I could not find a link to this article so in the fear of losing it i am posting it in whole
802.11 & Cellular: Competitor or Compliment? by Bill Gurley
“Same as it ever was… Same as it ever was…Same as it ever was…” -Talking Heads, Once in a Lifetime
One of today’s most popular debates is the potential long-term impact of the wireless Ethernet technology known as 802.11, or WiFi. Some suggest that the technology is over-hyped, but others continue to preach that this is not simply a new technology, but rather the next big thing – a “disruptive” technology that will change the world as we know it. One industry that is particularly attuned to this debate is the cellular industry. While some analysts have suggested that 802.11 is competitive with cellular data offerings, such as 3G, the currently trendy perspective – driven largely by the cellular industry itself – is that the two technologies are wonderfully complementary.
It is important to realize that the 802.11 vs. 3G issue is not black and white. WiFi will neither be completely competitive nor completely complimentary with data offerings from cellular carriers. There will likely be elements of both. So the real question is just how competitive and/or complimentary will WiFi be? And what will be the resulting balance of power in terms of both revenue and market capitalization? Consider the following analogy as a way to answer this all-important question: WiFi is to 3G as the personal computer was to the mainframe.
The first obvious similarity between the rise of 802.11 and that of the personal computer is the commitment of an entire industry to a single open-standard architecture. Just as the large box computer market had three or four leading stovepipe architectures (IBM, Tandem, DEC, etc), the cellular market has three or four competing technology architectures (GSM, W-CDMA, CDMA-2000, IDEN, etc). Meanwhile, 802.11 is an interoperable standard that is embraced by the entire technology industry. Cisco, Dell, Intel, IBM, Toshiba, Microsoft, Broadcom, Texas Instruments, Intersil, Agere, Hewlett-Packard, Proxim and numerous other public and private companies all support this single, open-standard, interoperable platform.
From a market dominance standpoint, open-standard technologies have many advantages over more closely controlled architectures. First, the presence of many competing players with interoperable products allows the consumer to benefit from the aggregated R&D dollars of all participants implying accelerated product improvement. Also, with aggregated R&D dollars chasing the same core technology, Moore’s Law is sure to shine. As a result you receive the dual benefit of rapidly advancing innovation AND rapidly declining costs. Ironically, falling costs lead to an ever-expanding market as lower prices attract more users and make more applications feasible. The expanding market attracts more suppliers, which results in more R&D dollars and the entire cycle starts over again.
In the early days of the PC, most people considered the initial market for the IBM/Intel/Microsoft-based personal computer to be fairly narrow. In 1980, no one envisioned that one day you might run your company’s entire ERP system, or power a massive array of web sites on this technical architecture. However, the “increasing return” forces highlighted above eventually created a product that, from both a feature perspective and cost perspective, was appealing to a much broader array of applications and uses than ever previously envisioned.
This exact thing is currently happening with 802.11. This tiny, and increasingly inexpensive radio is already shockingly versatile. The same $30 radio can be used to serve wireless connectivity in your office, connect both your PCs and your multimedia in your home, and provide coverage to a police force across an entire downtown area. Add a Pringles can as a directional antenna (no kidding!), and this $30 radio is capable of providing high-speed, line-of-sight connectivity at a distance of 10 miles. In fact, the majority of the volume in the line-of-sight fixed wireless market has shifted almost entirely to low-cost 802.11 radios.
Originally designed to connect PCs and handheld computers to the network, people are now using these low cost radios for an increasing number of diverse customer applications. In remote areas, 802.11 can be used as a DSL alternative. Hospitals are using 802.11 not just to connect doctors and nurses but also to connect hospital equipment. 802.11 is a wonderful solution for industrial sensors in remote monitoring as well as control equipment designed to intelligently manage such things as power and HVAC systems. You can even use 802.11 to build an extremely cost effective video surveillance network. It is also feasible that 802.11 will be the driving force behind a wholesale conversion to VOIP and an entirely new phone structure in the enterprise.
The invention of the microprocessor, which eventually powered the personal computer, led to an “order of magnitude” leap in price performance with respect to computing. And with the help of Moore’s Law and the open market dynamic outlined above, this advantage has only increased over time. By 1990, a single Intel microprocessor had the processing power of a mainframe built just four years earlier – and obviously at a fraction of the cost. While people have tried to build faster, richer, better computing architectures since the birth of the PC, all have fallen short due to the seemingly “unfair” advantage of the open-standard technology.
Companies deploying 802.11 systems are finding a similar experience. In the case of fixed wireless space, companies are replacing proprietary closed spectrum radios with 802.11 systems that are easily ten times cheaper than that of their predecessors. With respect to broad-based wireless data coverage, a similar case can be made. Most 2.5/3G system providers are currently pricing high-speed data services at somewhere between $1-$4/megabyte (despite the fact that some experts put the opportunity costs of the voice capacity of this transmission at $1.30/megabyte). At this price, a one-time daily synch of your email inbox could cost as much as $100. Hook a $600 T1 to an 802.11 node, and assume only 10% usage, and you will find you can deliver high-speed data service for as little as $0.01/megabyte. The numbers are even better for a $50/month 384k DSL line.
The economics of 802.11 are advantageous in two other critical and material ways. First, as 802.11 operates in an open, unlicensed spectrum, providers pay no fees for spectrum licenses. Proprietary licenses obviously cost billions and billions of dollars and are a key component of the excessive cellular cost structure. Second, and equally important, the price of 802.11 customer equipment is rapidly approaching zero. This compares with a number as high as $400-$500 for a cell phone that exploits a new technical breakthrough. So, why is WiFi almost free? The answer is (1) because the price of the parts could go as low as $5 over the next five years, and (2) because the entire PC industry has a vested interest in seeing 802.11 dominate over 3G. As such, within a matter of years, 802.11 will be a standard feature on every single portable computer, with a marginal cost to the consumer of $0.00.
The last key similarity between the personal computer and 802.11 is denial and lack of respect. All along the way, large system proponents would argue that the PC was a “toy” that would never be capable of real-world computing. Likewise, the large cellular providers argue that 802.11 is fine for delivering service to a coffee shop, but will never be a true alternative to broad based city-wide coverage. This is despite the fact that Metricom has already proven you can build a high-speed citywide network leveraging open spectrum radios. They should also know that the 802.11 market doesn’t need deep pocketed believers. Thousands and thousands of small ISPs are experimenting with a variety of 802.11 deployments. While service providers wait for solutions to high-end problems like roaming and billing integration, “cowboy” providers like San Francisco’s Surf ‘N Sip have already deployed at 170 locations.
So is 802.11 complimentary to cellular? Perhaps it is complimentary in the way that chips and salsa are complimentary to a Mexican entrée. They are not directly the same, but if you eat a ton of chips when you first sit down, you just aren’t as hungry for the Chile Relleno later. Prior to the emergence of WiFi, there were thousands and thousands of presentations given on why data services over cellular networks made sense. In those presentations, executives described scenarios and locations where users would interact with these valuable services. The problem is that the primary locations in these historic scenarios are the exact same locations where 802.11 hotspots are already deployed today (airports, hotels, coffee shops). 802.11 has stolen the heart of the market, leaving high speed cellular data services to fill the niche that Satellite phones fill in the voice market.
It is undeniable that the arrival and continued success of the personal computer did not obliterate the mainframe. It did, however, curtail the growth by “stealing” the incremental market opportunity. The combined market capitalization of IBM, Hewlett-Packard (Compaq, DEC, Tandem), and Sun is $201B (and this includes some PC value). The combined market capitalization of Intel, Microsoft, and Dell is $469B. So you tell me, was the PC competitive or complimentary to the mainframe?