3Com's enterprise net chief speaks out

Edgar Masri, the new senior vice president in charge of 3Com's enterprise business unit, spoke with Network World editors last week about 3Com's challenges in the enterprise network market.

Q. What's the plan to increase enterprise sales?

A. The challenge with any announcement like the one we just had [CEO Eric Benhamou announced that 3Com will not meet Wall St. expectations for the company's third fiscal quarter) is it casts concern and uncertainty on all parts of our business, including the ones that are doing well. The enterprise business is not without challenges. We truly feel we have two businesses there. My job is to ensure that they look like one.

The reason it's two businesses is that we have a workgroup and, in association, [small and medium enterprise (SME) business]. Then on the other hand you have the core chassis business and in correlation, the large enterprise.

We have actually increased our lead in the workgroup. We increased it at the expense of Cisco, we increased it at the expense of Bay. Contrary to what some of the competition had been mentioning, I believe on the workgroup and SME, we're in a very strong position. We've strengthened it in the past three months.

When we win the core, 95% of the time we win the edge. If a company like Cisco has such a strong position in the large enterprise that means a lot of what they're doing in the workgroup is by definition coming from those large deals. So if we are increasing our edge solutions that means it's selling into environments where the whole network is very much based on the workgroup, and that's the SME marketplace.

Q. What are the different market dynamics between SME and large enterprises?

A. When we increase our sales of workgroup products, we could make two implications out of it. Either the SME market is growing - and we think it is - but also it means we are selling well into large enterprises because we have a competitive solution at the edge. When our workgroup grows, it is definitely an indication that the SME market is growing; but the fact that we're not winning the core doesn't necessarily mean that the large enterprise market is slowing down.

Over the past three or four months though, there have been many indications that the enterprise market is slowing down. There are some timely factors that are causing that. Y2K - we are hearing systematically from almost every bank we talk to that they have set deadlines for when they can procure product and deadlines for when get the product into the network.

Our sense is that there's going to be a bit of a slowdown. Even if there is not an imminent disaster people are starting to fear Y2K, and what you do when you're afraid is slow things down. I don't expect it to be a major stoppage.

The other factor is more along the lines of LAN and WAN. Right now, very few people really need 100M bit/sec switching. Instead of buying 10M bit/sec shared or 10M bit/sec switched, people are buying 10/100. Why buy 10 megabit when you can afford 10/100? So people don't really need 10/100, but they can afford it.

The change in the core is more expensive and more impactful. You don't change your core thing. At the edge you can just go and add 100M. You cannot do the same thing on the switching fabric in the core. The bandwidth needed in the LAN is not as intense as people would like to think.

That's why for the next six months or a year, there will be a slowdown in the enterprise. People are absorbing the bandwidth that they bought. Also, the intensity of the need of bandwidth on the LAN has been overrated. Is that a fundamental slowdown? No. It's just a cycle. People have now got very nice bandwidth at the edge. As the gigabit solutions solidify - you start having more gigabit and Layer 3 - you will have things pick up. But that's not going to happen in the short term. It's going to happen more in nine months or so from now.

Q. What are the specific issues that are challenging 3Com in the core?

A. I have to admit very bluntly that in this market, things go by two- to three-year cycles. We missed, in the past six years, two key cycles: basic high-density chassis with nice routing technology; and 10/100 in the core. I believe it's an unforgivable mistake for a company like 3Com because we guessed that shift and those dynamics on the edge. Cisco bought Kalpana and Grand Junction, which are Fast Ethernet companies, and we through homegrown technology built 10/100 and came from being behind them in switching at the edge to being ahead. But then when you came down to that basement we did not have the 100M bit/sec switching. And that's why we have a measly two-something percent market share in that space and Cisco has the lion's share. I blame ourselves more than I blame some of the other competitors because we made the right bets and we guessed right at the edge; but we didn't follow it through. All of our other competitors, except Cisco, floundered too, but they missed it on both ends, and they still actually have in some ways better share than we do. Cisco did the best job because they went and bought Crescendo and just hit it right on. They made the right guess.

We have corrected our miss with the CoreBuilder 9000, one of the better platforms out there on the market. It's a high availability, gigabit switch. We actually learned from our mistake and figured we have to push hard on gigabit.

The second challenge is, as we have built ourselves towards large enterprises, we have still gone and approached it from a channel perspective; sell it to resellers and they can take care of it. We are just now building the customer supply chain where there is effective preselling, effective postselling and very effective service and support. We never had all those pieces built to the critical mass that is needed. Large enterprises rely a lot on post sales service and support. We have spent the past year fixing a lot of this and I feel like we are now just emerging with a system that is reasonable. But we've lost some momentum.

Q. What is it about convergence that has 3Com better positioned than Cisco, for example?

A. We believe we have the most comprehensive coverage in this space from a phased approach as well as from a market approach. We have been very modest here and recognized that we are not going to go at it alone when it comes to voice, especially in the large enterprise. We believe relative to basic data communications companies we are much better positioned because we have the support and the partnership of a company like Siemens. Relative to the voice companies, we believe we have the advantage of the speed with which data communications companies go and run with things. The best example we have is where we're going to start to see LAN telephony and convergence happening, which is in the SME.

We had the Selsius system - this company that Cisco bought - in our labs a year ago because we were trying to see how we could deploy LAN telephony even internally. Despite their claims that it could go to 1,000 users, it was breaking down at about 40 users. We believe with the NBX we have a superior solution.

Relative to the voice companies, the speed at which the data communications companies work, the fact that this is happening at the low-end is a paradigm shift that if we go and execute well on, we are going to be quite successful. I believe the NBX acquisition is going to make a fundamental difference in this area.

We're going to make sure we execute right on it in SME because we have a strong position there and we want to use it to get the large enterprise. It's a tough battle but we have also tough partners.

Q. Do you think you need to make an acquisition in the enterprise area?

A. I'll give you a quick answer and say no, but I'll qualify it. No for the very simple reason that for the past four or five years what we totally missed out on is that it is not the technology only, it is the infrastructure.

If we go and buy a company that has a nice infrastructure then we are capable enough to totally leverage it as opposed to what we did when we acquired Chipcom and Synernetics, which is to buy them and them try to mold them into the 3Com mold and therefore lose what they bring the most, which is a system approach and system sales capability.

Let's say Xylan was on the block for us. It would not have helped us because we would have had the same thing happen that happened with Chipcom. It's not just a bad reflection on 3Com - although I'm willing to take the responsibility that our DNA is so much like, 'We know how to do it with the channels, etc. Let's just integrate those people into us.'

The bigger problem is all those companies, including Chipcom, had started as start-ups, had won and were just facing the issues of infrastructure that we're just solving. The only two companies that have solved that are Cisco, because they've just managed to execute very well, and then Bay through acquisition by Nortel.

For us to go and buy any of those companies we haven't solved the fundamentals. If, for whatever reason, Siemens blows up - the head guy figures out 3Com is not the company I want to work with-we would have to partner very, very strongly with a company that has an infrastructure like that.

Q. Are customers today concerned about 3Com getting acquired?

A. I think customers should be concerned about 3Com getting acquired. I can give you a long answer to say that we have so much value, that anyone who looks at what we are doing would say, 3Com may be bought. This concern should be a legitimate one, but should it be a negative concern? No, but there is a risk with any acquisition that you have uncertainty.

Whoever would try to acquire 3Com would have a lot of money to be able to support us. You saw when Bay was acquired by Nortel they had a challenge for three months or so, but now I'm hearing a lot of good things.

I'm not saying that we'd like to be acquired. All I'm saying is that given the slowdown in the market and our announcement yesterday [by Benhamou about the financial outlook], there is a risk that some companies will start looking at us and see that 3Com is undervalued. But I like to think our customers know we're not going to do anything stupid.

We believe that we still have enough strengths to go ahead independently. To put myself in the head of Eric [Benhamou] I do think we could go it alone for two reasons. One, there is a great paradigm shift taking place and we're banking on it. And second, even in the large enterprises, customers want to have options. And you're not going to have one player providing all the options.

Cisco's current position is not sustainable. It will be eroded. It may still be in the 40% to 50% range. Look where the growth is happening: Gigabit and Layer 3 switching. Where it's stagnating is at Layer 2 Fast Ethernet at the core. This is where Cisco's strong and this is where they were the only player in town. But now I'm sorry, they are not.

Q. What's the one weakness of Cisco's you want most to take advantage of today?

A. I would like to think there's going to be arrogance, but they've been very good at executing. If Cisco was too arrogant, it would have messed up.

I say this paradigm shift to converegence. It's not because they are more exposed than anyone else; but because they are trying to go at it alone. It's a tough one in large enterprises.

They probably will do well in the SME because I believe, even if we have a superior solution, Selsius will figure it out.

But the big battle is in the large enterprise. The battle in the large enterprise is not just to get the solution, it's to have that aura that you can do it. That battle is going to pit them against Lucent and Nortel. And that's a different ball game. T

hey may still do well because in the past they've had major challenges and they've come out of them strong. All they have to do is make fewer mistakes than Lucent and Nortel, but from all I'm seeing Lucent and Nortel are taking this challenge very seriously.

You just see around here how Lucent is trying to hire people. You are dealing with an aggressive player, this is not your stodgy AT&T.

Masri bio
From 3Com.