XTV: Xerox’s Attempted Recovery From
“Fumbling the Future”
Following failures to capitalize on its own innovation, Xerox formed Xerox Technology Ventures to look for spin-off opportunities. Professor Henry Chesbrough outlines the history of XTV in this Business History Review excerpt.
By the end of the second spin-off regime, Xerox's position in the copier market had begun to improve. While its share of the market would never return to 80 percent levels, Xerox was able to regain more than ten market share points, beating back the Japanese and fighting off Kodak and IBM. During this period, Xerox's Palo Alto Research Center (PARC) delivered some important technologies for the corporation. One of the most critical inventions was laser printing for Xerox's printers and copiers, an effort that led to new Xerox products accounting for over $2.5 billion annually in revenue. 37 This confirmed the rationale back in 1970 for PARC's charter as a separate research center. As Robert Adams, then a Xerox senior vice president, recalled, "The laser printer alone paid for all of the other PARC research projects many times over. If some of the innovation results fall off the wagon, so what?"
Fumbling the Future was published in 1988, and the impact of this book reverberated throughout Xerox, stimulating many responses. As the story was recounted in the business and financial press, PARC's technological contributions to Xerox's imaging business, which were acknowledged in the book, were overwhelmed by Xerox's mismanagement of the computer applications for its technology. In response to the criticism set forth by the two authors, who were consultants for McKinsey, Xerox decided to improve its ability to profit from technology that was not utilized in its core businesses.
In reorganizing to capitalize on its technology, Xerox created a new entity in 1989, headed by Robert Adams, that effectively established a third phase for managing spin-offs. Adams had created a new business for Xerox out of PARC's laser technology, which had fostered a multibillion-dollar product line within Xerox. He had the support of Xerox's chief executive officer, David Kearns, for establishing the new internal structure. This structure would work much like an internal venture capital firm. Xerox would exploit Adams's skills in obtaining value from new PARC technologies by allowing him and his team to scan Xerox's technology base and identify investment opportunities that might otherwise have been missed. 38 This structure was called Xerox Technology Ventures (XTV).
Fumbling the Future was published in 1988, and the impact of this book reverberated throughout Xerox.
XTV was given $30 million to manage. It was set up formally as a corporate division within Xerox, but Adams negotiated special terms that mimicked many aspects of independent venture capital firms. Eighty percent of the gains of the fund would go to Xerox, while the XTV principals would share the remaining 20 percent among themselves. For investments below $2 million, the XTV principals had sole discretion in whether and when to invest. For investments over $2 million, an oversight group called the Management Board (consisting of the XTV principals, the Xerox chief executive officer, chief financial officer, and a senior staff executive) had final authority. This promised a far shorter deliberation process than was typical of the Innovation Board in the second period. 39
The intention of XTV was to align Xerox's incentives with those of the spin-offs and to employ venture capital processes to create value for its peripheral technologies. If there were promising technologies within Xerox that were not being utilized in Xerox's businesses, XTV provided a new path for them to enter the market. If that technology subsequently created significant economic value, XTV's investment enabled Xerox to participate substantially in the resultant value. As Kearns stated in 1993, "XTV is a hedge against the repeated missteps of the past." 40
Armed with this new structure and the charter to hunt for commercial possibilities within Xerox's technology portfolio, Adams and his partners began reviewing the company's inventions. The first opportunity they uncovered was Advanced Workstation Products (AWP). Xerox's high-end printers actually were complex computer systems that contained extensive networking connections. However, these internal computers were expensive because they were built with proprietary components, a manifestation of Xerox's approach to systems design. The development costs of these proprietary components had to be borne by the low unit volumes of its printers, in contrast to industry standard parts costs that could be spread over large volumes. Tony Domit, vice president of a network business unit, began developing a method that used off-the-shelf components to perform the Xerox networked printer-controller functions. Domit's solution used an IBM personal computer, some special chips, and a controller board to emulate the functions of the Xerox system. As Domit recalled, "The [Xerox proprietary] approach was costing around $15K and our approach was less than $1500, not including the IBM PC." 41 Domit's low-cost, "not invented here" system generated friction with internal Xerox units that were developing new extensions to their proprietary solutions. The internal development organization felt that an off-the-shelf solution would never perform well enough to substitute for its proprietary technology. This belief had frustrated Domit's ability to advance his inexpensive, informal approach within Xerox. Under the sponsorship of Adams, and armed with funding from XTV, Domit got the opportunity to develop his strategy outside Xerox's internal development units. Domit became the CEO of AWP, which gave him the resources to turn the print-controller system that he had designed into a final product. XTV invested $3.5 million and received 80 percent of the equity in the company. Becoming independent also gave the start-up organization the freedom to compete head-on with the internal Xerox development units in Rochester. To their surprise, Domit's group was able to raise the controller's performance to the level of the proprietary Xerox system—at one-fifth the cost. 42
This price performance breakthrough gave Xerox a strategic benefit, reducing the costs of its high-end printers. Subsequently, in January 1991, one and a half years after AWP was incorporated, Xerox bought the company back for $15 million. Domit and other AWP employees, who had received a 20 percent equity stake in the new company, now profited handsomely from their work. Furthermore, the XTV partners received roughly $2.5 million from the transaction. Certain Xerox managers, who did not wish to be identified, regarded XTV's compensation as grossly excessive relative to the time and effort the partners had invested in AWP. 43
XTV's third investment, Documentum, proved to be its primary money-maker.
XTV's third investment, Documentum, proved to be its primary money-maker. Documentum was a company formed as a direct result of the strategic intent of XTV. Xerox had done extensive research on document management, but none of the projects had resulted in new products or services. To be sure, some project results had filtered into Xerox's printers and copiers, but only as individual components of a larger system.
Robert Adams considered that the PARC document-management solutions offered pivotal know-how that could open doors for Xerox's high-end networked laser printers, while producing an outstanding business opportunity. Adams invited Howard Shao and John Newton to come to PARC and canvass its current research on document management, with a view to identifying a potential spin-off. Shao and Newton had both worked together at Ingres, a database software company. Shao had been a member of one other startup company and had spent some time at CitiCorp. The two men spent about six months talking with researchers at PARC. Much of that time was occupied with interviewing prospective customers who had significant problems with document management. For example, a pharmaceutical company that was applying for FDA approval had to manage hundreds of documents that ran into tens of thousands of pages. As Shao said later, "Xerox PARC had all of the pieces to implement a document management solution. They just didn't know what to do with the pieces, and how to put them together." 44
Documentum was formed in January 1990, and XTV took an initial 80 percent stake in the company. Beyond the founders' systematic and thorough approach to marketing, the Xerox association was critical to the startup's success for two reasons: it gave them direct access to customers who were interested in their intended products, and it gave them credibility with those customers. 45 However, Documentum's products were designed to work with IBM systems and with those of other high-end systems suppliers, as well as those of Xerox. This compatibility with such systems was inconsistent with Xerox's preferred strategy to develop and sell complete Xerox solutions. It also required Documentum to rewrite all the software code, in order to create products that were compatible with IBM and other third-party hardware.
With subsequent rounds of financing, XTV's ownership percentage declined. This was the only XTV spin-off that was syndicated to outside venture investors. By January 1996, Documentum had its initial public offering, creating a market value of $323 million at the end of the first day of trading. It would prove to be the most successful investment of all the XTV companies. Even though Xerox's share had declined to 38 percent by the time of the IPO, its stake at the end of the first day of trading amounted to $106 million, the largest financial return to Xerox made by any of the spin-offs described in this study. Xerox then progressively sold off its holdings in Documentum once it became public.
Despite XTV's tremendous financial success with Documentum, the venture nonetheless had its critics within Xerox. One senior research manager felt that Documentum had "stolen" some of PARC's best work. He believed that much of Documentum's success had been achieved through sales to Xerox's top corporate accounts by Xerox's own sales force at the expense of other Xerox products. 46
A final contributory factor in the demise of XTV was that its financial success may not have directly translated into gains for Xerox's shareholders.
The other significant financial success in XTV's portfolio, Document Sciences (DSC), was started in the fall of 1991. Other XTV investments in Xerox spin-offs were less successful. All of XTV's portfolio companies were created and funded by the end of 1992, with the result that its focus shifted to providing them with additional funding when and if additional funds were justified. During the XTV period, though, three PARC technologies emerged from Xerox without XIV's involvement: LiveWorks, Customers and Technologies, and Xerox ColorGraphX. Each of these companies was formed after XTV had committed its available funds. XTV's participation was not sought in any of the three ventures, either by Xerox or by the Xerox researchers who were involved.
LiveWorks arose out of an ambitious project to commercialize group collaborative technology concepts that had been incubating within PARC since the mid-1980s. This project was sponsored by the new Technology and Market Development (TMD) organization, formed under Roger Levien at Xerox headquarters. Rather than being culled from research efforts that were viewed as marginal by Xerox, LiveWorks commercialized research in an area deemed central to Xerox's strategy: ubiquitous computing. Ubiquitous computing was built on multimedia, networked, and portable electronic-document access technologies, and on previous PARC research in fields like cognition and anthropology. The culmination of these efforts in 1991 produced two innovative systems devices: the E-Case, a portable electronic-document reader, which would later become the technological basis for a spin-off called Uppercase; and the LiveBoard, a sophisticated workgroup interaction tool that was the basis for LiveWorks. The LiveBoard was a large electronic whiteboard with a sixty-seven-inch diagonal display that users could access or write on from across a room with special electronic pens. Users could simultaneously view people writing on LiveBoards at two separate sites. This allowed work groups in separate cities to collaborate in real time. The devices, which were produced starting in 1994, cost between $40,000 and $50,000 each.
PARC researcher Richard Bruce, who had been leading the project since 1989, became president of the group, which started with about thirty full-time "consultants." The group had already completed twelve preproduction LiveBoard units by the time it officially became LiveWorks in August 1992. However, it faced significant challenges, including the lack of a viable manufacturing plan and significant cost overruns.
An outside CEO was brought in to replace Bruce once the venture was formally launched as an independent entity. The company continued to struggle to meet its costs and to supply LiveBoards of acceptable quality. Despite reaching revenues of $14 million, the company continued to incur significant financial losses. By 1996, the outside CEO had left and was replaced by Timothy da Silva. 47 After investing over $35 million in LiveWorks, Xerox officially shut it down in July 1997.
Customers and Technologies was formed by two Xerox researchers who left without formal permission from Xerox in 1994. They attempted to use the findings of Xerox's research on document management by offering a consulting service to commercial clients. The company operated for a number of years but never achieved significant revenues. ColorGraphX was spun off from Xerox under Barry Lathan in 1996 but was then incorporated back into Xerox's large-format printing business, Xerox Engineering Services, three years later. It similarly did not make much money.
38. It also acted as something of a consolation prize for Adams, who was passed over for a promotion within Xerox's management, according to William Spencer. Interview at his residence in San Francisco, 17 Aug. 2000.
43. I heard many charges and accusations made "off the record," and generally, exclude such criticisms from this account. I include this one for two reasons: first, that it was made by three separate individuals at different stages of the research; and two, it foreshadowed the later decision to terminate the XTV structure, despite its financial success. This last decision is difficult to understand, absent these sentiments.
45. Interview with Larry Mansinter, 29 Feb. 2000, at his offices at Documentum. Mansinter had worked on software projects at PARC since 1973 and had been involved in the Documentum spin-off process from the technical side of Xerox.
46. This perception is not shared by respondents from Documentum. While such a dispute cannot be adjudicated ex post, there was an organizational attempt by Xerox to utilize its document management technology more directly. About one year after Documentum was spun off, Xerox created an internal business division called Xsoft. Xsoft's products would compete in the same space as Documentum in document management software. In the words of Xerox president Allaire, "[B]y consolidating document software development and marketing [into Xsoft), we can respond more quickly to the ever-changing needs of our customers" (Business Wire, 22 Aug. 1991). In the event, Xsoft's financial results were quite minimal, compared with those of Documentum. By 1995, Xsoft sold off many of its businesses, and others were broken off into separate Xerox New Enterprise companies in the fourth stage.
47. Some of the facts surrounding the failure of LiveWorks were unavailable to me because of the terms of a settlement of a lawsuit between Xerox and three of the outside professional managers (which did not include Bruce) of the failed venture.
Excerpted with permission from "Graceful Exits and Missed Opportunities: Xerox's Management of Its Technology Spin-off Organizations," Business History Review 77, Winter 2002.