Roth Promises A Nortel Ready To Compete (Inter@ctive Week)
From Inter@ctive Week for December 22, 1997 by Carol Wilson
On Oct. 1, John Roth became president and chief executive officer of
Northern Telecom Inc., a manufacturer of telecommunications network
hardware and software, with $12 billion in revenue in 1996, that is now pushing hard into Internet and data networking technology. A Nortel employee since he joined the company as a design engineer in 1969, Roth came up through the ranks in Canada to become president of Bell-Northern Research Ltd., commonly called the Canadian version of Bell Labs. He was the first president of Nortel's Wireless Networks organization and is credited with helping the company catch up to other global players in wireless technology after a slow start. Roth was named president of Nortel's North American operations in 1993 and began an extensive reorganization of operations before being named Nortel's chief operating officer in 1995. In February 1997, he was named president of Nortel in addition to continuing as chief operating officer. Roth spoke with Executive Editor Carol Wilson by telephone.
You are taking the helm at Nortel at a time of tremendous change. How
would you characterize today's competitive manufacturing landscape?
It's changing dramatically. The biggest thing that is happening -- if we look back a number of years ago, there were a bunch of old-line telecom suppliers that made everything from soup to nuts. Almost all of us could trace our heritage back to being owned by [former AT&T Corp.] Western Electric or competitors to Western Electric. Within each country, there were national telecom companies and national supply companies.
That situation carried on for quite a long time. The arrival of digital switching was the first break. Then came deregulation -- national carriers now have competition; alternate carriers are springing up. There has also been a lot of upheaval among the traditional manufacturers. Philips is busy selling off its business. ITT and Automatic Electric are gone. Ericsson is becoming a wireless company. The Japanese companies are less prominent today than they used to be.
You still compete heavily with Lucent [Technologies Inc.]. But many
industry analysts believe your real competition these days is coming out of the data networking world.
That's the next big change taking place. There were two big waves --
first was the growth of wireless, which is becoming tremendously
pervasive. We are still very early in the life of this technology, and
new applications, such as the one we are launching with Teligent [LLC], which uses wireless for broadband links to buildings, is just one example of what is still to come. The wireless segment will continue to be a strong growth component for the industry.
The second wave is the growth of IP [Internet Protocol] networks, which is the protocol used for most PC to PC or PC to server, whether it's on the Internet or just an extension of data over a wide area network. Data has always grown -- ever since people could measure -- about 30 percent per year, but it was always a very small part. Late last year, data traffic became equal to voice traffic on the network. And it grows at 10 times the rate of voice. Data traffic grows with PC penetration, and, with an increase in MIPS [processing power measured in millions of instructions per second], phones grow with people.
A lot of the applications that used to be on SNA [mainframe] networks
and X.25 [packet-switched] networks are moving to IP traffic. That is
the big movement -- we are in that business; Lucent has just announced
that it wants to get into that business, but it is not there today.
The people we meet include Cisco [Systems Inc.] and Ascend
[Communications Corp.]. But this is a new industry -- it's still in its infancy.
If you look at what people call the World Wide Wait, the capacity of the Internet has in no way kept up with the demand. This is a tremendous opportunity for our company to take our skills and help build a network that has more of the characteristics of the dial-tone network than the Web network. I call it the Webtone. You don't have to wait; you don't have a fast busy; you don't get dropped by the network.
If we can do that, the Web will become a tremendous global force.
Going back to Cisco, both GTE [Corp.] and Qwest [Communications
International Inc.] are going to build high-capacity fiber-optic
networks -- using us and Cisco. Neither Nortel nor Cisco is going to be able to handle the whole thing alone.
Competitors like Cisco don't take the traditional telecom industry
approach where everyone gets together and agrees on a standard and then you go build products. The data networking world has been driven more by de facto standards. How will Nortel change to compete in that kind of world?
We've done a lot of things. If you look inside Nortel what you see is 38 business units, each of which is headed up by a team. We used to have a central laboratory, which contracted to do work for specific units. We have disassembled the R&D lab and put those people into 38 business units. That way, we put the people who are designing the products in the same business unit with the people talking to the customers, and I can get my designers listening to our customers.
The characteristic of those customers has been that they won't move
until there is a standard. That's changing. They're seeing, with
competition, the need to move faster, sometimes ahead of standards. Some behave that way -- others are not prepared to do so.
We really pay attention to those customers that have a clear vision of
where they want to go.
Also, we have one of the broadest product lines. We passed Lucent
globally in wireless now. We dominate the long-distance network with our fiber-optic systems -- our 10-gigabit system is dominant. The key is how we can put them together for the customers.
Our approach is not to fall in love with a given technology. When we
went into wireless, we focused on being digital wireless. There are
multiple digital standards. Some suppliers said they would pick the one they felt was best. We decided to become experts in all three, because they're all really good. Our customers like that, because they know we can be objective about all this and not try to sell them a standard, because that's what we have.
The same is true when you talk about ATM [Asynchronous Transfer Mode]
and native IP. The traffic is IP. The service the customers are
interested in carrying is IP. Then you look at where the ATM fabric is
valuable and where it's just overhead. We understand where native IP is a really good answer and where, for the services you are going to offer, it makes sense to use ATM.
But datacom companies like Cisco and Ascend have also acquired ATM
capabilities.
When Cisco bought StrataCom [Inc.], that made it a competitor in that
space. In that area, we think we have a much better background in
building reliable networks. Just connecting a gigabit router doesn't
mean you have a network. The routers that are in use today were built
for the computing world that we have in our office. At Nortel, we have
one of the most extensive data networks -- we are the third largest user of Microsoft [Corp.] Exchange -- and the mean time to failure is way too short, the unavailability too high.
I can build a router, but I'd better build a router that doesn't go
down. This is the realm in which this technology is moving, and
certainly companies like Cisco move very fast, but they move very fast
in terms of datacom products. They are not bulletproof, not
24-hours-a-day, never-goes-down products. That is where Nortel and other companies need to be.
Do you consider Microsoft to be a competitor?
No, we are a really good customer of Microsoft's and it is a good
customer of ours.
Nortel recently announced a 1-Meg Modem product that you are bringing to market well ahead of any industry standard approval. Is that a departure for you?
People are now saying the video to the home may or may not have been a
profitable business case, but Internet to the home at data speeds in
excess of 56K [kilobit-per-second] modems that we see certainly is an
opportunity. But, some of the technology that is being pushed to create 6 megabits per second to the home was first developed for video
channels. If you back that off to a megabit, you can do that for
Internet access, and it makes a good service.
Some competitors say a slower-speed service is great, but shouldn't you use Discrete Multitone (DMT) line coding, the North American standard, so that the product is compatible with future, higher-speed products, and the service provider can upgrade?
The DMT technology is not here today -- people are talking about it, but it's not here today. DMT consumes a lot more power, needs energy and cooling, and, so, it's a question of is it worth waiting for or could people make money today with a 1-Meg Modem service?
There are customers saying, 'I'm going to go with 1-Meg Modem and I'll
make so much money; I'll recoup my investment and upgrade my customer,
whom I've already captured, when the higher-speed stuff comes along.'
In a way, we did take a page out of Cisco's book. It's a question of
wait for the standard or enter the market with something compelling and see who bites.
We think this is a neat way for an ISP [Internet service provider] to
sell different grades of service and for the user to get what they need. No one is making money on the Internet today. If we don't allow the various players to make money, it's not economically sustained.
You've signed some major contracts with companies like Qwest that are
building networks from scratch. Is that easier than dealing with
telephone companies that always have to take legacy systems into
account?
Qwest has a network that is 40 percent cheaper than next newest. People like Qwest have nothing to defend. They have the clarity of an attacker. People who have to defend worry about cannibalizing existing service. If they really sat down and thought about it, they can run these things as
an overlay; let their legacy systems fade away. If Qwest can start from
scratch, so could anyone else if they made that choice.
But hasn't the history of the telephone network been that nothing goes
away -- that it all migrates forward?
That's one of our trademarks that we've had over the years -- to be
evergreen -- the ability to migrate over the years. That doesn't mean
you don't throw something away. If you go into service, spend a certain amount on equipment -- but, with the cost you spend to capture customers and train people, actually the equipment purchase is fairly small. If Nortel equipment can be upgraded and everything else stays in place, then you've maintained your investment in people and the methods of procedure that people use, which is the real investment.
Does the telecom industry as a whole need to change its approach to
standards-making in order to enable services to go into the network much more quickly?
I think what you are highlighting is that, for example, ISDN [Integrated Services Digital Network] was conceived as a standard in 1980-1981. But look at the position today of ISDN vs. the Internet. There were no standards on the Internet at the beginning.
The committee work on standards is often overtaken by market events.
That 1-Meg is so appealing -- you may have 6 [megabytes] in three years -- I'll take this one right now, thank you very much. The key here is to create an industry standard. There are two ways to do it -- Cisco created an industry standard, but it sold [the product]; it didn't go to an industry forum to get a stamp of approval.
With our 1-Meg Modem, we are talking to all kinds of people --
complementary players and competitors -- about licensing that
technology. We're not looking to get rich off it; we'd like to get on
with it. No one will buy proprietary technology.
There was a period of time when Nortel seemed to be constantly
reorganizing itself -- your competition took that opportunity to say you were internally focused. Is that period over?
[Laughing] Well, I reorganized when I took over Nortel. In North America in 1994, I started redefining the piece of the company that [former Nortel Chairman] Jean Monte asked me to run.
But some of the reorganization was market-focused. In 1991, we started
the wireless group -- we started late. We thought it wouldn't be that
hard to overtake wireless guys, but we weren't going to do it through
our traditional method of bringing products to market. I acquired
technologies and very quickly built the wireless.
What I learned from that was the importance of speed to market. Also I
realized we are a tremendously capable company. Many of the challenges
involved transforming an organization that was highly functionalized,
very rigid,and focused on the CEO into an organization that focused on
the customer. Last year we grew in North America by 30 percent.
We've taken that management model across the company. I think it's a
template that works rather well. It is still relatively new -- people
are still learning how to make it work.
In addition, there is a strategy: When you have to win the bet, you
cover your bets. With ATM vs. IP -- for sure we are going to build IP
networks, whether they run on ATM switches or IP, I don't care. We'll
make it happen both ways and give the customer two really good offers.
'I can build a router, but I'd better build a router that doesn't go
down. This is the realm in which this technology is moving' |
