In 2012, one of us — Gregor Gimmy, a California-based serial entrepreneur and former IDEO consultant — accepted a new role at BMW’s corporate R&D headquarters. Gimmy’s task was clear but highly demanding: to reimagine the way BMW innovates.
At the time, BMW had no dedicated, company-spanning unit to leverage the creative power of startups. This meant that the company was leaving out huge innovation potential — thousands of startups with billions of funding — that could help BMW innovate anything from core vehicle technology (batteries, sensors, artificial intelligence software) to manufacturing innovations (internet of things, cybersecurity, robotics). To fill the void and build such a new BMW startup unit, Gimmy partnered with an experienced innovation manager from BMW, Matthias Meyer.
The prevalent model of startup cooperation in recent years has been corporate venture capital and accelerators (CVC&A). From 2012 to 2015, the number of global corporate venture capital deals almost doubled, and their investments quadrupled, to $29.1 billion. Among the 30 top companies in seven of the largest industries, almost half had a VC-fueled accelerator in 2015, up from just 2% in 2010.
However, despite these massive resources and C-level attention, all this corporate venturing seems to be failing to accomplish strategic innovation goals. In recent years, many high-profile companies, including Volkswagen, Yahoo, Turner/Warner Bros, and Coca-Cola, have shuttered their VC arms or accelerators entirely. In fact, in 2011 BMW had also founded a corporate venture capital (CVC) unit, called BMW iVentures, but it exclusively invested in service startups. Gregor and BMW faced a crucial question: “How can the BMW Group, as a company, co-innovate with startups?”
Based on his own experience, a reading of the emerging research, and dozens of conversations, Gimmy was convinced that the innovation impact of corporate VCs had been disappointing not because external startups do not have value to offer to large companies — after all, startups often develop precisely the sort of innovative solutions that large incumbents need the most — but because they could not master critical requirements to meet strategic innovation goals, three of which are particularly essential:
First, his time in Silicon Valley had shown Gimmy that corporate VCs and accelerators struggle to lure the best startups from private venture capitalists. Talented founders know that corporations simply cannot replicate the deep experience private VCs have in starting companies, nor their expertise in assisting startups with complex challenges such as deal making, business modeling, resolving disputes among founders, executing a successful IPO, and so on. In addition, top startups are heavily oversubscribed, and competition to get into a deal is fierce. Finally, corporate VCs are unable to pitch the promise of a self-fulfilling prophecy the way successful private VCs can. As a result, top founders prefer independent, noncorporate accelerators, and, to date, no corporate accelerator has truly accomplished incubating world-class startups.
Second, against all hope, CVC&A have inherent difficulties with regard to innovation transfer and integration. Only around 20% of technologies funded by CVC&A grab enough attention of business units to start co-innovation pilot projects with portfolio startups.
Third, corporate VCs and accelerators are costly and complex to operate, turning them into a slow and expensive innovation tool. Our research — more than 60 in-depth interviews with executives at large corporations and founders of startups as well as numerous innovation projects with large and midsize companies — shows that it takes at least 12 months between the first touchpoint of the CVC with a startup and the kick-off of a pilot with the business unit. And the fixed cost from “touchpoint-to-pilot” are immense. For example, in the case of a $100 million CVC fund, which can close five to 10 investments a year, these costs typically range from $1 to $2 million per startup — not including the administrative and variable costs of the pilot itself.
Gimmy’s insight was that the key to overcoming these challenges and to improving corporate co-innovation is not for companies to compete more effectively with private venture capitalists or accelerators, but to focus on what they can offer to top startups that others cannot. He noted that, in general, startups require three things to grow: capital, coaching, and clients. Private VCs and other professional investors can provide the first two. But only large corporates can offer the last one. Top startups already have market leading solutions. They need clients — that is, purchase orders — far more than advice or capital.
Creating the Venture Client Model
Based on this insight, Gimmy proposed a radical rethinking and new corporate venturing tool to boost corporate innovation: the “venture client” model. Its principle is very simple, yet it has a profound impact on enabling incumbents to attract top startups, and on assuring high integration rates quickly and at low fixed costs. In essence, the venture client, instead of equity, buys the technology of a startup when it is still a venture to do so. In this arrangement, private VC funds and accelerators do the complex work of sorting high-potential startups from the also-rans and nurturing them through their nascence. Corporates then enter further downstream by partnering as the first big client with young companies that have either graduated from an accelerator or received professional VC funding. The first purchase is a “minimum viable purchase,” since the incumbent buys just a sample of the startup’s solution for validation in a real pilot project conducted by the business unit. Selected startups become real suppliers, with purchase orders and supplier numbers, from day one of the venture client program.
Gimmy knew that many of the most significant breakthroughs in innovation result from collaborations between large corporates working with small startups as a client rather than an investor. In fact, BMW’s greatest co-innovation success, 12 years earlier, had been in a “venture client” relationship with an early stage startup: Mobileye, now a leader in collision-avoidance technology. Similarly, Charles Schwab was the early adopter client of Siebel’s revolutionary CRM system. And Apple’s Steve Jobs recounts in a conversation known as “the lost interview” how he met a startup in a Silicon Valley garage and decided to drop all internal projects and partner with the young firm, known today as Adobe, a deal that eventually allowed Apple to become the world leader in desktop publishing. Could these experiences be replicated and scaled up at BMW?
To find out, Gimmy and Meyer designed the new organizational unit with a set of processes and a startup program based on the venture client model. The idea was to identify and collaborate with early stage startups — usually young VC-backed companies with a functional prototype but no track record or business model. The two managers secured buy-in from top BMW R&D executives to launch such a unit, and they received top-management approval for the unit in July 2014. Everyone understood that being the flagship client for startups at this stage in their development would attract top startups and give BMW first access to cutting-edge technology, product customization advantages, time-to-market, and pricing, while reducing the risk profile of traditional corporate venturing. Obviously, such “venture client” collaboration was possible at BMW before the launch of that unit, but it was lengthy and cumbersome due to complex administrative barriers.
A New Name and a Unique Brand Identity
The first challenge to making the startup unit successful was to become known in the global startup community and to connect fast with promising startups. So Gimmy decided to create a new name and a unique brand identity for the venture client unit that was independent of the corporate parent yet clearly recognizable as part of the BMW Group: the BMW Startup Garage. He also launched a new website and social presence on Facebook, Twitter, and YouTube next to the social media presence of the core BMW brand.
In the meantime, the website has become not only the central showcase of the Startup Garage but also the focal hub for startups to submit information about their innovations. Apart from the website, the BMW Startup Garage team also stays in regular touch with leading venture capitalists and regularly appears at international conferences, as well as hosting their own startup events. It also proactively reaches out to startups that have relevant solutions for strategic innovation initiatives such as autonomous driving, electric mobility, or additive manufacturing (3D printing).
The next phase of the BMW Startup Garage process involved whittling down startups to find those that have a strategically relevant technology, product, or service. Gimmy decided that, in order to outsource its filtering to professional investors, BMW would not consider a startup unless it has graduated from a prestigious accelerator, has received professional venture funding, or includes a successful serial entrepreneur as part of its senior leadership team. Even so, less than 5% of all potential startups will make it past a screening process. For those that do, the last step prior to acceptance into the program is to define a pilot project between the startup and a business unit. Once the startup proposal is accepted, the Garage issues a supplier number and purchase order, recognizing the startup as a bona fide vendor. (The team has an agreement with central procurement that allows it to issue supplier numbers and purchase orders without a tender for the startups.)
Harnessing Big Potential While Minimizing Risk
The initiative is still young, but this is only one of many early signs of success: More than 1,000 startups have been evaluated since the launch of the program in 2015, and BMW expects this figure to grow to over 2,000 per year, with more than 80% coming from outside Germany. By summer 2017, 90% of accepted startups have met BMW expectations and are continuing to grow their business with BMW.
Starting as a new pilot innovation unit, the BMW Startup Garage grew organically to include startups for all BMW divisions (including services, IT, manufacturing, and HR). The model is also showing important synergies with BMW iVentures, as it lowers the investment risk and keeps deals moving. Moreover, it has been highly praised by startups, VCs, and the media. The number of startups assessed and transferred has increased over 300% in the first two years. Other corporates are starting to adopt the venture client model, among these other large multinationals such as Daimler, Continental, Caterpillar, and Viessmann.
We believe that the venture client approach is unique and has enormous potential, not just at BMW but also at other companies. Perhaps the most important reason for its success is that all stakeholders — startups, private venture capitalists and corporate — focus on what they do best while minimizing risk. Moreover, through venture clienting, startups gain valuable insights into technical, process, and quality requirements and the workings of an actual corporation, which helps them become savvier and more adept at business development. The venture client approach also gives them real-world feedback on their products and services. In addition, an established early client is a terrific motivator for founding management teams — and can be very useful in recruiting additional talent.
All in all, we recommend companies to consider the venture client model as a way to co-innovate with startups. Of course, before doing so, the challenge will be to change executives’ mindset about what startups really are, and what corporate venture capital can offer the startup: A top startup is a market leader who needs clients — “purchase orders” — no further investment and, definitively, no belittling.