Open Innovation, page 12
3Com
3Com Corporation was the first of several highly successful spin-off companies based on technologies created at Xerox PARC. Robert Metcalfe was a young computer scientist at Xerox PARC when he invented the Ethernet local area network (LAN) technology.18 Used within PARC as early as 1975, this technology connected different parts of Xerox’s computers and its copiers. Sensing the latent opportunity of Ethernet and impatient with Xerox’s indecision about commercializing PARC’s pioneering technologies, Metcalfe left PARC in January 1979. He formed 3Com Corporation (“computers, communication, compatibility”) in June of that year.
While pursuing his vision for 3Com, Metcalfe had to find ways to support himself. He was soon engaged as a networking consultant to Digital Equipment Corporation (DEC) by Gordon Bell, then the leading technical figure at DEC. In 1980, with Bell’s encouragement, Metcalfe successfully persuaded Xerox to grant him a nonexclusive license to the Ethernet technology, on which Xerox held four strong patents, for the sum of $1,000.
Xerox’s agreement to this proposal reflected a strategic choice rather than an oversight. Xerox was a large user of DEC computers and was eager to promote a technology to link Xerox printers and workstations to DEC minicomputers. DEC’s help would be vital to accomplishing that.19 By licensing the Ethernet technology, Xerox could promote its Star systems products. Spurred by Metcalfe’s efforts, Digital,Intel, and Xerox formed an alliance (DIX) to define a standard for Ethernet LAN communication and to promote its widespread adoption asan “open standard” by the computer industry.20 By comparison, the IBM PC would not be announced until August 1981.
Armed with the DIX alliance, 3Com began to seek venture capital in October 1980 in order to begin developing hardware products. In the absence of established markets for either PCs or workstations, the business plan for 3Com was necessarily vague. The search nonetheless paid off in February 1981, with first-round funding of $1million from VC investors who looked beyond the formal plan and were attracted by Metcalfe’s vision and charisma, as well as his team’s strong technical talents.
Metcalfe’s venture was hardly an instant success. 3Com’s first products connected DEC minicomputers to Ethernet LANs, using Intel chips. This was a market in which a company sold primarily to scientists and engineers who used Unix operating systems and who did much of their own programming. Distribution was accomplished through direct sales or value-added resellers. Ungermann-Bass was the leader in this market, with 3Com lagging behind, partly because of 3Com’s much smaller direct sales force.
3Com realized much greater success in the IBM PC marketplace, selling its Ethernet adapter cards to be installed inside IBM-compatible PCs in corporate networks running Novell’s operating system. The core value proposition became the ability to share files and laser printers (which in those days were very expensive) via an Ethernet network protocol that was compatible with the nascent IBM PC standard. Later, Ethernet would also enable companies to use e-mail within their LANs,and still later, Ethernet helped networks connect to the Internet.
Once the PC business began to boom and 3Com had shifted away from its initial focus on workstations, 3Com began to take off as well. 3Com stock was first sold to the public in 1984, and the company was still operating as an independent company in 2002, with a market value at the end of 2001 equal to one-third of Xerox’s market value.
Did Xerox make a mistake by licensing Ethernet for a mere $1,000? As this account shows, the latent economic potential of Ethernet was far from obvious at the time that Xerox decided to grant the license. In fact, Xerox was advancing its own strategy for its Star networked systems by agreeing to the license, in order to connect its equipment more effectively with DEC minicomputers.
Ethernet’s value arose because the technology was commercialized in a new business model outside of Xerox workstations, DEC minicom-puters, and the Unix operating system. The key ingredients of that model stood in sharp contrast to the business model of Xerox, which exploited unique proprietary technologies and sold them through a direct sales system to its leading office equipment customers (table 4-3). The latent value in the Ethernet technology did not materialize until the technology was targeted at a different market, which offered a different value proposition, utilized an open-technology platform populated by many third parties, and was sold through a new set of distribution channels. It seems reasonable to infer that a business model similar to 3Com’s would not have evolved had the technology remained within Xerox. And Xerox could not have anticipated the value latent within the technology, unless it had conceived of a radically different way to take that technology to market.
TABLE 4-3
* * *
Summary Evaluation of Xerox and Selected Spin-Offs on Key Business Model Attributes
Xerox 3Com Adobe Metaphor
Identified Market Segment Corporate and government market Corporate PC Market PC, MAC, and laser printer market Knowledge workers in corporations
* * *
Value Proposition High-quality copies at a low montly lease rate Establishes file and printer sharing between IBM and PCs Enable output of richer document types Enables nontechnical queries of corporate databases
Elements of Value Chain Developed entire copiersystem,including supplies, sold through a direct sales force Focused on Ethernet protocol and add-on boards Focused on supplying fonts to laser printer manufacturers and software firms Developed and sold entire systems, from hardware to software distribution
Defined Cost and Margins Modest profit on equipment,high profit on supplies, or per “click” High volume, low unit cost Very high fixed cost, very low variable cost High fixed costs, high margin, low unit volume
Positioned in Value Network First mover in dry-copy process; did not require or pursue partners Set the IEEE 802 standard; utilized PC distribution channel Defined the Postscript standard for scalable fonts No third parties or complementors utilized
Formulated Competitive Strategy Competed on technical product quality, product capability Compete on standard, new channels Strong netword externalities, high switching costs Compete on superior technology, usability
* * *
Adobe
The spin-off of Adobe from Xerox followed a path similar to that taken by 3Com. Adobe’s founders, Charles Geschke and John Warnock, left PARC in 1983 to commercialize a page-description language that be-ame their first product, PostScript. PostScript allows printers to use digital fonts to reproduce a wide variety of characters generated from a PC.
The technology embodied in PostScript came from Interpress, a page-description software project developed while Warnock and Geschke were at Xerox PARC. (The project had drawn on earlier work they had done at Evans and Sutherland—this would later complicate Xerox’s ability to control the ideas exclusively for itself.) Interpress was an internal, proprietary protocol used to print fonts generated from Xerox workstations on Xerox printers. This was an effective usage of the technology, because it linked tightly with Xerox’s own business model and gave Xerox’s products a competitive edge over other systems. But the potential value of the technology was limited to that of an important proprietary component in a larger Xerox system.
While at PARC, Warnock and Geschke had argued repeatedly with Robert Adams, then the head of Xerox’s printing division, over whether to make Interpress into an open standard. Adams had strongly resisted, contending that he couldn’t see how Xerox would make any money if the company “gave away” the font technology and weakened one of the most distinctive features of Xerox’s own systems. After debating this inside Xerox for more than a year, they agreed to disagree, and Warnock and Geschke left PARC. As Geschke remembered it,“Certainly, within Xerox, none of this was going to happen. They wanted to have an industry standard, but they wanted to control everything at the same time.”21
Arguably, Adams was at least partly right: It may well have been that Xerox’s business model could never have benefited from making the technology an open standard. The business model that eventually realized significant economic value for Adobe differed substantially—both from Xerox’s business model and from Warnock and Geschke’s original intentions when they left. Indeed, Adobe’s initial business model had contained many elements that were similar to the model then dominant at Xerox, but subsequent events persuaded the founders to change it. AsGeschke recalled,
Our original business plan was different. We were going to supply a turnkey systems solution including hardware, printers, software, etc. With this in hand, we were then going to build a turnkey publishing system. It turns out other people were trying to do this at the same time—there would have been a lot of competition if we had gone this route. . . . In many respects Steve Jobs and Gordon Bell (my teacher in graduate school) were key ingredients in getting things going the way they did. Gordon said, “don’t do the whole system,” and Steve came to us and said, “we don’t want your hardware, just sell us the software.” We said, “No!” Later Steve came back and said, “OK,then just license it to me.” That’s how the business plan formed. It wasn’t there in the beginning.”22
Selling and supporting a turnkey publishing system, complete with its own hardware and software, would have required a direct sales force and a field service network very much like the one Xerox managed in its copier business. In Geschke’s view, such a system would have taken along time to be developed and would have encountered a lot of competition. The font technology on its own might not have been that valuable in this configuration, since it was merely a component in a larger system—as Ethernet originally was inside of Xerox.
Instead, selling font libraries to original equipment manufacturers (OEMs) allowed the font technology to capture significant value by leveraging the efforts of computer OEMs like Apple and IBM and printer OEMs like Canon and Hewlett-Packard (HP) to create a new value network around desktop publishing. Adobe occupied a single important piece of this value chain, focusing on supplying the digital font libraries to laser printer and software manufacturers. As the manufacturers of PCs, printers, and software made faster and more powerful products, Adobe’s position became increasingly valuable.
This very different approach to commercializing its technology also made Adobe a valuable company. Adobe Systems went on to become a public company in 1987 and continued to operate as an independent company in 2002. At the end of 2001, its market value was approximately equal to that of Xerox.
As with 3Com, the business model that eventually created significant economic value out of PostScript for Adobe differed greatly from the Xerox business model. Had Adobe persisted with its initial intentions, which had strong similarities to Xerox’s model, that latent value might never have materialized.
Metaphor: A Xerox Spin-Off with a Xerox Business Model
3Com and Adobe created value from Xerox technologies only after they transformed their business models substantially from the one that Xerox usually employed. In contrast, the founders of Metaphor commercialized some promising user interface and database query concepts developed at Xerox PARC through a business model quite similar to theone at Xerox. Metaphor is thus an important contrasting case of how effective Xerox would have been if it had pursued its technologies further through its own business model.
Metaphor was created by David Liddle and Donald Massaro in 1982. It developed a series of technologies that allowed nontechnical users to create sophisticated queries of large databases. This enabled a new group of users to mine corporate data for a variety of new purposes, such as market research, pricing analyses, or analyzing trade-offs between possible new product features. Before, users would have to rely on corporate programmers to write report generators to extract data from a mainframe to get the data they needed. Because the programmers had many projects to perform for mainframe users, these requests typically landed in a large queue. Users were frustrated by the long lead time it took to get the requisite mainframe data they needed to do their jobs, and the technical programming required to generate the data was too arcane for them to access the data directly. Metaphor’s technology let knowledge workers utilize a point-and-click graphical user interface to construct their own database queries directly to the corporate databank. The ability to extract useful corporate data directly was a potentially powerful value proposition. The technology would allow users to bypass the report-generation programming queue, would create faster access to data, and would empower the users with the ability to experiment with new combinations of data. It was one of the first true client-server applications, employing the graphical user interface technology out of PARC to construct previously arcane and complex database queries in an intuitive fashion.
Metaphor’s ambitious technical approach was accompanied by a business model that would have been familiar to Xerox. This included the development of a proprietary software product and the sale of that software bundled in with proprietary hardware as a turnkey solution for the customers. Metaphor intended to reach customers through its own direct sales force. As with Xerox’s business model, Metaphor had a strong systems approach to commercializing its technology and a similar approach toward proprietary technology. Essentially, it built an internal value chain and eschewed an external value network. Liddle defended this approach as the only viable means at the time to implement the company’s product strategy: “The problem wasn’t one of a business model. When we started Metaphor, standards weren’t available and the only choice was to do the entire system—that’s the way every body did it then. It’s not like today. What’s more, this kind of product couldn’t be sold at a retail level. The only way to sell it was with a knowledgeable sales force. . . . There was no packaged software at the time; we had to make our own equipment.”23
While Liddle’s defense seems plausible, many aspects of Metaphor’s circumstances appear to be similar to those facing Adobe. In 1983, when Warnock and Geschke left PARC (a year after Liddle and Massaro left), there were no standards for fonts or for generating computer characters mathematically on laser printers, either. Nor was there an obvious way to distribute such a product. And, as noted previously, Adobe’s initial plans were to develop the entire system as well. Its value network had to be constructed de novo. Warnock and Geschke believe that, in hindsight, they would not have succeeded had they continued with their initial business plan. They were also aware of Metaphor’s situation and felt that Metaphor employed this approach as a direct result of their experience in Xerox. In the words of John Warnock, “Metaphor took the Xerox business model with them.”24
This probably was a mistake. Despite its innovative technology and its potentially powerful value proposition, Metaphor was not one of the great commercial successes spun out of PARC. The company did manage to survive from 1982 until its sale to IBM in 1991, but its financial performance was meager, and it burned through a great deal of venture capital. Although the amount that IBM paid in 1991 was confidential, it did not reach the amount of capital cumulatively invested in the company. While there are undoubtedly many explanations for Metaphor’s performance, its failure to explore alternatives to the Xerox business model stands as one plausible explanation—particularly in comparison with the value network that Adobe erected for its font technology. Metaphor’s lack of success does not seem to reflect the limitations of its technology; rather, its disappointing fate lay in its inability to find the model that would unlock the latent value embedded in that technology.
Implications of the Business Model for Open Innovation
Chapter 3 argued that firms that wish to employ an Open Innovation approach need an architecture to integrate internal and external technologies and to fill in the missing pieces. The analysis in this chapter shows that this architecture extends far beyond the traditional boundaries of technical management to encompass marketing, sales, support,and even finance. The customer segment chosen and the value proposition offered have important ramifications for the particular attributes of a technology being developed. The value chain that is constructed around the offering determines the value being created and the ability of the firm to claim a portion of that value for the firm. The resulting margin structure casts a long shadow over future initiatives, which are judged in part on whether they can continue or enhance these margins.
These issues imply that R&D managers must play an important role in the development and execution of the business model. As John Seely Brown noted in the introduction to this chapter, these managers must regard “the architecture of the revenues” as a vital element of capturing value from technology. These issues also imply that R&D managers cannot abdicate their part of the responsibility for crafting an effective business model. Just as the business model itself must span the technical and economic domains, so must technical and business managers themselves reach outside their areas of responsibility to work toward an effective model.
Technology managers need to include experiments in alternative business models. This is as important as the experiments they conduct inside their labs to evaluate technical risks. While it is certainly valid to consider making all the elements of the value chain to deliver a new innovation internally, it is equally valid to explore the possibilities of focusing on one or more pieces of that chain, and possibly utilizing external elements for the rest of the chain. This will also require technology managers to create processes to explore the social domain far more thoroughly, from customers to third parties, and the surrounding elements of the value network. It is vital for business managers to create mechanisms to expose technologies to external companies and to imbue technology developers with greater understanding and empathy for the social context in which their ideas will ultimately be applied.
