Williams Communications - Second time around: Part 2
January 30 2002

Interview with Matt Bross, Senior Vice President and General Manager of Emerging Markets.

This is part two of an interview with Williams Communications. Part 1, with Howard Janzen, Chairman, President and CEO, was published on October 15th 2001.

Optical Keyhole conducts interviews on the basis of readership interest only. They are not paid for by the participating companies, nor is there any swap for newsletter subscriptions or advertising.


Williams Communications is a multi-service long-distance network service provider, with a U.S.-based network. The company seeks to leverage optical transmission network technologies via a multi-vendor strategy, and by partnering with equipment vendors in order to tailor and integrate products to meet its individual requirements. The company's strategy is closely allied to its unique use of technologies, network topology, and long-term vendor and customer relationships.


Williams Communications targets what the company calls 'bandwidth-centric customers' who wish to address a larger market, vary costs whilst doing so, and implement services more quickly than could be achieved alone. With Williams, all of these benefits can be realised with the customer having no loss of control over the network. From an operational perspective, it will appear as if they had built the network themselves.

According to Matt Bross, "Williams sees its customers as having invested capital in certain geographies. Now they are at the point where they need to address a larger market on a variable cost basis, quickly, and without spending the $7 billion on a network that Williams has. They want to achieve all of this while retaining full control over that expanded network."

Comparing this strategy to that of Broadwing, Matt Bross commented, "Having fat plumbing in place, as Broadwing talks about, is interesting, but the multi-service capability - building an efficient and profitable network - is what Williams Communications is focused on."

Regarding Williams lack of presence outside the U.S., and the threat of missed opportunities that this may hold, Matt Bross cites the failure of companies such as Interoute, Viatel, and GTS Hermes in Europe. Williams has followed a strategy of gaining access to international 'broadband on-ramps'. This strategy has allowed the company to see consistent growth in traffic terminating into, and originating from, those markets, whilst avoiding financial risks. Williams has positions in China-U.S., Japan-U.S., TAT-14 (Europe), APCN-2 (from Japan to China) and Southern Cross (Australia) cables, combined with "the best" connectivity from U.S.-based networks to those cable landing points.

This connectivity, which has gained Williams such customers as KDDI in Japan, has also become a valuable commodity following the advent of opposition from environmental groups in the U.S. to new developments. As a result, some companies have experienced difficulties in gaining additional connectivity.

Summing up Williams Communications' strategy, Matt Bross said, "Our strategy of focusing on deriving revenues from our North American plant has proved to be the correct strategy - for our shareholders and for our financial performance."

"Our commitment to bandwidth-centric customers has differentiated Williams. Our commitment to not competing for our customers' business is key - it is another aspect of our focus on the customer. Our links with SBC, for example, stretch from the top to the bottom of the company - the relationship is nurtured, is long-term, and our livelihoods are inextricably linked."

Network development

Williams Communications claims to be highly efficient when lighting new fibre, said to be a $100 to $200 million proposition, by obtaining greater optical reach for each dollar spent. Matt Bross says this can be seen as a greater contribution to Williams' margins as proven by its quarterly numbers.

"Other service providers have treated bandwidth as though it were free. To me the efficiency of any OC-48 equipment, as it is loaded across a transmission system, relates to efficient use of the optical spectrum available. Williams Communications does this extremely well."

Williams adopts a technology-focused approach to the network, believing that innovation is dependent upon how the network is put together as well as what network components are put together. By understanding the linkages between optical, electrical and software systems, the "higher-order brain functions", Williams says it is able to operate in a heterogeneous environment whilst offering a homogeneous set of services.

Multi-vendor approach

In order to leverage product innovations and put them together in unique ways, Williams takes a multi-vendor approach since there is no guarantee that the next innovation will come from any particular equipment supplier.

Matt Bross believes there will be more significant and meaningful innovations in the future than at any point in the past.

"The innovation genie is out of the bottle. Recently there has been a lot of cloning of innovations, resulting in products with a short lifespan. With contraction in capital markets we will see growth in truly innovative products."

But capitalising on the next innovation is not simply a matter of testing a vendor's technology for speed and power consumption. Williams' approach requires constant coaching with both existing and new suppliers.

"I see such testing as being like a black and white photograph. Instead, we coach the vendor early. We study all of its managerial, operational and financial capabilities - key factors in technological innovation - in order to get the full-colour picture of the vendors products."

Side by side with Corvis

The company's Technology Farm System allows the most innovative technologies to be tapped early and pushed into the market. An example of this is Williams' work with Corvis, where an intimate understanding of that supplier's technology was developed, and together the companies assessed how best to operate the network to yield operational and cost efficiencies.

Describing Williams' work with Corvis, Matt Bross noted, "In building a network the cost of amplifiers is less important than the span of the network. Dr Huber (CEO of Corvis Corporation) was excited about putting his technology into the Williams network because he knew that it would work better in our network than in any other. Operators such as Level 3 and Qwest, who both installed networks after Williams, simply purchased the technology of the day when building out their long-haul networks - the result is that they realise lower margins than does Williams."

Network architecture

Williams Communications network architecture differs from other players in the market, particularly in the deployment of optical amplifiers. As Matt Bross explained, "The Raman or distributed amplification effect peaks at around 50 miles along the fibre. This is the point where amplifiers begin to reach their limits and the signal decays. By capitalising on our amplifier spacing of around 45 miles, versus 60 to 80 miles for our competitors, Williams eliminates regenerators and the associated operational and equipment costs. Despite installing the same elements in the network as other service providers, Williams' attains greater efficiency and revenues as a result of this unique network architecture."

Terabit routing

Mr. Bross stated that Williams is currently evaluating terabit routing as a concept, and is continuing to test products "until they break". He was not able to disclose which suppliers' products are currently under test, though approximately 18 months ago, when Williams Communications' IP network was deployed, terabit routers from companies including Corvis, Avici and Charlottes Web were evaluated.

Williams views 'the router' as one of the fastest changing elements in the network. According to Boardwatch Magazine's Network Assessment results, Williams Communications' Cisco-routed network is ranked first on overall performance among next generation networks. Within the network, the optical layer is used to simplify routing and to simplify the adjacencies between routers - a topic studied by Corvis' Dr Huber.

According to Matt Bross, "Within the network the optical core can be employed to simplify layer 2 and 3 issues - which Williams has carried out very effectively. Terabit routers will, in the future, increase the efficiency of optical networks."

Williams does not currently have digital cross-connects installed in its network. All services are provided to customers by routing traffic across the network as data - whether it is voice, data or video.

Williams uses circuit emulation to move DS1, DS3, TDM, and private-line traffic over the network core. This is an area in which the company leads, having several years experience with the technology.

Discussing a move to terabit routing, Mr Bross said "The deployment of terabit routers will have two key advantages in the core network: they will handle more private line traffic, decreasing the use of other multiplexing technologies; and they will take much of the traffic arising from native IP, such as Gigabit Ethernet."

Williams is content to continue testing terabit routers whilst many of its competitors are already deploying these products. Matt Bross feels that other companies do not conduct such rigorous testing as Williams. He explained, "Customers expect a specific level of testing to be carried out on, for example, a T1 line that they lease from a service provider. These levels of service are currently lacking in terabit packet-routed cores."

Mr Bross added, "Williams offers 'real private network' services to its customers. Our customers experience no loss of control over the network that we provide - with a VPN they would."

Continuing on the topic of terabit routing, Mr Bross said, "Williams knows as much, if not more, about terabit routing as any player out there but we have to know the limits of the products. To this end, we have a 'terabit wind-tunnel' - it is like bending a wing until it breaks. You have to know when it will break before you put it on a plane!"

Market opportunities

Dark fibre

Mr Bross was asked for his comments regarding the stated view of Mr Ellenberger (of Broadwing) that the market has turned away from dark fibre in favour of lit networks. He responded, "We have never built our business case around dark fibre sales, unlike some of our competitors. It is an opportunistic thing; we do not focus at all on selling dark fibre. We have always based our business upon the development of DWDM technologies."

Evidence of this strategy is provided by the fact that over 90 percent of Williams' revenues at the end of the last financial year were derived from recurring business, and not from the sale of dark fibre.

Commenting further on the company's use of DWDM technology, Mr Bross said, "We pioneered the offering of DWDM-based, scaled services. Operators such as Excel and Teleglobe would light dark fibre along a route, while Williams is able to offer 40 Gbit/s bandwidth over the same route to these companies at a lower cost than they could achieve themselves, whilst still enjoying a nice margin."

He added "Where a company such as Level 3 has built a network around 10 or 12 conduits on a route, we would build ours around 3 conduits."


The emergence of 3G and other converging networks is seen as a further opportunity for Williams. In the view of Mr Bross, "3G is an access technology - converged networks, whether cable, DSL, T1, T3, private line, fibre or 3G, are all moving to packet-based technologies. Williams aims to connect to any and all of these networks in their 'native tongue'. 3G wireless, specifically, offers an opportunity for us as a further 'broadband on-ramp' to our efficient, bad-to-the-bone network."

Multimedia traffic

With regard to the impact of multimedia services and resulting demands on the network, Williams feels it has taken a very practical view of how media will drive traffic volumes. Williams Communications' Vyvx Broadband Media business is the market leader, by market share, for services such as video back-haul, and advertisement, production and pre-production processing. The company therefore claims to have a realistic view of how the technologies and the market are going to develop.

"Taking Napster as an example, they proved that there is huge demand for very cheap entertainment services. The main issue here is that the science experiment has taken place before the business case. This happens repeatedly."

He added, "Broadband and the media is a huge market opportunity - people are willing to spend money for convenient access to entertainment."

Indeed, Williams is confident of being in a position to enable the bandwidth revolution. A study by the University of Berkely has estimated that the sum total of data amassed since human kind began, has reached 12.5 Exabytes, and that in the next two and a half years a further 12.5 Exabytes of data will be amassed. Williams' strategy is firmly geared towards facilitating this 'Bandwidth Big-Bang'.

Head to head with Broadwing

Williams and Broadwing are direct competitors in the optical transmission market, but each has followed a different strategy in rolling out their networks. Despite installation of the same network infrastructure as Broadwing, Williams claims to deliver higher bandwidth and greater reach over its network.

When implementing its network, Williams posed the question 'Where are the greatest advances in efficiency going to come from when building optical networks?' The company concluded that one factor would be advancements in the fibre itself; the second would be advancements in DWDM technology. Williams judged that advances in DWDM would offer the greatest benefits. Therefore, as the network was rolled-out, Williams placed optical amplifiers closer together along the route than was standard practice.

Matt Bross explained, "In a network the cumulative noise in the optical system is the square of the launch power. As a generalisation, one bit of information has to consist of 92 photons of light hitting the photon detector at the destination end of the fibre for it to be counted. Increasing the number of wavelengths, multiplied by the data rate of each wavelength, means that more photons need to hit the detector at the destination in order for the data to be discriminated. Longer distances in an optical network therefore require more power at launch."

"The Williams network may have, for example, a 17 decibel average span loss, but this compares to a 23-24 decibel average span loss in competitor networks. With a few colours of light, combined with low data rates, the loss rate is relatively unimportant. Adding more colours of light, combined with higher data rates, results in an exponential increase in noise, which becomes critical. The Williams network infrastructure has been engineered from end-to-end with the aim of minimising span loss. The end result is the transmission of more light, further, while delivering higher data rates."

Rick Ellenberger, CEO of Broadwing, has claimed that Broadwing's optical transmission network is the most sophisticated in the world. Williams Communications, on the other hand, has the stated aim of being the technological leader in optical networking.

Supporting the Williams Communications' claim, Matt Bross commented, "For the same dollar spend on the same equipment, Williams' Corvis transmission gear is delivering 700 billion bits/s, while Broadwing is achieving 400 billion bits/s. In financial terms, this translates to Williams network reaching more capacity for the same dollar spent. Additionally, when the lower bandwidth network reaches full capacity the operator must re-deploy capital to upgrade it. Meanwhile, Williams will be continuing to enjoy margin-rich traffic - fixed costs have been laid off while at the same time the network is returning strong incremental margins."

"A second quantifiable metric is that for the same dollar spent on Corvis technology, Broadwing is able to achieve a reach, without regeneration, of 4,200 kilometers, whereas Williams has achieved 6,400 kilometers. This means that, as you move to all-optical networks, adding elements such as add-drop multiplexers, optical routing and switching, there is more link budget available. This in turn allows greater flexibility in the network topology and restoration plans. Therefore, the overall reliability of the network and of services offered over it, are higher. This level of reliability is delivered for a lower expenditure. These are hard facts supporting our claims of technological leadership."

Summing up the differences between Broadwing and Williams, Matt Bross stated, "Broadwing has built a service-specific network, while Williams is a multi-vendor, multi-service, network provider."

Williams has worked at the multi-vendor approach for reasons of both business continuity and economics. Williams is able to manage transport network equipment from Sycamore, Nortel, Ciena or Corvis. The importance of this approach is that it avoids reliance on a single supplier, and can also reduce capital expenditure costs. For example, when an equipment supplier has supply chain issues, as has occurred recently with shortages of optical components and filters, the effects on Williams are reduced.

As evidence of the success of this strategy, Williams was able to accelerate build-out of its network to complete rollout one year ahead of schedule, despite supply chain problems with some equipment vendors.

Optical Keyhole conducts interviews on the basis of readership interest only. They are not paid for by the participating companies, nor is there any swap for newsletter subscriptions or advertising.

This article is the copyright of Optical Keyhole. It may be freely distributed by any means in an unaltered form.