist whose work has appeared in Nature, Science, New Scientist, the Guardian, the Daily Telegraph, MIT Technology Review, and the New Statesman, amongst other publications. He is also the author of two books, Austerity Ecology and the Collapse-porn Addicts, a progressive defence of economic growth and industry within sustainability discourse, and most recently with co-author and economist Michal Rozworski, The People's Republic of Walmart, a history of the economic calculation debate.


Share this article and let us know what you think. We're here to help and answer any questions you might have. Any suggestions or feedback? Let us know what you think and we will use your input for the future improvements.

Contact us

Join us

The return to journalism, the pursuit of truth and the utmost respect for solid, peer-reviewed science. You're just one click away from receiving the best of The Habtic Standard straight to your inbox. Subscribe to our monthly newsletter now and keep up to date with the latest corporate wellbeing insights from our experts around the globe.