Lean by Design

I want to come back to an overarching theme for this blog, one that I have alluded to when writing on changing business models in Telecom, but one that warrants further explanation – that is; the disaggregation of the value chain in Telecom.

Telecom has lagged the computing / IT industry in some of the major electronic technology trends through out its history. Heck, the big vendors were still designing and building electromechanical switches in North America as recently as the early 1970′s, long after mechanical tabulators had given way to stored program electronic computers. And while the computing industry was revolutionized through the work of hobbyists – developing what later became known as personal computers in the mid 1970′s – a similar phenomena did not occur in the telecommunications field until very recently with the advent of open source PBXs and associated User Groups and home brew clubs.

Similarly, the move to specialize the value chain in Telecom was slow in coming. When I began my career in Telecom, (over 20 years ago) we were still designing and building virtually everything that made up a communications platform – right down to the screws. It was a vertically integrated business, with all the benefits and pitfalls that come with that kind of structure. In the United States, the Service Providers thru their common standards and research arm (Bellcore) were trying to pry a greater percentage of the value stack (and control) away from the dominant suppliers, through standardizing new network architectures that shifted some network control functionality away from the vendors towards the operators. This effort – embodied by their IAN, AIN, and IN architecture initiatives – proved to be very costly and arguably, ineffective, for the Service Providers. In parallel, certain economic influences were beginning to have a greater impact. First, processor technology was becoming much more complicated and specialized companies, including Intel and the predecessor of Freescale, were increasing the performance gap with product components produced by vertically integrated companies like Lucent and Nortel. Memory, and then processor technologies were the first to be replaced with third party components. From there, successive levels of integrated technologies fell to companies that created defecto standards through their focused niche dominance. After processors fell to this trend, it was only a matter of time before embedded system companies arose to supply the boards, chassis and integrated platforms based on increasingly higher order standards, like CompactPCI, and then ATCA. Nokia has been a leader in recognizing the business value of this trend, and has outsourced larger and larger components of the hardware value chain (both design, production and support). ATCA hardware platforms are now being procured from embedded system suppliers and contract manufacturers, representing a huge component of the hardware value chain. For the most part, hardware design and manufacturing is now supplied by contract manufactures and ODMs.

Revenue is, however, dominated by network edge products (in wireless, approximately 70% of revenue comes from base stations), and this domain has trailed the core network in terms of disaggregation. It remains the last bastion of an integrated value chain for large Telecom vendors, and the large vendors have jealously guarded it. A few years ago, Nokia attempted to rally the industry behind a base station standard – called; Open Base Station Architecture Initiative (OBSAI). The intent was to desegregate the value chain, thru standardizing base station components and interfaces, and in the process, enable greater performance, lower R&D costs, and improved time-to-market for new products. Wireless base station development had become very costly and complex with rapidly changing technology. Only the very largest players could manage the R&D costs. The ecosystem resulting from OBSAI would introduce greater competition, improved costs, and enable companies to specialize on a given technology niche. We can only speculate as to all of Nokia‘s motives, but it seems likely that this also represented a way of unseating the industry leader – Ericsson – by changing the game underneath them.

The initiative has been slow to take root, primarily because of a counter effort by a group of vendors around another so-called standard architecture, called; Common Public Radio Interface (CPRI). One that was far less open, and would have the effect of largely maintaining the status quo. I suspect that another reason for its slow uptake, may have been a realization, that the advent of a new and more open wireless technology – WiMAX – was likely going to have the same desegregating effect, if it became successful.

Key elements of product disaggregation are:

1. Components
2. Enabling Technology (Tools)
3. Low volume hardware subsystems
4. High-volume (largely access) hardware subsystems
5. Network Services and Facilities Software
6. OSS software
7. Communication Applications

Of course, there is another dimension of the value chain that could be considered. Manufacturing outsourcing, and outsourcing R&D further complicate matters, as they are obvious forms of disaggregation, and also enable further disaggregation of the value stack. Absent of any unique and valuable proprietary technology, once manufacturing of a subsystem has been outsourced, it is only a matter of time before that subsystem will be designed and eventually marketed by a someone else. Today, despite claims by vendors to the contrary, there is no proprietary system level technology that carries significant value for their customers. To understand this fact, one only needs to examine the procurement behaviour of their customers.

Elaborating on the elements of disaggregation – we find the following:

1. Components: Almost all components of the value chain have been externalized. Starting with powerful general purpose processors, now even highly specialized elements, such as RF chipsets, of the value chain are being ‘disaggregated’.
2. Enabling Technology (Tools):  Virtually all enabling technologies have been externalized (the sign of a mature industry). Twenty years ago, companies developed and maintained their own email, and customer support solutions, and in some cases their own software library systems, language and Operating Systems. Today, all these technologies have matured to the point that specialized companies offer ‘sufficient’ solutions, and superior solutions do not provide sufficient differentiation to justify the incremental cost and distraction.
3. Low volume hardware subsystems:  In Telecom today, virtually all the low and medium volume hardware subsystems are provided by general purpose computing platforms (blade servers), or about to be met through ATCA platforms. Again, Nokia has often provided leadership in regards to disaggregating their value chain.  It is instructive to know that Nokia sourced a preATCA platform for their server based products, called; Flexiserver, but a year ago made the commitment to move to an ATCA platform. Today these platforms are provided by smaller embedded systems suppliers, but given the size of this market and the relatively small number of large customers, we can expect to see a consolidation of vendors, either thru M&A activity, or as a result of the large contract manufacturers entering this space.
4. High-volume (largely access) hardware subsystems:  The majority of innovation in the ‘facilities’ portion of the Telecom business, has occurred in the access domain. Whether it be the rapidly changing modulation and encoding technologies of broadband wireless, or xDL technologies, this remains an important area for the large Telecom vendors, and one where they are still very competitive. It is also an area that is witnessing growth in start-ups, primarily focused on component technologies.
5. Network Services and Facilities Software:  Although a fairly small domain when compared to the others, especially on a revenue basis, it is important to note, that managing network services, or in other words, conditioning the hardware, is closely tied to the hardware, and network integration. It is a necessary element, but one that is difficult to separate from the hardware. The control software component is embedded in the OSS software component and is governed by standards.
6. OSS software:  Over the last 10 years or so, OSS software has largely become the domain of specialty companies occupying a key relationship with the Service Providers and large corporations.
7. Communication Applications:  Ok, now we finally get to the growing battleground of communication services. Blossoming of the internet a dozen or so years ago, followed the introduction of commercially available Web Browsers, and touched off a battle that is still being fought over the nature of the future network, and its implications for existing service providers and vendors. In the early years it was characterized as a battle between the Netheads vs. Bellheads, with the Netheads arguing for a stupid network with smart applications at the network edge. More recently, these same opposing forces have touched off a firestorm of debate over so-called Net Neutrality. This is arguably the most important potential disaggregation of all. Who will be supplying, and profiting from communication services? It has yet to be fully decided, but the evidence to date does not favour the traditional telecom vendors and Service Providers.

There are two forces that are handy capping traditional service providers in this battle. They are:

• The earnings model associated with voice has largely broken down. The internet has disaggregated the communication service model, effectively separating connection capabilities from services, including voice. This in turn has invited a dramatic increase in competition as it is no longer necessary to own expensive connection facilities in order to offer the voice service.
• Service providers have traditionally operated a ‘closed garden’ approach to offering services, where services are provided within the core of the network, owned by the Service Provider – this is essentially the IMS model. They have a large investment in the infrastructure as well as the culture of providing high quality services in a controlled fashion. Limited access has ensured oligopolistic practices and profits in a relatively stable domain. The internet and services like Skype, and others, have shown that services can be rapidly and effectively provided with sufficient quality from the edge of the network, to compete disruptively with Service Provider offers.

Traditional Telecom Vendors are faced with similar challenges. Specifically, having surrendered (for good reasons) much of the lower parts of the value chain, the obvious conclusion, is, and has been, to head for the high ground of Services and Applications. Unfortunately for them, the very disaggregation that they have actively participated in, has contributed to edge applications becoming more practical and appealing, thus inviting additional competition. Like, the traditional service providers, these vendors have operated in a vertically integrated environments that facilitated oligopolistic practices. They are not well positioned to compete with more agile companies that have typically developed in highly competitive software businesses.

These challenges do not necessarily spell doom and gloom for traditional Telecom companies. It does, however, mean that these companies must take a sober second look at the new environment that they are operating in, and coolly assess their prospects of success under different business models. As an example, I have suggested that service providers are not well positioned to develop services/applications in an open and highly competitive environment that appears to be evolving, but they are in an excellent position to provide the high speed pipes and connectivity facilities to enable these applications. As Martin Geddes has claimed in his Telco 2.0 Manifesto (and I agree), utility businesses can be very profitable, and could be more so with the appropriate focus. So, contrarian as it may sound, providing a super duper ‘dumb’ pipe may be a great root for today’s service providers to follow. Similarly, the traditional Telecom vendors should focus on the portion of the value chain where they bring the most competitive advantage, and have the fewest handicaps. One possibility is to focus on facilities and related technology, and less on communication applications. If they determine that ‘applications’ is their business, then they better aggressively undergo the necessary changes to their business. Hang on, because it is going to be a violent ride!