Case Studies and Speculation on Strategic Control
Here, I will try to demonstrate the usefulness of this conceptual distinction between architectures and platforms by using proof by examples taken from the gorillas's Qualcomm and Microsoft. For the moment, please set aside the similarities in the terms architecture and platform, and contrast the idea of an architecture with the concept of a platform. If you do so, you can imagines instances of architecture that controls the value chain initially that may or may not give birth to a platform that becomes the means for sustaining strategic control of a valuable standards-based franchise.
Consider the Case of Qualcomm. Within the telephone industry, a disruptive technology with partially overlapping and partially different trajectories of performance, wireless telephony, challenged the traditional wireline paradigm by introducing a discontinuous innovation: mobile voice communication. This wireless network tied into the established public switched telephone network, which greatly increased the value of the wireless network as it somewhat increased the value of the PSTN. A new but only partly discontinuous technology was born: wireless telephone connectivity. By partly discontinuous I mean that wireless was a discontinuous innovation, but the tie in of the wireless architecture to the wireline architecture made it partly continuous as well. Of course, this model may prove to be disruptive of wireline (at least, where it is not in place) because wireless is not only a complementor to wireline, adding minutes of use and customers to the PSTN, but also a potential competitive substitute for wireline (for example, in the Leap Wireless business model).
Although, wireless telephony was not nearly as far advanced along the traditional trajectories of progress that it shared with wireline, lacking qualities like clear voice quality, reliability and services like 911, call waiting, and so on, its customers' high priority on mobile access availability led them to accept less functionality, lower reliability, and fewer features at a higher cost just to possess one specific feature of the new technology: mobility is what they wanted and would pay for.
The first generation of wireless architecture, called AMPS, was very close to traditional radio frequency analog architecture, except, of course, for its tie in with PSTN. This form of cellular architecture used FDMA architecture, which limited the number of users in a cell to one call per each analog frequency within the assigned band.
The second generation of architectures introduced improvements in the number of users based on incorporating digital technology. Within digital technology, two major architectures developed: TDMA and CDMA. The TDMA architecture increased the number of users by dividing time into millisecond (?) units, multiplexing many calls together for transmission, and then demultiplexing them back into unified-by-time digital messages. Using TDMA technology, a European version, GSM, specified different communication standards than those chosen in the American or Japanese versions of TDMA technology, which used different frequency bands and different standards for coding TDMA technology into transmission-reception architecture.
Using its proprietary but open second-generation digital architecture, Qualcomm helped launch initial CDMA services in Hong Kong, soon followed by Korea as a national wireless standard, and, subsequently, in the U.S. with the carrier Sprint. Although Qualcomm did not invent CDMA technology, it developed significant features of CDMA technology for its wireless architecture. These features were both patented and selected (in the absence of significant competition) as the de facto 2G CDMA standard architecture. Qualcomm's CDMA architecture had specific competitive advantages following from the use of a spread spectrum with a coded signal to permit more users to multiply access the system. Not only does this CDMA architecture continually evolve in terms of improved performance in the traditional CDMA trajectories of performance, improved power control, rake receivers, and soft-handoffs, variable vocoding, and the like, but also Qualcomm's CDMA platform agilely adapts to opportunities for more profits by including position location, multi-media applications, multimode and multiband functionality, and BREW's simplified operating system software in its enabling ASICs for base stations and handheld products.
Each of these additions to its platform is based on architecture. Only some of these architectures are proprietary to Qualcomm, and others are not. For example, when QCOM purchased SnapTrack, it bought their patented IP and know-how. This expanded CDMA ASIC platform along the different trajectories of performance of this new technology, position location. Also, it added a new source of revenue and and possibilities for continued product improvements. Thus, a new proprietary open architecture for position location was added and integrated with the initial CDMA architecture within Qualcomm's platform.
The addition of multimedia functionality was based on several non-proprietary communication standards, each of which has added new trajectories of technological progress, which are specific to each medium. Qualcomm will attempt to make its instantiation of, say, MP3, or other open non-proprietary multimedia standards, into its own de facto standard by including its functionality within its platform of CDMA ASICs. When a customer purchases a Qualcomm ASIC with these features, it adds to the value of the CDMA franchise. If these multimedia products generate a significant network effect in sales, Qualcomm might make its MP3 instantiation into a de facto standard for CDMA mobile devices because it is nested within its CDMA architecture, which is nested within its ever expanding CDMA platform.
To build its multimode chips Qualcomm was required to license patented aspects of GSM's architecture from Nokia (although this standard is multi-proprietary and defined by the ETSI consortium, or it may just have avoided threatened lawsuits). This negotiation with Nokia, who is the most powerful competitor and member of ESTI, to access rights to use GSM was part of the exchange for giving Nokia, as a handset OEM, a license to use Qualcomm's CDMA 3G technology. However, Nokia pays Qualcomm for the use of its proprietary standard, whereas QCOM does not pay Nokia for its patents on aspects of the GSM standard, so QCOM has the strategic control. My point is that this TDMA architecture will be included soon within QCOM's CDMA platform.
Multiband capability is based on Qualcomm's proprietary RadioOne technology, which offers new RF functionality while it also reduces the number of necessary components in the handset. Both the new RF archtecture and the reduced-component architectural advances will be included in the future, more inclusive Qualcomm platform for multimode CDMA, WCDMA, GPRS, GSM, and other earlier architectures that operate across a diverse bands of radio frequencies. Note that multimode and multiband functioning, all with architectural implementations, are nested in Qualcomm's platform. This means that both CDMA 2000 and WCDMA will become part of QCOM's platform. This possibly makes both architectural versions potential QCOM de facto standards. This is so because, as yet, no one else has demonstrated Qualcomm's engineering ability in producing 2G or 3G chips, much less in handling the difficult asynchronous problem in using the WCDMA standard. Because Qualcomm controls both its extensible platform and the technological performance of CDMA architecture, its strategic control has two bases of support: technical and strategic. Qualcomm is developing the ability to use its ASICs to solve interoperability problems across bands and modes in different regions of the world, if it is successful in leveraging the power in the technology and of the market, it may dominate the 3G market.
Qualcomm's platform continued to expand by the addition of BREW as a simple telephone operating system. This addition to the platform permits application developers to forgo the necessity of porting their software to each OEM's CDMA telephone, creating the advantages of a" write once, use in all CDMA handsets."
Thus Qualcomm's penchant for creating new technologocal innovations radically expands the scope and life of its mobile connectivity platform along new trajectories of performance; just as Moore's law drive the continual product upgrades that come from the continual normal evolutionary advances in CDMA or, say, position location or RadioONE technologies's trajectories of progress.
Ideas are powerful tools that permit the birth, combination, and advance of original architectures . Qualcomm has two distinctive discontinuous innovations that led to two original architectures. Each breakthrough generated strong sources of competitive advantage. The first was the solution to the near-far problem that led many so-called experts to say that CDMA violated the laws of physics. If you have many equally amplified signals, strength of the near signals mask those signals that are far away. I presume it was QCOM's Richard Viterbi's who had the key theoretical insight that amplification that varied proportional to distance from the base station could solve this previously unsolvable problem. This insight, when tied to superior engineering to produce the necessary engineering algorithms, gave birth to workable CDMA architecture. The advance in DSP technology along its established trajectory made this inventive solution workable, creating what can be called the first CDMA architecture.
This CDMA architecture handled data far better than TDMA architecture, leading the ITU to make CDMA into a worldwide 3G standard because of the growing importance of data. Viterbi's second insight was that voice and data had different properties, with data being more bursty. The CDMA channel had been set to yield a reasonable compromise between both types of digitize input. However, Viterbi knew that two separate channels would permit the variable optimization for either voice or data. Unlike WCDMA's 5MHz band, that required new and extremely expensive bandwidth allocations, the Qualcomm architectural decision to use 1.25MHz channels permitted an architecture that can dynamically shift between voice and data use throughout the day as demand changes. This discontinuous architectural innovation greatly increases the value of the present CDMA 2000 platform in comparison to the promised WCDMA platform.
Qualcomm's superior scientific and technological knowledge of CDMA puts them firmly in charge of inventing and controlling the crucial architectural trajectories of performance for CDMA. No other company has demonstrated the technical ability to make sound and sustained evolutionary advances along established trajectories of performance, much less to make inspired discontinuous or disruptive architectural shifts. Qualcomm not only innovatively created the original CDMA wirelss arachitecture, it also, with HDR, has produced a second new set of technology trajectories for dynamic use of independent channels for voice and data.
This technical control of Qualcomm's CDMA architecture has made it inevitable that Qualcomm's instantiation of CDMA architecture has become the de facto standard in 2G and threatens to dominate 3G because of its ever broadening and long-lived third generation platform for wireless mobile communication. Qualcomm has retained full intellectual property rights to its inventions, and it will be well rewarded for them.
My key point here is that it is Qualcomm's platform that provides strategic control as the de facto standard for the implementation of it proprietary open CDMA architecture.
The CDMA architecture, with its proprietary open power control, rake receivers, soft handoffs, plus its newer variable vocoder , and the like, was the origin of its broader platform for 3G that also contains it proprietary open position location, its propriety open BREW, its proprietary open RadioONE, its proprietary open HDR, and the like, and so on into a promising future of additional sustaining and disruptive advances.
Sustainable strategic control requires an open-ended platform that runs ahead of the competition and translates technological advances into architectures and performance trajectories that can be translated into up-graded and continually up-gradable products that improve performance, reliability, and ease of use within architecture and across multiple architectures of a platform. When any one of these differentiated competitive advantages aligns with customers' priorities, it can create network effects and increasing returns that generate high switching cost and produce hypergrowth in sales.
I conclude that Qualcomm's CDMA architecture created differentiated competitive advantage and that Qualcomm' CDMA platform both deepens and extends the breadth its sustainable competitive advantage.
[Now let me shift from analysis to theory. Here I speculate about the meaning and significance of the star node in a value web, slightly recasting and augmenting Moore's suggestions.]
Qualcomm has sustainable strategic control because it is the star node in its value web. In the landscape of the mobile communication business a value web includes a traditional value chain plus a value net of competitors and complementors. The star node, as I am defining it, is the nexus of communication, concern, and power. It is analogous to being a sociometric star along several axes. In a sociometry, you ask people to nominate individuals who fit a descriptor and you plot a graph of single and mutual selections. Say, you ask the group to nominate the person who is most liked, or who is seen as most powerful, or to select someone as leader, and the like. If a single individual emerges as the center of affiliation, achievment, or power, then she is a sociometric star. By analogy, Qualcomm is the star node in its value web. It is the star node because its platform draws the companies in its business landscape into its orbit of strategic control, both complementors and competitors, both value-delivery partners and customers as partners. Even its competitors must follow its lead or attempt to counter them. The star node is the center of creative power, whose strategic power is magnified because it includes the power of its satellites operating as a value web.
As the star node, Qualcomm is not just the de facto standard but instead is the de facto standard setter. It has increased in power as it moved from architectural leader to platform leader. Qualcomm has strategic control because it owns the standard, controls the value chain, and creates the next big thing while everyone else waits for its moves to organize the advance of its platform, which is the stabilizing point of connection for members of the value web. The point of origin of the value proposition of each member of the value web resides in the star node. Qualcomm used its create-and-set-the-standards business design to become the standards-based-standard-setting star node in an ever-expanding and long-lived value web.
So, I argue that architectures become de facto standards and that platforms become de facto standard-setters. If I have not yet convinced you, lets look at Microsoft once again.
Microsoft is the prototypical gorilla. We have seen that it has a standards-based business design. Let's look at the evolution of its platform across multiple architectures, drawing once again on Slywotzky and Morrison (p. 269): [Unfortunately, I cannot format this table so that it translates online without jumbling my effort to format. So, I must improvise.]
Microsoft Business Design Reinvention: In 1975, the era of PC BASIC, MSFT selected programmers as its customer; captured value with software sales; differentiated (strategic control) by creating a standard, whose scope was language. In 1981, the era of MS DOS, MSFT selected programmers and PC manufacturers as its customers; captured value with software sales and its installed base; differentiated (strategic control) by creating a standard, whose scope was desktop software. In 1990, the era of Windows, MSFT selected programmers, PC manufacturers, and Mac and PC users as its customers; captured value using its installed base; differentiated (strategic control) by creating a standard and with application that used the full capacity of its operating system, whose scope was desktop software and bundled applications. In 1995, the era of NT and Explorer, MSFT selected programmers, PC manufacturers, and Mac,PC, and network users as its customers; captured value using its installed base and increasing return to scale; differentiated (strategic control) by creating a standard, brand, and the relationships with OEMs and developers; whose scope was desktop software, bundled software, communication, network, and exchange.
In 1975, Microsoft began it ascent by getting a contract from MITS' then it embraced BASIC, extending it into PC BASIC. By selling it software widely to programmers, it created the de facto standard programming language for the new PC. In 1981, Microsoft acquires QDOS, preinstalling it through multiple PC Manufacturers as an embraced and extended--MS DOS. Because it had embraced CP/M, it was able to enlist former CP/M application programmers as DOS ISVs for new applications. DOS became the de facto desktop operating system. In 1990, Microsoft launched Windows and successfully made it the de facto GUI standard for the desktop. Microsoft began to harvest excess returns as it extended its value capture of an installed base and broadened customer selection to include differentiated desktop applications by creating versions tailored to the operating systems and bundling them into a low-priced Office suite. In 1995, Microsoft continued its platform development, extending it from the desktop into enterprise computing with NT and Exchange, attacking Netscape in the Internet market with Explorer, and seeking to use a scaled-down OS, Windows CE, to attack the consumer market.
Microsoft has used multiple architectures within its Windows platform. It has branded the name "Windows" as the name of its platform although its architectures had been variously called PC Basic or MS DOS before Windows. Both new applications and variants of the OS include the platform name, Windows, as part of its name. More recently, Microsoft is using "X" as an equivalent or substitute for "Windows" when designating platform additions. So, we have X.net, the X-box, and XP coming to the fore.
According to Slywotzky and Morrison (p. 274), "A standard is something that drives an entire industry." Their "something" is, I argue, a platform of expanding standards for various architectures. As the classic standard-setting company it becomes the premier star node of the preeminent value web.
When you contrast Qualcomm and Microsoft, a major difference is apparent. Qualcomm created its innovative architecture; whereas, Microsoft's strategy was to adapt existing architectures to its business purposes by embracing and extending them. Also, Slywotzky and Morrison made a compelling case that Gates was driven by getting the customers and then aligning his priorities with theirs to create profits. In fact, Microsoft can be credited with inventing the create-the-standards-business-design, often called the Microsoft model.
Thus, The concept of architectural control does not do full justice to the extensive scope of Microsoft's power. Although I have been making a case for "platform" as more inclusive, central, and long-lived than "architectures," here, I mean something more than that, even more than being a star node in a value web that attracts the entire power of the web into the platform of the star node. It is as if strategic control itself follows the law of increasing returns.
In addition to owning the standard and controlling the value chain, giving Microsoft gorilla power, it also has multiple superdominant positions and the world's second best brand, giving it exceptional platform power. Superdominant positions indicate that increasing returns to scale have created a winner-take-all (or most) market. Increasing returns are not easily reversed. Microsoft's power is securely in place until a discontinuous innovation appears or a disruptive technology successfully attacks from below. (For example, disaggregating the computer into multiple single or restricted-purpose information appliances is one such disruptive threat.)
But also, this mindshare known as brand is a powerful element of Microsoft's strategic control. How does that work? Let's see. Perhaps we can follow the logic of the growing web of increasing strategic control to see where it lead: proprietary control of open architecture plus competition equals rapid advances in sustaining a discontinuous technology; open standards permits wide use within a significant and growing market; the value chain standardizes around a de facto standard to organize a preferred single solution as its whole product in a newly formed market; a network effect generated by solution adoption by pragmatist customers drives the tornado; hypergrowth validates the value chain's de facto standardization as the market tips strongly to its solution as the single winner; the single winner is locked-in by high switching costs coming from both the costs of switching and the loss of competitive advantages in the offering; the control of a proprietary open standard with high switching costs increases to another level when it includes the power of the whole value web; now the beginning architecture transforms into a broader and long-lived platform through both sustaining advances along normal technology trajectories of performance and the addition of new trajectories of progress by adding additional architectures under the scope of the platform; as the platform grows, the value web follows its lead as the stabilizing center of power and creator-adaptor of new standards that only it can set; the standard-setter become the star node of a value web; thus, the expanding platform of the star node generates increasing returns in profits and strategic control.
This ascension to star power creates strong emotions from customers, complementors, value web partners, and competitors. Who is more hated and feared than Bill Gates? Who is more avidly followed and lionized than Bill Gates? Who has more exhilarating and terrifying mindshare than Microsoft?
I believe that this is so, in part, because a platform becomes invested with the excitement and joy associated with its brand (or mindshare) and with the fear and anger associated with the gorilla's platform power. Let me try to define mindshare. Mindshare is the affect-invested (including both positive and negative affects) expectation that a specific platform is the only possible winner in technology competitions. The star node is idealized or demonized depending on the construction of its power as good or evil. Recognizing the existence of such mindshare is either thrilling or terrifying. Yes, Microsoft, but not Qualcomm as yet, does seem to have this kind of mindshare.
[I wonder, if you understand this strategy, could you use it to actively create a standard-setting business model for constructing a gorilla? If some manager could do that, could some investor reduce her risk of early entry into a gorilla-candidate by searching for companies who were deliberately following this strategy? But, I am getting carried away and must stop before I become unraveled. Wild speculation blows my mind. Does anyone know a good psychologist that I might discuss this with? vbg]
I hope this helps, if not please help me.
Don |
