Breakthrough Ideas (continued, with two [[significant additions]] set in double squared brackets)
The Evolution and Selection of A Winning Platform
The basic platform provides sustainable competitive advantage, but the continually expanding platform enlarges the value added to create nonstop, renewable, leveraged, and enduring competitive advantage. Architectures are organic: adapting to imperatives and evolving by selection.
A successful platform continually develops when management understands that a standard-based business design requires a dynamic path of adaptive change that can be described by a set of general imperatives. That is, the general imperatives are broadly defined, but common, all-purpose economic, technical, or strategic features or sub-goals that identify basic and indispensable guidelines to criteria of adaptive fitness. As such, they serve as a set of pragmatic guides that define excellence and fitness.
The general imperatives for a platform strategy are: increase value; power performance; expand functionality; deepen functioning; extend reach; generate generality; separate specialty; isolate layers; foster reliability; reduce complexity; simplify forms; appreciate aesthetics; standardize interfaces; selectively open proprietary standards; enable modularity; plug to play; set standards; subsume standards; seek compatibility; attract complementors; decompose to recompose; combine building blocks; interrelate architectures; optimize functions; upgrade products; commoditize complements; embrace change; cannibalize to avoid obsolescence; embrace R&D; inspire creativity; seek architectural breakthroughs; create beauty; value elegance; add value; increase scale; reduce costs; excel in quality; build brand; facilitate customer convenience; speed time-to-market; foster user-friendliness; offer complete solutions; integrate end-to-end; provide complete standardized solutions; reduce size; tightly integrate modules; and conserve power.
As a company develops adaptive fitness as defined by implementing these general imperatives, complementors in a value web select it to become a key member of a developing open system information solution. The inclusion of the minimum features necessary to offer an initial version of a generic whole product defines the membership of a value chain. Only then does the value chain turn its attention to discovering the generative source of power within its web.
Generative power is the potential to continue adding value to the value web’s economic proposition. The power to lead/manage/control the value web is the upshot, the end result of generative power, of adding the most value to the value web.
Generativity resides in a general-purpose architecture because of its capacity to both generate amplifying force and produce offspring. Imagine the idea of a “wheel” as a general-purpose architecture. From the water wheel and the millstone wheel to gears, pulleys, and rotary engines, the idea of a wheel generated force and produced innumerable offspring. In the same way, because of its generativity, the value web of suppliers, customers, and complementors selects a rapidly evolving architecture as providing an elegant solution, as its central technology interface, as crucial to its value proposition, as defining its standard-of-value, and, if truly generative, as its standard-setting star node.
The star node’s organic, rapidly growing, vitally enabling, pragmatically stabilizing, and predictably expanding platform of standard-setting architectures attracts and holds the value web in its economic orbit. The architectural building blocks in the platform gracefully and elegantly combine with the value web’s products and services to create an integrated whole, the complete system solutions that can enable a potential tornado of accelerating demand. The architecture of the star node’s platform configures the open system, accelerates its growth, defines the industry, and shapes its future.
Qualcomm is not only in the role of protagonist in its value web; it is specifically its star-node (a conceptual honorific that I introduced in the April 2002 issue of the RTW Report). The star-node supplies the crucial architectural interface in an open information system.
[[The architecture comprising an open system consists of a set of modules that are structurally independent but work together as a complex information system. According to Baldwin and Clark (2000, p. 63), “A module is a unit whose structural elements are powerfully connected among themselves and relatively weakly connected to elements in other units. Clearly there are degrees of connection, thus there are gradations of modularity.”
For Baldwin and Clark, the two key ideas in the concept of modularity are: (1) interdependence within and independence across modules; and (2) the breaking up of complexity through (a) abstraction, (b) information hiding, and (c) interfacing.
A master problem solving strategy of human beings when faced with complex problems (or complex systems) is to break the problem apart into more manageable units. Thus, Baldwin and Clark (p. 64) pointed out:
“A complex system can be managed by dividing it up into smaller pieces and looking at each one separately. When the complexity of one of the elements crosses a certain threshold, that complexity can be isolated by defining a separate abstraction that has a simple interface. The abstraction hides the complexity of the element; the interface indicates how the element interacts with the larger system.”]]
The interface of a star-node offers an architectural layer of abstraction that integrates and simplifies a set of fundamental general-purpose functions. Serving as the generative engine of growth, this architectural platform enhances the value web’s shared business proposition because of its fundamental capacity to evolve while keeping its interfaces constant. At the same time, the star-node’s power ultimately derives from its ability to add the most value to its value web and its industry’s economic proposition.
For instance, to enable multipurpose usefulness to customers, both Microsoft’s Windows and Intel’s PC microprocessor are not only crucial, fundamental, general-purpose platforms, but also serve as abstraction layers that facilitate the transparent augmentation of the technology system and value proposition by complementors in their value webs by simply connecting to the visible information contained in its interface. By providing this abstraction layer, these gorillas insulate their value web from architecturally higher-level complexities, thereby, simplifying the process by which the value web adds its additional and specific contributions to an open and evolving total system.
[[Similarly, Qualcomm’s crucial airlink interface was introduced through its commercialization of a spread spectrum set of design rules, parameters, and interfaces called CDMA. CDMA abstracts and hides the complexities of spread spectrum modulation as it provides a simple interface for connecting and interacting with complementors’ modules.
For example, when optimizing the power control of CSMs and MSMs to solve the near-far problem Qualcomm had to resolve a series of complexities produced by interacting physical variables. Interdependencies between variables can produce a cycling in which the tentative solution to one problem creates another, but the attempt to solve the second problem undoes the first solution, and so on. Discovering exactly how various interconnected physical parameters change in relationship to changes in one another’s changes, that is, unraveling the complexities of interdependencies among parameters, requires careful experimentation, problem solving, and selection of the best tradeoffs currently available. Engineers gain experience from solving such problems in previous designs: how a solution was found, what the ramification of that choice were, how the artifact performed, how to avoid many of the routes to failure, as well as what is likely to lead to success.
Design rules are designed to promote parameter independence. A set of incomplete or imprecise design rules permit unforeseen interdependencies to occur. This creates unexpected and unwanted cycling, requiring further iterations, consultations, and revisions. Only after accumulating sufficient theory- or experience-based knowledge, can an architect foresee, given that the requisite experience has been sufficient, the expectable problems and how to solve them.
To proceed toward increased modularity, an architect makes a set of decisions in the form of design rules that specify certain choices; he must enforce these rules, making them binding, before the rest of the engineering tasks get underway. As Baldwin and Clark (p. 68) put it, “Instead of the original interdependency, a hierarchical pattern emerges in which certain parameters are ‘privileged’—they affect other parameter choices but they themselves cannot be changed. These privileged parameters [like the CDMA chipping rate, the wave form, the bandwidth of the spectrum, or synchronous handoffs] are the design rules,” and (p. 70), “Imposing a design rule when one is ignorant of the true underlying interdependencies can lead to design failure.”
Following Baldwin and Clark, design rules establish the architecture of a system, its basic structure: what modules are in the system and what roles they are to serve. Parameters define the specific dimensions of the architectural structure. Interfaces provide detailed descriptions of how the different modules will interact, including how they must fit together, connect, and communicate. A set of integration protocols and testing standards allow system integrators to assemble and test the functioning of the system originally and, over time, the functioning of the various modules.
As part of their implementation of the principle of abstraction, the architects partition the design parameters into two categories: Visible or Hidden Information. The design rules that specify the parameters and interfaces are visible information for modules lower in the nested design hierarchy. The rules for interfaces govern the connection and interaction of modules with the crucial architectural module, crucial, not only economically, but also because its global design rules are higher in the design hierarchy, making its rules necessarily both visible and prior to the design rules and parameters in modules lower in the hierarchy. The structural independence of hidden modules permits their rapid evolution, changing through experimentation with altered parameters and new designs.
The implications of this hierarchical structuring of visible and hidden information, according to Baldwin and Clark (p. 76), are:
“Because they constrain later actions in every other part of the system, the parameters in the top level of the [design hierarchy] diagram need to be established first. Thus higher levels both ‘are visible to’ and ‘come before’ lower levels.
Finally by definition, changing visible information requires changing the parts of designs that ‘see’ that information either directly or indirectly. Therefore, changes at the top of the diagram will have far-reaching consequences, and are bound to be difficult and expensive. In other words, visible design choices are relatively irreversible. Conversely, changes at the bottom of the diagram are limited in scope, hence cheaper to implement, at least in comparison to visible change. Hidden choices are reversible.”
This implies that the architectural control gained by setting the global design rules for CDMA2000 spread spectrum is powerful indeed, because its choices were both scientifically sound, tested, and proven, as well as relatively irreversible. Unfortunately, the far from wisely selected design rules of WCDMA are also relatively irreversible, but are necessarily less precise, complete, and useful because they are more “design around Qualcomm’s IP” and less theoretical or experience-based. Because these design rules for WCDMA were not derived from GSM design rules, a costly swap-out of the infrastructure and handsets for the spread spectrum airlink is required.
To translate these ideas into gorilla-speak, to say that Qualcomm has a proprietary open architecture with high switching costs means that proprietary intellectual property resides in the design rules and selected parameters of CDMA, which is openly licensed to a value web of complementors who add their value by using the CDMA interface’s visible information to connect and communicate with the CDMA RF interface that provides mobile connectivity. The high switching costs for the set of companies comprising the value web result from their dependence on hierarchically higher visible information in the CDMA interface. Of course, not every bit of intellectual property information is visible; it is only made transparent to complementors through abstraction. The required information is selectively open in a way that permits design CDMA airlink networks, infrastructure, chipsets, handsets, and the like to be designed according to the design rules and specified parameters when it is licensed..
Although based on the visible information in an architectural model, the structural independence of the tight connections within hidden modules permits their rapid evolutionary change by changing parameters or experimenting with new designs. Hidden modules invite competition and independent experimentation along their own evolutionary trajectories. This is the commoditization of the value chain that is driven by competition within a rapidly changing product market (like consumer electronic digital cameras or camcorders, which can increase the number of pixels, the storage space, or running time, or adopt a new UI) that can be attached to the crucial air interface to expand their value proposition to include mobile connectivity and transmission of images.
Any change in CDMA’s architecture means a change in visible information that requires the value chain to realign their products and services to the new visible information of the improved interface. Thus, Qualcomm controls the architectural design rules, which, in turn, control the interfaces with the hidden modules that connect and communicate with it using the visible information in Qualcomm’s design rules and parameter specifications. High switching costs for complementors are a function of the cost of realigning a hidden module with the visible information of a new crucial architectural interface. This means that drastic and incompatible alterations sharply raise switching cost; for instance, the change between GSM and WCDMA was profound; whereas, as it underlines the powerful competitive advantage of the evolutionary continuity in Qualcomm’s transition from 2G to 3G CDMA. Another significant aspect of switching costs remains the difference between switching from a more valuable to a less valuable value proposition. Each augmentation in CDMA2000 that is not found in WCDMA represents a loss of value, and, hence, a switching cost.
The delta between the intrinsic values captured in the two sets of 3G design rules for CDMA and WCDMA project into a probable future. Imagining this projection reveals a sharp difference between two net present option values. What’s more, Qualcomm’s architectural advantage from its original spread spectrum breakthrough, taken together with the visible information in the evolutionary architectures of its next two breakthroughs, called 1xEV-DO and the World Phone, continues to construct and compound its advantaged architectural control. Also, Qualcomm expands that delta of intrinsic value by each augmentation of its platform, from position location, the Internet Launchpad, and BREW, to QChat and with every application that uses those modules and its crucial air interface.
What’s more, the net present value of Qualcomm’s intrinsic value was radically changed by its harmonization strategy of offering a complete standardized solution regardless of modes, bands, or generations. The world phone platform subsumed the WCDMA and GSM design rules into a nested architecture by including the subsumed modules or the necessary translator modules in a tightly integrated radioOne modular chipset, which Qualcomm controls by the virtues of both its patented and licensed intellectual property and its ability to rapidly advance and expand its platform in ways that its rivals, who have less experience with either CDMA or WCDMA spread spectrum, cannot match. Yes, based on its intellectual property embedded in its architecture’s design rules which only it controls, Qualcomm exemplifies a proprietary open architecture with high switching costs.]]
Knowledge can be defined as systematically or methodically ordered information. A particularly useful form of ordered information is advantaged information. By analogy to mechanical advantage, think of knowledge as leveraging information to get more lift from the effort. Advantaged information equals the ratio of power divided by simplicity. Consider how a scientific theory explains or predicts many phenomena (its power) by using relatively few principles (its simplicity). Knowledge incorporated into a few rules can explain many complex phenomena. (If you don’t yet believe this, see Dr. Stephen Wolfram’s (2002) tome, A New Kind of Science.) From this angle, knowledge becomes the beautiful or elegant use of advantaged information that is leveraged by the factor of its power to simplicity ratio. The power to create wealth lives in this idea.
Applying these ideas to a platform’s architectural control, either increasing its power (for instance, through amplifying its performance or expanding its functionality or optimizing its modulation for data transmission), or simplifying its architecture (for instance, by adding a new layer of abstraction or sculpting unnecessary elements from its architecture, as in, say ZIF) amplifies a platform’s information advantage.
Also, you can similarly think of seeking competitive advantage as the use of advantaged information to leverage organizational knowledge and capabilities to increase the ratio of the power of differentiation to the efficiency of cost reduction. Simply put, advantaged information is efficiently powerful. This is the power in breakthrough ideas that drive wealth creation.
Therefore, possessing architectural control not only means that you can keep changing the game each time you introduce elements that increase functionality or expand the platform, but also that you can change the game through simplification of the architecture, say, by deep craft integration, by creating layers of abstraction, or by carving away no longer necessary components. So, Qualcomm used the power of information-advantaged ideas in each of its three breakthrough ideas.
Thus, owning and controlling an economically-crucial-and-technologically-fundamental-and-abstract-layer of architecture explodes a scale-dynamic magnification of the power of the platform. This explosion in added value magnifies abstraction’s delta of simplification by both the uniqueness-power of the delta of differentiation and the efficiency-power of the delta of plunging costs. This multiplicative-scale-dynamic of abstraction-times-differentiation-times-cost-reduction produces an exponential and offscale increase in added value, with a corresponding elevation in the power of the star-node.
Mammoth potential power resides in the idea of abstraction itself. Abstraction insulates a lower layer from the complexities of modules higher in the architectural platform below them by making that complexity transparent through the visible information of its interface. For instance, when the BREW platform abstracts its set of complexities, it insulates application developers, who can ignore the details of telephony and the deployment and management of applications by following the rules for interfacing with BREW.
Interfaces permit communicating connections among elements across a system’s modular layers. Integration tightens the interoperability within a module. By integrating its RF and processing functions into an integrated circuit, Qualcomm both ensures swift and cross tested interoperability of, say, the multimedia functionality and standards with the hardware or software in the integrated ASSP itself. Connecting to the BREW interface opens up the potential of simplified and convenient integration to the application developer, who is shielded from all this complexity by simply writing his software solution for a hidden module, say, sending digital camera photos to friends and family, directly to the BREW API. Because BREW works transparently across devices and networks, the developer does not have to be concerned specifically with either how other system layers function or with the diversity inherent in standards-based devices or networks.
My claim is that a fundamental property of a star-node’s interface as a nexus of connection is that it is simultaneously conveniently abstract, which produces transparency for lower layers in the design hierarchy, and architecturally fundamental because, not only of its visible information, which tends to endure and ensure sustainable advantage, but also because of its generativity. The star-node owns and controls the architecture that sits astride both sustainable technology trajectories and the corresponding plunge off their cliff of costs. Originally, the value web chooses to connect to this general-purposes platform because it is economically crucial at a critical point in time. Eventually, the industry accepts the platform as the key interface in its future, as the standard-setter, because it offers stability simultaneously with rapid and predictable evolutionary growth that drives the industry’s profits.
The star-node prospers and sustains its prosperity by tightly integrating the expansion of the platform to the platform’s nuclear interface, the source of its generative power. This tight integration of new elements to the familiar platform tightens it grip on leadership/management/control. This friendly-dominant form of power ultimately remains based on its power-as-mastery of the architecture and its relational-power to nurture the value web as a set of customers who requires the best possible service by its leader.
Qualcomm was selected as the star-node by its value web because of its fitness, a fitness demonstrated by excelling on each of these dimensions: (a) providing the most added value now and in the foreseeable future, (b) promising the most potential for continually subsuming the expansion of new elements and for adding ever-increasing performance to the information/communication system, and (c) commercializing and tightly integrating an architectural interface that abstracts the general-purpose functions into a platform for distributing a complete standardized end-to-end solution that brings, conveys, and passes along most of the benefits of its breakthrough ideas to its value web, industry, and end-users. |
