Covad Communications was only founded in 1996. However, thanks
to its ability to perceive the benefits of DSL, in particular
the speed of deployment, the net-centric services provider has
rapidly gained market share at minimal cost. Michael Bowen,
telecoms equity analyst at BT Alex Brown, comments: ?Covad was
really one of the first to go out and grab a lot geographic
space if you will in the central offices of the ILECs. So they
really were anywhere from 12-15 months ahead of the
competition. They have a tremendous number of central offices
that are service-ready in the neighbourhood of I believe, 370
central offices as of this last quarter, which places them a
little bit ahead.?
Covad Communications has been focusing on the provision of
services to small businesses and ISPs. However, they are now
looking at the consumer market. Bowen notes: ?They are not
competing directly against the RBOCs yet. They have said that
they are planning to roll out a consumer-orientated product,
albeit using a co-marketing agreement through ISPs. Most of
their revenue is coming from businesses, rather than consumers.
The RBOCs have to date focused on the consumer segment
primarily because they do not want to cannibalize their T1 and
DS3 revenues so far. They are not really going after that
particular market at this point. But at some juncture we may
well see that happening. Companies such as Covad, Rhythms and
Northpoint are already provisioning lines. They have got a head
start on the RBOCs and thus far, on the regulatory side, things
are becoming a little bit easier for these competitive
companies to provision lines out of the RBO central offices.?
Priced at $18 at the time of its IPO in January 1999, the
company's shares are now trading at over $80. Covad
Communications remains an attractive acquisition target, as
Bowen explains: ?It would be foolish not to view them
eventually as a take-over target if they do well, much like
Teleport which was taken over by AT&T.?
Bowen believes that companies such as Qwest and MCI-WorldCom
are more likely than the RBOCs to acquire Covad: ?You probably
won't see an acquisition by an RBOC because from the regulatory
standpoint, even when they do get 271 intra-latta long-distance
entry, they really do not want to perceived by their regulators
as snapping up their competition. The whole idea is to promote
competition in the local exchange market. But, for somebody
like an AT&T, Qwest, MCI-WorldCom which will possibly
become very serious about provisioning xDSL services, there is
no doubt in my mind that these companies will get looked at. I
am sure they get calls already even in the nascent stages of
the industry. Quite frankly, companies such as AT&T, Qwest,
MCI-WorldCom will probably wait until they gain more market
share and it is a lot cheaper to date, even at these
valuations, to purchase a company such as Covad and Northpoint
than to build your own.?
In an interview with Global Telecoms Business, the president
and CEO of Covad Communications Robert Knowling outlines his
vision of the high-speed data market and the steps that the
company is taking to achieve the goal of being the leading
provider of high-speed access.
Could you tell us about your recently launched
"Telesurfer"service? How are you attacking the residential
market in the US?
Knowling: We launched Telesurfer about a month ago.
Back when we were doing our IPO, we announced that the consumer
launch would take place at the end of 1999 or perhaps in the
first quarter of 2000.
A couple of things have happened in the business that made me
rethink the proposition. In essence we launched the Telesurfer
service to dabble our toe, if you will, into the consumer water
space. We look at it as a very disciplined approach.
Only a few of our select partners sell the Telesurfer Service.
We are only offering two speeds of high-speed access. We are
hoping to gauge consumer interest, learn some intelligence
around pricing and some things about back-office operations.
We do believe that over time the consumer will be a large
component of our business mix, but we are primarily testing
ourselves in the market: my reason for doing so is that we are
deploying the networks so rapidly and that we currently have
DSLAM capacity. So the incremental cost of adding a consumer to
my network, with a high profitability of success in terms of
filling the order, made a lot of sense to me.
What are Covad's competitive advantages? How do you
see the demand for high-speed data services?
Knowling: My competitive advantage is quite frankly
the speed of deployment of my network. I have a first mover
advantage owing to our early leverage of the Telecommunications
Act. Let me outline the process to illustrate why we have this
first mover advantage.
First you have to obtain CLEC status: secondly you have to get
inter-connection agreements with the incumbent local carriers.
Then you have to pick an equipment platform, deploy equipment,
hire employees and resolve a number of other business issues.
This process takes anywhere from 12-24 months, depending on the
territory. It is not easy. I still do not have inter-connection
agreements from two telcos and we have been in inter-connection
agreement discussions and negotiations for over two years.
Anybody wanting to be a clone of Covad has these barriers to
entry. This is just a function of the regulatory process to
enter this business. So our first mover advantage is very
In addition, the central office of a telephone company does
not have infinite capacity to co-locate equipment and there is
a tremendous backlog in terms of the telco's ability to allow
CLECs to obtain space. So if you are waiting to get your status
in terms of CLEC, inter-connection agreements and others, you
go to the back of the line. It truly is a process of first
come, first served.
Most importantly, unlike the telcos, when we move within a
territory, we provide blanket coverage in the area. Blanket
coverage means that we will go into 90% of the central offices
within that area. For example, look at the San Francisco
region. We will not only cover the city of San Francisco, but
also 35 miles north of San Francisco, the East Bay and 50-60
miles south of San Francisco to cover all Silicon Valley. There
are 2.2 million homes and businesses within that footprint. Our
network is deployed and we pass over 1.9 million of those
So blanket coverage is a tremendous advantage when we deal
with Internet service providers, as they don't need to look at
a Swiss cheese map to determine where Covad has coverage. The
telcos are not going to deploy pervasive coverage, at least
from my experience so far. They are basically going to
cherrypick and only deploy DSL in high-density consumer,
high-end residential communities.
In your opinion, what are the benefits of DSL
technology? Could you tell us about the network you are
building? When will it be complete?
Knowling: The key benefit of DSL is that it is the
most economic choice in terms of high-speed bandwidth
deployment. When you look at the cost characteristics of
satellite, wireless and cable modems, the DSL model is
significantly cheaper to deploy than any of the others. For
example, on average it costs me about $100,000 to equip a
central office that serves 40,000 subscribers. From a cable
modem standpoint, to have visibility to all 40,000 subscribers,
you would literally have to lay infrastructure into all those
sub-divisions and business parks that the serving central
office is geographically placed to serve.
With DSL technology, I can put a DSLAM in that central office,
which allows me to equip every loop with DSL, leveraging and
using the infrastructure that is already in the ground and has
already been paid for. That is the beauty of the
Telecommunications Act of 1996: I can buy that unbundled
element, which provides me with a lower deployment cost.
In addition DSL can be deployed rapidly. By the third quarter
of this year I will have deployed DSL technology in over 1,000
central offices in the country. That will represent over 25% of
the addressable US market.
Regarding cable, let us look at AT&T and TCI. TCI has
cable infrastructure that passes somewhere in the neighbourhood
of 20% of the addressable market in the US. The upgrade cost
for AT&T to equip that cable plant with high-speed access,
telephony and others, comes to $6 billion.
And that is not the total upgrade cost: that represents only
40-50% of the upgrade of the cable plant. If you throw on top
of that MediaOne, when that acquisition or merger is finalized,
AT&T will then own cable franchises that will pass 62
million homes, which amounts to about 50% of the addressable
market. This will add several more billion dollars for the
upgrade of the MediaOne cable plant. So for some $10 billion
over five years you have managed to equip half the addressable
market in the US with high-speed access. If you compare the
economics of the varying technologies, clearly DSL is the most
Which DSL suppliers are you using?
Knowling: We are using a company called Diamond Lane,
which was recently purchased by Nokia. That is one of our
principal suppliers. We also use Cisco DSLAMS.
You just said that Covad has a significant first mover
to market advantage. How is Covad leveraging that advantage? Do
you believe that the company is advantageously positioned,
compared to ILECs and CLECs?
Knowling: The position that we have today is truly an
advantage. My business is doubling about every 90 days. The
ability to get in market first, get a large footprint, sign up
the premier ISPs and go to market with strong marketing within
a territory has proven to be a successful model. Quite frankly
those who come after me will find it a difficult proposition.
Once you have engaged in a strategic framework with an ISP, I
think that it really does prohibit the second player from being
as effective as you are. You have basically gone in first and
have first right of refusal in terms of who you want to work
with within a territory.
Compared to the CLECs and the incumbent telcos, the telco is
paying a lot of attention to the cable modem players. And their
DSL offering is geared towards consumers. And I don't blame
them for being focused in that regard.
I came from the telcos. I helped to shape a lot of strategies
that are taking place in at least a couple of those companies
today. And my point of view is that they have a huge revenue
stream with their private line services, their T-1, ISDN, frame
relay, fractional T-1. And I do not believe that you are going
to see them aggressively cannibalize those high-margin revenue
Covad has over 200 ISPs and 50 corporations as
customers. How quickly do you expect these numbers to increase?
Presumably you will focus primarily on the small business
market, owing to the international requirements of large
Knowling: That is correct. We currently have over 200
ISPs deployed and we have both national players, as well as
those ISPs that just operate within small regions. We have
added over 100 ISPs in the past three months, so growth is
Can you foresee a situation in the future, where you
would be competing against your current partners?
Knowling: I will tell you that we are going to be
very disciplined about what we do. If you can build a critical
mass business through indirect partners, it then begs the
question: why do I need to become an ISP, which will put me
competitively against the very people generating the volume? If
you think about the leverage of Covad using 25 ISPs in a
territory, compared to going clean on the street with my own
employees, sheer logic indicates that I am going to reach large
amounts of customers more quickly through the indirect model.
When do you expect to go EBITDA positive? Which
benchmarking tools do you employ to gauge various efficiency
and profit margins?
Knowling: Our model around EBITDA is that on average
it will take about 24 months within a particular region. We
have 22 regions that will be deployed this year. We are
currently in 11 regions. San Francisco is the region where we
have the most experience. And San Francisco will go EBITDA
positive in the third quarter of this year.
In terms of benchmarking tools I pay attention to @Home in
terms of my financial modelling. When you really parrot back
and look at the amount of network deployed, the ramp rate of
growth, the capex cost and acquisition cost of a customer, I
still believe that we are trading at a discount to them. But I
believe that if we can continue to get at least a few quarters
under our belt to be able to demonstrate to the street the
growth and the targets that are being achieved, I think that
gap will be closed.
I don't pay too much attention to the fibre-based CLECs,
because we are very different to them. They are voice plays
with central offices and other lines of business. So @Home is a
more comparable company. We have truly created a new company
and a new space. I would say that we are the new hybrid between
the Internet analyst and the CLEC analyst in terms of what they
are starting to chart. I would go as far as to say that we are
more aligned with Internet companies than with traditional
You announced in April 1999 plans to issue five
million shares of common stock. How will this issue dilute
earnings/dividends? When will the issue occur?
Knowling: Let me explain the reasoning behind the
issue. We have a few unnatural shareholders that are in the mix
of investors. As a part of the first high-yield deal, warrants
were given to the bondholders. Those warrants represent about
5,000,000 shares. On July 22 our lock-up expires and there will
be 37 million shares that are theoretically able to go into the
market when we reach that date. The majority of those shares
are owned by my strategic investors, Warburg Pinkus and
Crosspoint. A significant number are held by the founders and
employees of the company.
What I am trying to do with this offering - and there are no
proceeds that go to the company - is let the warrant holders
get out in advance of this lock-out. I am trying to manage the
number of shares that have the potential to hit the market all
at once that could negatively affect the stock price.
Are you planning to go back to the equity markets in
order to raise more funds in the near future? How successful
was the IPO in January 1999? How much money did you
Knowling: I plan to go back to the equity and debt
markets, depending on market conditions. The good news is that
I have the ability to go to either. Our IPO in January was
tremendously successful. We raised $150 million for the
company. I think that we priced at $18. I am not sure where the
stock price is today. It hovers somewhere between $80-100.
How is your customer strategy going to change going
forward? Who will be Covad's core customers in two or three
Knowling: I believe that you will see the
continuation of the indirect wholesale model of Covad, with a
tremendous amount of volume coming through indirect channels.
At the same time you will see other entities bringing customers
to the network. I believe that over the next two to three years
you are going to see a number of different industries and
sectors start to understand the benefit and the growth of the
Internet and how the Internet transforms these industries.
For example, the banking and health care industries are going
to transform themselves. You will see major players that will
not only facilitate the growth of DSL, but start to partner and
in fact subsidize the high-speed connection, once these
segments start to understand how they can change the
professional services of doctors, health care suppliers and
bankers. Just look at what is starting to happen in the
investment banking world. E-trade is just growing in leaps and
bounds to the point where there is even consideration of
allowing Internet trading around the clock every day.
I remember how just a year ago I could call my Schwab
investment broker and we could talk for 20 minutes about quotes
and a number of different things. Today you cannot reach a
Schwab broker. In fact the announcements on the network will
tell you that there will be a lengthy hold. If you want to
hold, they will even tell you that it will be a 20-minute hold.
But then they will say that you have access to the Internet,
where you can actually transcribe and conduct business.
I think that you will see entire businesses shift. And when
they shift, they will start to facilitate the acceleration of
lots of high-speed access over digital subscriber lines.
Do you aim to target the consumer market? If so, how
will you claw back the head start enjoyed by cable modem
providers, such as @Home, which has already secured about
350,000 subscribers in the US?
Knowling: Let us be bold and give them 1,000,000
customers. There are 55 million Internet users today. By 2001
there will be well over 120 million users. So from my
viewpoint, there is a huge opportunity in terms of the
available, addressable market that one can entice one way or
another through cable modem or DSL. In the next two or three
years it is not the issue of which technology the consumer will
choose. The more immediate issue concerns what is available.
Are you planning to offer voice services in the near
future? How do you view the voice over DSL debate?
Knowling: We are looking at voice. We are
experimenting with voice. Voice over DSL is not a future
technology: it exists today. The question for us is not will we
do it, the question is when. We are aggressively looking at it:
we are talking to a number of partners. We will be ready to
make an announcement one day relative to our entry into that
particular field, as we offer voice services over DSL. I think
that is going to make DSL terribly exciting, when you start to
add enhanced capabilities to the service. The DSL pipe, if you
will, has the ability to do a number of different things from a
content perspective that just enhances the value of the DSL
Concerns have been raised about the potential speed of
your offerings for long-distance traffic, compared to the reach
of the incumbents. You are currently using a modified type of
xDSL which is compatible with ISDN to gain access to customers
that cannot be reached using higher-speed xDSL. How will you
resolve this issue?
Knowling: DSL is distance-limited to about 18,000ft
for the higher speeds. For anybody over 18,000ft we offer an
IDSN equivalent, called IDSL. Covad's IDSL product is far
superior to ISDN. Let me take you through the profile of those
two technologies. ISDN is 128kbt, but it is not true 128,
because you get degradation of that bandwidth speed. ISDN is
not an always-on service. It is a dial-up service. ISDN has an
installation charge of anywhere between $700-1,000. There is
also a monthly recurrent charge of anywhere between $100-500,
depending on how much of a user you are, as this is a service
almost like pay-per-view.
If you look at the Covad 144kbt service, there is no
degradation of the bandwidth - it is always on and is secure -
the installation cost ranges between $400-600 and at retail
Covad's recurrent monthly fee is $95. With my ISP partners and
the volume that they can buy, the cost of that service can be
reduced to the $49-59 a month recurrent fee. So if you stack
those two services up, I have a far superior service than the
From a technology standpoint, the industry will solve the
problem of the limitations and reach of DSL. The second problem
you face concerns DSL through a digital loop carrier. I can
actually get through the digital loop carrier with my IDSL
product. But I believe that we will be talking a year from now
about a digital loop carrier being a moot point. The business
community will solve that technology hitch.
The management team of Covad comprises a mix of
technology and communications expertise, which represents a
break from the traditional base of telco-only executives. What
are the advantages of such a mixed management team?
Knowling: Well, it is by design at Covad that we have
gone and looked at the mix that we need to be effective in this
space. First of all we really are an Internet-based company. So
we felt that it is critically important to find people from the
data communications and computer industries. We have people on
our senior team that were literally giants in those fields,
from Intel, Adobe, Sun Micro Systems and Lotus. So we have a
fair share of people who truly understand the Internet, have
been vital players in key companies in terms of the emerging
technologies around PCs, chip technology, software development
and other areas.
Now at the end of the day from an operating standpoint we run
a phone company. So we have brought that element in from an
operational perspective. I have discovered, as I look across
the industry at other CLECs - and I have said this to the folks
on Wall Street - that I have assembled here at Covad probably
one of the strongest teams that you will find in the industry.
It is by far one of the greatest assets that we have. I often
believe that it is one of the reasons why we have been
successful: I have people who were senior executives in those
respective industries and know the space, know how to navigate,
can easily get into strategic partnerships, talk at the
executive level and understand how to grow businesses.
There has been a lot of investor interest in DSL
start-up stocks such as Covad, Rhythms NetConnections and
NorthPoint. Why do you think there has been such hype
surrounding these stocks? In your opinion what has accelerated
the popularity of DSL as a technology?
Knowling: Well I think that you have actually touched
upon the answers. DSL has actually been around for about 10
years. The telcos have been talking about DSL for the past six
years. You have companies such as Rhythms, NorthPoint and
Covad, which have literally taken all the hype around
high-speed access, digital subscriber line technology and we
have literally gone out and delivered on it. I believe that if
you look at those three companies and analyse where they are in
their development curves, clearly Covad is the leader. We have
a nine-eighteen month advantage over those two firms. And I
believe that the street is absolutely in love with the fact
that we are closely in line with the Internet. In fact the
bottleneck problems of the Internet are being solved by these
companies, which are bringing this new technology to the
market. It is terribly exciting.
How do you differentiate your strategy from Rhythms
NetConnections and Northpoint? Rhythms is already saying that
10-15% of their customers are using 1MB speed and above? Is the
demand for these high speeds out there?
Knowling: Well the sweet spot that we have seen has
been 384kbt. We have not seen the demand and interest of
customers that want 1.5MB, 4MB, 6MB. We just haven't seen such
How does my strategy differ from those two? From what I know
about Northpoint, they are a wholesale player. I am not so sure
that you will see them going after the remote LAN business of
large Fortune 1000 companies. They could be in that space, but
Covad doesn't run into them here. However, if you look at the
numbers, I think that they have about 3,000 subscribers. I
can't really tell if there is a mix of retail, as well as
wholesale, in their business plan.
I do see Northpoint in the ISP space. I will be in a market
talking with an ISP and inadvertently they will say that they
have received such and such prices from Northpoint. So my sales
force does run into the Northpoint gang.
Rhythms on the other hand, according to their first quarter
report, has about 1,200 subscribers. I think that they are more
of an integrated player, where they are doing more than just
DSL. They are doing managed data services, PBX extension. They
are doing a number of things, which are in my opinion difficult
to do. But I don't run into them in the market very much yet.
When do you think that the voice quality of Internet
telephony will be indistinguishable from that of traditional
Knowling: This is an interesting area I have listened
to voice over ATM and voice over IP. And I have to tell you
that the quality of the service is as good as anything that I
have ever seen. I believe that voice over ATM is going to be a
huge play. Currently I think that it is the better of the two
in terms of the quality of the service, just in the voice
quality. But voice over IP is going to be big as well. And I
think that both those capabilities are going to be tremendous
threats to the telcos.
Surely the Bells are likely to capture a large
proportion of your market share with their own DSL
Knowling: I think that they are going to be terribly
successful in the consumer space with no marketing, just
because they are telcos. But again, if you look at my strategy,
a 4-5% success rate in a market allows Covad to build a $1
billion business enterprise. So we don't have to go out and win
30-40% share of the high-speed game. That is the wonderful news
about the business model. You have two factors that are
working. Firstly Internet usage is still growing by a multiple
of ten. Secondly you have a technology where Covad has a first
mover advantage in terms of its ability to enter the market
first. Thirdly, the telcos are going to focus primarily on the
consumer space, because I do not believe that they will
cannibalize those high-end revenue streams that we talked about
In some regards, I believe that there is going to be a lot of
room at the table. There are going to be multiple winners. I do
not believe the folks who write on this particular topic that
there is going to be an ultimate winner. I think that they are
wrong. A number of businesses will be created, which will build
a lot of shareholder value, as there is that much opportunity
What are your hopes and ambitions for the company over
the next five years? Do you have any growth projections for the
company? Where will Covad be on the US telecoms landscape? What
trends do you see emerging in US telecoms in this time
Knowling: My hopes and ambitions for the company over
the next five years are for Covad to become the household word
in this country when people start talking about high-speed
access and enhanced value-added services. Today the darlings in
the space are companies such as AOL and @Home. We need to make
sure that there is a tripod and that people start talking about
Covad in the same breath. So I want to build a brand and
company that is truly viewed as the number one provider of
high-speed access on the earth. That is a pretty bold
statement, but I believe absolutely that we can enjoy that
I am not really permitted to talk about growth projections.
But James Henry from Bear Stearns has modelled out where we
will be in five years. His five-year model shows well over
1,000,000 customers in five years and a free cash flow business
that has revenue of over $1 billion. Those are his models. I
believe that those are public documents.
Finally consolidation is the trend that I see emerging in
telecoms. Unfortunately, five years from now I think that you
will only have three big telcos. It is pretty clear that
AT&T will be one of those three. It is clear a
south-western Bell will be another.
And at least in my opinion MCI WorldCom will be the other big
telco. So what does that say for the Bell Atlantic's and US
West's and Sprint's of the world? I don't know. But I think
that you will see a lot of great, big players in five years.
Maybe in five years AOL will be one of those big players.
Is there anything that you would like to add?
Knowling: There is one thing that I would like to
re-iterate in terms of more in-depth understanding about the
management team. In addition to the management team we talked
about, we have grown this company to about 450 employees. And
we have really brought in some outstanding talent in marketing
The infrastructure and foundation that you need to really
drive business growth are in place, because we have the people
in the company who have been there. We have true marketing
talent in the business and not telecoms marketing. We are
talking about people who have come from the software space,
from the PC industry, people who understand indirect marketing,
how to generate demand and other issues. This allows us to have
a better army and better arsenal than most of our competitors.