Williams Communications - Second time around
15th October 2001
Interview with Howard Janzen, Chairman, President and CEO of Williams Communications.
This is part one of an interview with Williams Communications. Part two, with Williams' CTO Matt Bross will appear in the next few weeks.
OpticalKeyhole.com and the Optical Networks Daily newsletter conduct 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.
History
Williams Communications' initial entry into the telecommunications arena began in 1985. At that time, only three other companies in the U.S. were building nationwide optical fibre networks: AT&T, MCI and Sprint. Williams can therefore claim the pioneering mantle of being the first company, with a core business outside of telecommunications, to recognise the opportunity of deploying its skill set and assets into that market.
Williams' approach was to deploy fibre through decommissioned pipelines. The majority of that business was reluctantly sold to WorldCom in January 1995, the offer being so beneficial to shareholders that it "could not be overlooked".
However, Williams, wishing to remain in the market, managed to retain a number of telecommunications assets that were not vital to the WorldCom deal, including approximately 50 percent of employees in the sector and Vyvx, Williams Communications' broadband media business. Significantly, Williams also retained a single fibre throughout the entire existing network and these factors combined allowed Williams to stay in the business and operate a network. As a result, Williams had a launching point for what the company describes as "the largest next-generation network platform in North America".
Unique positioning
Mr. Janzen describes the current Williams Communications as "a business targeting bandwidth-centric companies". He added that Williams has been very careful to align the company's interests with those of its customers. According to Mr. Janzen, a number of other players in the market also say they have adopted this strategy but in fact those operators are also chasing large enterprise and retail customers. As such, Mr. Janzen noted, this fundamentally puts them at odds with the customer set that Williams serves. He added that in theory Williams Communications has many competitors in its space, but because of a unique positioning, Williams retains a huge advantage.
Suppliers and technology
While Mr. Janzen agreed that optical switching is entirely possible today, citing developers such as Lucent, Calient, Corvis and others, and is a critical issue for networks such as his own, he believes that the technology remains in the early stages. Though Williams Communications will certainly deploy optical switching, as the need for higher speeds demands it.
The critical issue for Williams Communications currently is complete reliability and interoperability, both with existing vendors and those that Williams may identify in the future. Williams Communications employs its unique Technology Farm System to manage technology and enable deployment of new or alternative technologies faster than the competition, and Mr. Janzen notes, "In each case we do it with great care".
Supplier relationships
On the question of the relationship between technology deployment and Williams Communications' financial investments in leading-edge technology vendors, Mr. Janzen said that the investments were primarily a vehicle for building a strong and lasting relationship with the companies in question. The involvement in technology companies allows Williams to see deep inside the company, and to judge that company as a long-term viable supplier.
"Our Technology Farm system is about more than measuring speeds and feeds and power consumption; it is about vendors understanding exactly what Williams is looking for. We are driving them to build products that can specifically help us in our business."
"The fundamental reason we are so passionate about our system, and the relationships it helps build, is that this is the key driver for our opportunity to be the lowest cost, most efficient network. We have focused all our actions on making that the outcome."
A point put to Mr. Janzen was that service providers, in their inexorable drive toward higher functionality and ever increasing technology output from their vendors, but at an ever-decreasing price, might seem unwilling to pay for required innovations, and could drive the small startup company out of business. Mr. Janzen responded that the "innovation genie is already out of the bottle". He noted that the amount of wealth that has been created through technology innovation, even allowing for the current market meltdown, would ensure that innovation will not stop.
"Capital spending rocked along at a steady pace for 15 or 20 years, took a huge leap around 1995, and peaked 18 months ago. However, people who think there is no money out there to continue to drive innovation are very wrong as venture capital spending today is still astronomical compared to the previous 20-year period."
Mr. Janzen made the point that Williams Communications is sitting on network platforms with plenty of dark fibre waiting to be lit. "The demand growth is very large, and each time we deploy more capacity we need to be using the best technology."
In the last year alone, the Technology Farm team at Williams Communications examined more than six hundred solutions, including in-depth interactions with those companies. However, Williams is constantly looking for 10X improvements and with so much technology available can barely be bothered to look at systems only offering 50 percent improvements.
"We can get improvements anyway from tweaking processes and internal changes. What we look for is a 10X improvement in the cost to build, the cost to own and service velocity."
Mr. Janzen added that it is only when a vendor company has passed an initial screening that Williams Communications will then look at a potential "winner" and seek to develop a relationship with a supplier.
Ultra long haul
Contrary to many people's thinking, Mr. Janzen believes that ultra long haul (ULH) equipment suppliers do have a bright future in the U.S. "When you look at mesh networks, ULH becomes very important for paths around that network."
Dark fibre and network economics
Mr. Janzen said Williams Communications has been very clear in its strategy not to focus on dark fibre, preferring instead to pursue recurring, service revenues rather than one-off sales. Williams does in fact benefit from dark fibre deals, but only when there are service contracts as part of the overall package.
"This is what is hurting Level 3 right now - they had a lot of dark fibre sales, which are basically one-time events."
"I personally sat with a number of dot.com companies, when they were seeking to buy dark fibre, and our team showed them the economics which demonstrate what a bad decision it would be."
Mr. Janzen was adamant that such companies would never have been able to justify the costs associated with lighting and managing a network. The route many of them should have taken, but did not, was buying services that are easily ramped up and varied, instead of lighting a pipe that, to a great extent, was going to sit empty. Indeed, whilst many startups Mr. Janzen spoke to had admitted that they did not need dark fibre, they wanted it anyway, presumably for the prestige of owning their own network. Most of those companies have now collapsed, in part due to bad decision-making, Mr. Janzen noted, and a number of others have stepped back and restructured their dark fibre deals.
According to Mr. Janzen, it takes up to three years to turn dark fibre into a lit network, taking into account the creation of a qualified team or outsourcing of the project. In addition, the capital required to light up dark fibre is about eight times the cost of initially buying or leasing the fibre itself. Companies buying dark fibre today tend to be the larger players who need to complete the network jigsaw, and the market is very different now from how it was a year ago.
Pricing models
With service prices dropping steeply and last-mile connectivity prices rising, Mr. Janzen feels many carrier-pricing models no longer make sense. Over the next few years, he expects to see many changes in basic models, some resulting from regulatory issues.
Mr. Janzen does not believe that the transport of bits - pure transport with no value-added - will ever be free, but he did think it would continue to decline in price, noting, "The costs on our network have declined on average 40 percent per year". On a per bit cost, Williams Communications calculates that the cost of lighting fibre has dropped more than tenfold since 1998, and Williams' intention is to pass on a portion of that decrease to customers.
In any case, costs have dropped over time and continue to do so. Using the Bell companies as an example, revenue and profitability have increased dramatically due to elasticity in the huge volumes.
At the end of the day though, providers are successfully selling bundles of services, even if the cost per service is not exactly transparent to the user.
Bandwidth glut
On the topic of what is known as 'the bandwidth glut', Mr. Janzen explained that most of the business press does not understand, or is blind to, the difference between dark fibre and provisionable capacity.
"The time and capital to put dark fibre in service should clearly not be counted as provisionable capacity. That is like saying that because there is a lot of sand on the beach, there must therefore be a surplus of microprocessors."
Williams Communications' policy has been to place enough fibre in the ground to meet projected capacity at that time, whilst allowing technically more advanced fibre to be easily deployed in the future.
On average, Williams Communications deploys ninety-six fibres throughout its network and three conduits. One fibre pair is lit everywhere in the network, and on some routes, six fibres are lit. Two conduits are spare for future fibre deployment.
"From a dark fibre perspective, this is a low utilisation, but the real economics are driven by management of the lit utilisation of the network. The incremental costs for the additional dark fibre are really pretty low."
Rights-of-way
Following recent judicial decisions, the industry could be facing a huge bill to compensate landowners for rights-of-way (ROW). Mr. Janzen explained that this was because a lot of fibre was laid alongside railroads, and the railroad companies had assumed that they had the right to lay fibre. The landowners have argued, seemingly successfully, that the railroads had no right to lay communication lines, only the right to lay a railroad.
However, pipeliners have, according to Mr. Janzen, often negotiated with each individual landowner, and pipeline ROW deals that Williams Communications has negotiated have included rights for communication lines. When Williams Communications wished to acquire its own ROW, it acquired part of the pipeline team to do the negotiations.
Williams Communications does have some minor exposure to this potentially massive problem, particularly in Louisiana, but in each case the original ROW acquirer, Williams says, has a duty to perfect the contract. As such, Williams Communications is likely to remain relatively unaffected by the staggeringly large figures being touted in respect of compensation that the carriers are likely to have to pay.
Investment and capacity utilisation
Williams Communications' CEO feels it is a mistake to link the severe downturn being experienced by equipment suppliers with capacity utilisation, noting that IP network traffic is growing at roughly 100 percent a year.
However, obtaining funding in today's capital markets is very difficult and as a result service providers have had to dramatically cut capital spending. Mr. Janzen said that the discretionary proportion of Williams Communications' capital spending last year would have been as low as 10%, or even slightly less. "The rest had to be spent".
For the 2001-2002 period, Williams Communications has budgeted $2.0 billion for capital investment, of which roughly half is discretionary. The $2.0 billion was reduced from $3.2 billion projected earlier in the year. The reduced capital requirement reflects: the redeployment of warehouse equipment, availability of better-than-expected equipment pricing, and greater capacity realization from deployed technology.
Significantly, Mr. Janzen said, "We are cranking the dial down on as much of that spending as possible, to make sure that the company is adequately funded because nobody in our industry can go to the capital market today."
"As a result of such investment cuts, it is true that vendors are feeling the pressure, but this has nothing to do with bandwidth or any other type of glut, it has to do with management of already-installed capacity and taking all the slack out of the rubber band before we continue lighting."
"In fact, Williams Communications is adding capacity this year but at nowhere near the same rate, and this is what the suppliers are missing - they looked at the huge amount of money being invested in building next generation platforms and thought it would continue."
Liquidity
According to Williams Communications' second quarter 2001 earnings release, the company continues to project turning EBITDA (earnings before taxes, depreciation and amortization) positive on an annual run rate basis by year-end 2001 and becoming free cash flow positive by year-end 2003. In addition, with the revised cap-ex plan and the successful execution of the company's financing plan, Williams Communications expects to be funded into 2004.
As of June 30, 2001, Williams Communications had cash and short-term investments of approximately $1.3 billion and has an additional $525 million available under a bank credit facility revolver. In addition, the company recently raised $276 million on the sale-leaseback of the company's corporate headquarters in Tulsa.
Those dollars are reserved to execute the company's business plan - to fund capital plans and see Williams Communications through to being EBITDA positive by the end of this year, which is "well on track". Williams has some flexibility to use part of the cash for acquisitions, but "we could be pummelled to death if we make a move that Wall Street feels would leave us inadequately funded". Mr. Janzen said that there are a lot of opportunities and bargains to be had, but like everyone else, he feels that caution is the best policy.
However, Williams Communications, on October 11th, announced the acquisition of the majority of the facility and equipment assets of iBEAM Broadcasting, in which it already had a 49 percent stake. According to Mr. Janzen, "What we get from iBEAM would have cost us more dollars to do internally, and those kinds of moves we do have the flexibility to make".
Bandwidth trading
Mr. Janzen is of the opinion that a market for bandwidth trading exists only for the very low speed and mostly uninteresting types of service. However, the skills that a trader brings to those services are very important. Post spinout from its parent, Williams has therefore retained its own, previously energy focused, bandwidth trading people.
Williams Communications itself does foresee an opportunity to trade bandwidth and claims to have been profitably in that business since 2000. But Mr. Janzen believes that some of the large claims made so far in this market relate to the same bandwidth being traded a number of different times.
"The big number is not real, but it will happen slowly. Our sweet spot will be the high end, which will keep moving up, and we want to participate in it. However, the risk management and other processes that traders bring us are very valuable, and we have moved our pricing function into that group. Even if the group never traded bandwidth per se, I would pay for their skill set."
Network evolution
To Williams Communications, having an expanding full set of services is critical to both capturing and retaining customers. Williams' top ten customers, Mr. Janzen says, use four or more of its services, and the company is constantly seeking innovative products to sell in order to further secure relationships. So while ensuring that the network supports enough capacity is a key factor, Williams believes that the network as a pure transport mechanism is not a business model that will work in the medium to long term.
Bottlenecks
Mr. Janzen made the point that companies like his own, building long-haul networks, were often criticised for taking the easy way out. While commenting that "it didn't feel very easy to me", he conceding that it was "easier and less capital intensive than building last mile connectivity".
However, all service providers today have the role of easing bottlenecks. The first such bottleneck, data processing power, was solved by Intel and its peers, and unleashed all kinds of unimagined applications. Companies like Williams Communications solved the next bottleneck, which Mr. Janzen maintained was indisputably long-haul capacity.
"Now we are down to the last-mile bottleneck, but that won't be the last - it will just move somewhere else - perhaps back to the core."
In the U.S., 11 million households, or roughly 18 percent, are expected by the end of 2001 to have true broadband capability. "Once we start seeing numbers in the 30 percent plus range, we will see lots of new applications emerging, and coming on stream when there are enough customers to access them."
Vyvx unit
An appetite for adding services to the portfolio, and leveraging more products and services, means Williams Communications will consider anything that is consistent with customer requirements. For example, Williams media and broadcasting subsidiary Vyvx has a rich set of products that drive lots of network capacity, an ideal combination for the company.
Vyvx has itself been in existence for a dozen years or so, and Mr. Janzen is adamant that no other carrier has as much experience or understanding of this business. Vyvx was the first company to move broadcast video from satellite to fibre.
"The problems we have with international media traffic is with the other players, who are often unable to realise that this is traffic that will not tolerate an interruption - it has to be perfect and everything has to be ready exactly when it is needed."
Customer satisfaction is measured in all of Williams Communications' business units, and within Vyvx in particular, Mr. Janzen says, "it is off the scale, we cannot even use the conventional measurement anymore". Vyvx moves more than 80 percent of all professional sports, and about 60 percent of all live television, as well as about 35 percent of the spot ads that are seen on US television.
"We have captured a tremendous market share and we are ideally positioned to leverage these relationships into tomorrow's business by moving this content in different ways from how it is moving today - deploying Internet-type technologies to move rich content to consumers."
Future of video
Mr. Janzen does not consider the future of video overstated - quite the opposite, but believes the issue is one of timing.
"I will not go out on a limb on that one, because neither I nor anyone else knows when the market will really emerge. Last-mile broadband connectivity is the critical issue, but when that gets close to truly being deployed, I believe it will not only unleash video but full multimedia capability."
On the subject of local caching of non-live video content, Mr. Janzen considers that current economics support it but there would be a tremendous amount of traffic from moving content to local caches, keeping it updated, moving it to the right local markets etc.
"It is a question of costs - as the cost of bandwidth comes down, the cost of moving content will change dramatically. Today it may be cached, but in future it may well be cheaper, perhaps far cheaper, to move it all on the network."
Overseas expansion
Williams Communications has been noticably more reticent to delve into overseas markets than many of its competitors. Mr. Janzen responded to this point by emphasising that his company seeks to recognise where it has strengths and weaknesses - a key strength of Williams, and the subject of its focus, being ownership of "a best-in-class platform in the U.S. - the largest telecommunications market in the world".
Williams Communications' international strategy is centred on providing direct connectivity to all the financial capitals around the world.
"We do have that in place - we have offices around the world, and have assets, but what we have not done is to rip open the streets in London and lay a Williams Communications network."
To date, Williams Communications has worked through partnerships, which it regards as much more efficient in terms of capital. As an example, Mr. Janzen cited the recent broadening of an existing relationship with KDDI, allowing each party to leverage the others' network strengths. Williams also has dark fibre and wavelength capacity in Europe, obtained through a transaction with Telia. But before Williams reaches the stage of lighting the fibre, the feeling is it would need to have built, what Mr. Janzen described as, a "significant business in Europe".
Competition
Mr. Janzen cites companies like Level 3, Qwest, Broadwing and Global Crossing as competitors, but notes every one of them is a player in the enterprise market and not therefore focused purely on the bandwidth-centric customer segment. Indeed, a typical Williams Communications' customer tends to view the above carriers as its own competitor, though that does not mean they will never buy capacity from them. Some carriers, Mr. Janzen noted, do not care about the conflict of interest issue, while others, like SBC, are sensitive to the issue.
Level 3
Mr. Janzen believes that Level 3 has been unfairly hammered recently from negative sentiment that was largely based on such dubious concepts as "capacity glut". He cited independent research that looked at twenty-two individual city pairs, and in detail at how many were actually in service and what the utilisation rates were. The research report concluded that overall utilisation, in percentage terms, was in the upper 60s and Mr. Janzen said, "70 percent is basically where you are out of gas and you figure you are running out of capacity, so I think that Level 3's drubbing is somewhat unfair". "However, their success in selling dark fibre now turns into a big challenge as that opportunity is not there today."
Williams Communications, on the other hand, is positioned very differently from Level 3 with recurring revenues representing 95 percent of the total. Also, Mr. Janzen said, many of Level 3's dot com customers are presently experiencing financial difficulties, whereas Williams' customer set is very strong.
"Half of our revenues last year came from companies that were investment-grade credits, and this year 85-90 percent of our customers fall into the low-average credit risk category."
Level 3 is now trying to adopt a focus more in line with that of Williams Communications, even to the extent of using the same terminology - 'bandwidth-centric' - "but if you look at Level 3's revenue, and divide out the number of customers, they have a high customer-count, somewhere between 2,000 to 3,000 - clearly, they are still selling to enterprise customers."
Broadwing
Broadwing deserves credit for its aggressive deployment of Corvis technology, according to Mr. Janzen, but his feeling is that this is the primary basis for the claim of Broadwing's President and CEO Rick Ellenberger that Broadwing is the most 'technically alert' network in the U.S.
"To see the reality, just look at industry recognition - Williams Communications won SUPERQuest awards three years in a row for being the best built backbone network. This year Williams came in second place, again for the best built backbone network, so out of four years we have been number one three times and number two once - I don't see Broadwing gaining that level of recognition"
In 1998, Williams Communications won the award for multiservice aggregation with Cerent (now Cisco). In 1999, the award was won with Sycamore Networks.
"Actually, we influenced Sycamore's focus as that company was starting its business. We wanted to be able to sell optical waves, and we asked Sycamore to change its design slightly. Through listening to us, we were able to use Sycamore equipment to be the first in the industry to sell optical waves, and we are able to interleave those waves with OC-192 traffic."
In 2000, the award was won for deployment of CIENA's CoreDirector. At that year's SuperComm, Williams also demonstrated integration between Avici and Corvis, whereby Avici's router actually turned on waves in the network. "Our team actually wrote the software specs for that".
On Broadwing's deployment of Corvis equipment, Mr. Janzen noted that press releases had referenced the fact that Broadwing could send light 4,000 km without regeneration. "However, on our network we can send light 6,400 km without regeneration." Mr. Janzen believes that Williams Communications' is able to achieve almost double Broadwing's capacity rate, and he said the reason is "a deep technical understanding that our team has developed, with 17 years experience, which has led us to light our network in a very different way from not only Broadwing but pretty much anybody else".
Therefore, Mr. Janzen maintained that his network is able to harvest more capacity from a piece of glass than any other network. Partly this is due to fundamental decisions taken in design such as the spacing of amplifiers.
"When we lit the network we spaced conservatively because we would be able to drive a far larger number of waves down the network - and that advantage is not linear, it is exponential."
According to Mr. Janzen, other carriers tended to go with what the vendors recommended. He noted that Broadwing had said it put in the boxes and they worked. Williams Communications on the other hand, through extensive testing, identified almost one hundred different issues in relation to the Corvis equipment, some related to software and some to hardware. Williams had not turned on Corvis gear to real customers until all those issues had been resolved. Mr. Janzen said, "This is not unique to Corvis. Williams makes all its vendors run the gauntlet. We've yet to see a vendor's gear work right out of the box."
"Vendors like working with us, because we take their kit and break it, helping them deliver a better product."
OpticalKeyhole.com and the Optical Networks Daily newsletter conduct 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 Electronics International and Optical Keyhole. It may be freely distributed by any means in an unaltered form.
opticalkeyhole.com |
