• 06 Nov 2000
  • Research & Ideas

The Determinants of Corporate Venture Capital Success

Corporate-sponsored venture capital funds do not have to fail. But as HBS professors Paul Gompers and Josh Lerner explain, hybrid organizations such as Xerox Technology Ventures face considerable challenges on the road to success.
by Paul Gompers & Josh Lerner

Corporate-sponsored venture funds first appeared in the 1960s, about two decades after the first flowering of the venture capital industry. Ever since, they have mirrored the cyclic nature of the industry as a whole.

But there are important differences, write Gompers and Lerner, between corporate and private venture funds. Corporate efforts lack the partnership-based organizational structure that many have identified as critical to the success of private equity funds. On the other hand, corporations are able to exploit complimentarities with their existing lines of business, giving them—potentially, at least—a strategic edge that private funds lack.

These differences—good and bad—are illustrated by the case of Xerox Technology Ventures (XTV).

The case, write Gompers and Lerner, "highlights the fact that—contrary to both popular wisdom and academic arguments—corporate venture programs can still be successful without the traditional partnership structure."

But XTV's example, they continue, also suggests "some of the difficulties that these efforts encounter and the apparent importance of having a strong linkage between the fund's investment focus and the corporate parent's strategic focus."

Determinants of Corporate Venture Capital Success

The Xerox Corporation originated as a photography-paper business called the Haloid Company. 3 The Haloid Company's entrance into what would later become its principal business came in 1947, when it and the Battelle Memorial Institute, a research organization, agreed to produce a machine based on the recently developed process named xerography. Invented by the patent lawyer Chester Carlson, xerography involved a process by which images were transferred from one piece of paper to another by means of static electricity. Rapid growth and a redirection of the company's emphasis toward xerography characterized the Haloid Company in the 1950s. In 1961, in recognition of the spectacular growth of sales engendered by the first plain-paper copier, the firm was renamed the Xerox Corporation.

In response to IBM's entrance into the copier field in the late 1960s, Xerox experimented with computers and with designing an electronic office of the future. It formed Xerox Computer Services, acquired Scientific Data Systems, and opened its Palo Alto Research Center (PARC) in California. These efforts were only the beginning of the copier giant's effort to become a force in the computer industry. Throughout the 1970s, Xerox completed several acquisitions in order to further its project for an "architecture of information." Unfortunately, in assembling these noncopier companies and opening PARC, Xerox created a clash of cultures. Differences between its East Coast operations and its West Coast computer people would severely affect the company.

The focus for much of this division was PARC. In the 1970s, PARC was remarkably successful in developing ingenious products that would fundamentally alter the nature of computing. The Ethernet, the graphic user interface (the basis of Apple Computer's and Microsoft's Windows software), the "mouse," and the laser printer were all originally developed at PARC. The culmination of much of PARC's innovation was its development of the Alto, a very early personal computer. The Alto's first prototype was completed in 1973, and later versions were placed in the White House, Congress, and various companies and universities. Nonetheless, the Alto project was terminated in 1980.

Technology And Strategic Objectives

Inherent in the Alto's demise was Xerox's relationship with PARC. Xerox did not have a clear-cut business strategy for its research laboratory, and, in turn, many of PARC's technologies did not fit into Xerox's strategic objectives. For instance, the Alto's ability to adapt to large customers' computer systems was inconsistent with Xerox's strategy of producing workstations compatible only with its own equipment.

The establishment of XTV was driven by two events in 1988. First, several senior Xerox managers were involved in negotiating and approving a spin-off from Xerox, ParcPlace, which sought to commercialize an object-oriented programming language developed at PARC in the 1970s. The negotiation of these agreements proved to be protracted and painful, highlighting the difficulty that the company faced in dealing with these contingencies. More important, in this year a book documenting Xerox's failure to develop the personal computer, Fumbling the Future (Smith and Alexander 1988), appeared. Stung by the description in the book, Xerox chairman David Kearns established a task force with the mandate of preventing the repetition of such a failure to capitalize on Xerox innovations.

The task force reviewed Xerox's history with corporate venture programs. Xerox had invested in venture-backed firms since the early 1970s. For instance, it had joined a variety of venture capitalists in investing in Rolm, Apple, and a number of other firms. While the investments were successful financially, they were made on an ad hoc basis. In the early 1980s, Xerox established two venture funds with an external focus. These did not prove particularly successful, largely owing to disputes within the firm about appropriate investments. The task force, in member (and future XTV president) Robert Adams's words, rapidly "concluded that we needed a system to prevent technology from leaking out of the company" (Armstrong 1993). The committee focused on two options: (1) to begin aggressively litigating those who try to leave with new technologies and (2) to invest in people trying to leave Xerox. Owing to variations in employee noncompetition law across states (and particularly the weak level of protection afforded by the California courts), it was unclear how effective a policy of aggressive litigation would be. Furthermore, such a policy might reduce Xerox's ability to recruit the best research personnel, who might not want to limit their future mobility.

On the basis of the task force's recommendation, Kearns decided to pursue a corporate venture capital program. He agreed to commit $30 million to invest in promising technologies developed at Xerox. As he commented at the time, "XTV is a hedge against repeating missteps of the past" (Armstrong 1993). He briefly considered the possibility of asking an established venture capital firm to run the program jointly with Xerox but decided that the involvement of another party would introduce a formality that might hurt the fledgling venture.

Modeling XTV after venture organizations had several dimensions. The most obvious was the structure of the organization. While this was a corporate division rather than an independent partnership (like most venture organizations), the XTV partners crafted an agreement with Xerox that resembled typical agreements between limited and general partners in venture funds.

Minimizing Organizational Disruption

The spinout process was clearly defined in the agreement in order to ensure that disputes did not arise later on and to minimize the disruption to the organization. The XTV officials insisted on a formal procedure to avoid the ambiguity that had plagued earlier corporate ventures. The agreement made clear that the XTV partners had the flexibility to respond rapidly to investment opportunities, something that independent venture capitalists typically possess. They essentially had full autonomy when it came to monitoring, exiting, or liquidating companies. The partners were allowed to spend up to $2 million at any one time without getting permission from the corporation. For larger expenditures, they were required to obtain permission from XTV's governing board, which consisted of Xerox's chief executive officer, chief financial officer, and chief patent counsel.

Similar to independent venture organizations (but unlike many corporate programs), the program also had a clear goal: to maximize return on investment. The XTV partners felt that the ambiguous goals of many of the 1970s corporate venture programs had been instrumental in their downfall. They hoped to achieve a return on investment that exceeded both the average returns of the venture capital industry and Xerox's corporate hurdle rate for evaluating new projects.

Not only was the level of compensation analogous to that of the 20 percent carried-interest that independent venture capitalists received and the degree of autonomy similar, but XTV operated under the same ten-year time frame employed in the typical partnership agreement. Under certain conditions, however, Xerox could dissolve the partnership after five years.

The analogy to independent venture organizations also extended to the companies in which XTV invested. These were structured as separate legal entities, with their own boards and officers. XTV sought to recruit employees from other start-ups who were familiar with managing new enterprises. The typical CEO was hired from the outside on the grounds that entrepreneurial skills, particularly in financial management, were unlikely to be found in a major corporation. XTV also made heavy use of temporary executives who were familiar with a variety of organizations.

The independence of management also extended to technological decision making in these companies. The traditional Xerox product—for instance, a copier—was designed so that it could be operated and serviced in almost any country in the world. This meant not only constraints on how the product was engineered but also the preparation of copious documentation in many languages. These XTV ventures, however, could produce products for "leading-edge" users, who emphasized technological performance over careful documentation.

Giving Up Control

Like independent venture capitalists, XTV intended to give up control of the companies in which they invested. Transferring shares to management and involving other venture capitalists in XTV companies would reduce Xerox's ownership of the firm. Their goal was that, over the long run, after several rounds of financing, Xerox would hold from 20 to 50-percent equity stake. XTV sought to have under a 50 percent equity stake at the time a spinout firm went public. In this way, it would not need to consolidate the firm in its balance sheet (i.e., it would not need to include the company's equity on its balance sheet, which would reduce Xerox's return on equity). The Xerox lawyers had originally wanted only employees to receive "phantom stock" (typically, bonuses based on the growth in the new units' performance). Instead, XTV insisted that the employees receive options to buy real shares in the venture-backed companies, in line with traditional Silicon Valley practices. The partners believed that this approach would have a much greater psychological effect as well as a cleaner capital structure to attract follow-on financings by outside investors.

Between 1988 and 1996, the organization invested in over one dozen companies. These covered a gamut of technologies, mostly involving electronic publishing, document processing, electronic imaging, workstation and computer peripherals, software, and office automation. These not only successfully commercialized technology lying fallow in the organization but also generated attractive financial returns.

One successful example of XTV's ability to catalyze the commercialization of technological discoveries was Documentum, which marketed an object-oriented document-management system. Xerox had undertaken a large number of projects in this area for over a decade prior to Documentum's founding, but had not shipped a product. After deciding that this was a promising area, XTV recruited Howard Shao and John Newton, both former engineering executives at the Ingress Corporation (a relational database manufacturer), to lead the technical effort.

Shao spent the first six months assessing the state of Xerox's knowledge in this area—including reviewing the several 300-plus-page business plans prepared for earlier proposed (but never shipped) products—and assessing the market. He soon realized that, while Xerox understood the nature of the technical problems, the company had not grasped how to design a technologically appropriate solution. In particular, the Xerox business plans had proposed building document-management systems for mainframe computers rather than for networked personal computers (which were rapidly replacing mainframes at many organizations). With the help of the XTV officials, Shao and Newton led an effort rapidly to convert Xerox's accumulated knowledge in this area into a marketable product. Xerox's accumulated knowledge—as well as XTV's aggressive funding of the firm during the Gulf War period, when the willingness of both independent venture capitalists and the public markets to fund new technology-based firms abruptly declined—gave Documentum an impressive lead over its rivals.

Documentum went public in February 1996 with a market capitalization of $351 million. 4 XTV was able to exit a number of other companies successfully, whether through an initial public offering, a merger with an outside firm, or a repurchase by Xerox (at a price determined through arm's-length bargaining). A conservative calculation (assuming that Xerox sold its stakes in firms that went public at the time of the initial public offering rather than later, after prices had substantially appreciated, and valuing investments that Xerox has not yet exited or written off at cost, less a 25-percent discount for illiquidity) indicates that the $30 million fund generated capital gains of $219 million. Given the 80/20-percent split established in the XTV agreement, the proceeds to Xerox should have been at least $175 million, those to the three XTV partners at least $44 million.

Internal Rate Of Return

The same assumptions suggest a net internal rate of return for Xerox (i.e., after fees and incentive compensation) of at least 56 percent. This compares favorably with independent venture capital funds begun in 1989, which had a mean net return of 13.7 percent (an upper-quartile fund begun in that year had a return of 20.4 percent) (Venture Economics 1997).

These calculations of Xerox's internal rate of return (IRR) do not include any ancillary benefits generated by this program for the corporation. For instance, some observers argued that high-expected-value projects that might otherwise not have been funded through traditional channels (owing to the high risk involved) were increasingly funded during this period, apparently out of the fear that they would otherwise be funded by XTV and prove successful.

Despite these attractive returns, Xerox decided to terminate XTV in 1996, well before the completion of its originally intended ten-year life. 5 The organization was replaced with a new one, Xerox New Enterprises (XNE), which did not seek to relinquish control of firms or to involve outside venture investors. The XNE business model called for a much greater integration of the new units with traditional business units. The autonomy offered to the XNE managers and their compensation schemes were much closer to that afforded in a traditional corporate division. As such, XNE appears to represent a departure from several of the key elements that the XTV staff believed were critical to that company's success, such as a considerable degree of autonomy and high-powered incentives.

The experience of Xerox Technology Ventures has several implications for corporate venture capital programs more generally. First, the case makes clear that, contrary to the suggestions in writings by both venture capitalists and financial economists, corporate venture capital programs need not fail. As noted above, Xerox's financial returns were exceedingly favorable when compared to returns from comparable independent venture funds. Second, XTV's successes—such as Document Sciences and Documentum—were concentrated in industries closely related to the corporate parent's core line of business (document processing). This suggests that the fund's strong strategic focus was important in its success. Finally, the Xerox Corporation was unable to commit to a structure akin to that of a traditional venture capital partnership. Despite efforts by XTV's founders to model the fund as closely as possible after a traditional venture partnership, the fund was still dissolved early. This experience underscores the challenges that these hybrid organizational forms face.

About the Author

Josh Lerner is the Jacob H. Schiff Professor of Investment Banking at Harvard Business School.