08072007 Pacific Crest Technology Conference Christine Heckart
Pacific Crest Technology Conference
Tuesday, August 7, 2007
BRENDAN BARNICLE: Good afternoon. I'm Brendan Barnicle. I'm a software
analyst here at Pacific Crest. And this afternoon we're pleased to have Microsoft with us.
We've got Christine Heckart, who is a GM of Microsoft's TV with us, as well as Joe Binz
from Investor Relations.
We know we're going to do this the typical way we've done all day. I'll start with the
questions, and open it up for the group. Joe, just a quick housekeeping question we've
been asking all the companies this, it's probably more appropriate for you than for
Christine. With the kind of meltdown we've seen in financial services, and sub-prime
market, asking all the companies about concern about exposure to financial services,
particularly for the second half of the year?
JOE BINZ: Yes. It's hard to tell what the exact impact is going to be, but obviously
from our perspective we're large and diversified across a number of different verticals in
the commercial side, so we don't expect it to have a significant impact. The other thing
we have going for us is, we have a really rich product cycle, both in terms of what we've
launched and what we're going to launch in the next year around some of our server
products, and we think that will cushion us as well.
BRENDAN BARNICLE: Great. And, Christine, we saw you guys win an appeal
today, or yesterday I guess, in the Alcatel-Lucent decision, and interesting in what the
ramifications of that are, and is that something that I guess will be likely appealed again,
or are we kind of finalized on that?
CHRISTINE HECKART: Well, I'm in the TV business. I'm not one of the lawyers.
But I would say that it's likely for Alcatel potentially to come back, but Microsoft thought
it was a really good verdict. We were happy about it.
BRENDAN BARNICLE: Okay. The last time we had you in front of investors, it's
been about nine months. I was hoping we could just kind of start with an update of
what's gone on in the IPTV business at Microsoft, and the latest developments, and latest
CHRISTINE HECKART: Okay. If you haven't touched this business for nine months
to twelve months, a lot has changed. A lot has changed from Microsoft in the TV space,
as well as for the IPTV industry as a whole. If you go back this time last year, we had a
handful of pretty big service providers in the market in beta trials, and if you look at the
business today, we have 23 customers around the world, 23 different markets. Most of
them are the incumbent provider in their markets, the incumbent telco. Ten of those have
commercially available services, the others are in some form of trial, or limited
availability. And several of them have pretty good clips going in terms of subscriber
ramp. So AT&T is the one that most people here seem to be generally interested in, since
we're in the U.S., and they disclosed at the end of last quarter that they had 51,000
subscribers at that time that was ending June. They were ramping over 2,000 a week at
that time. They've got four million homes passed with the network now, on target for
eight million by end of next year, and their target is to reach 10,000 subscribers added per
week by the end of the year, and super confident that they're going to do that.
BRENDAN BARNICLE: Great. Then also, over the years, one of the big concerns is
then whether this works, in particular scaling up and scaling down. I was wondering if
you could give us an update on where you are in all those scale issues?
CHRISTINE HECKART: Yes. So really there is really a lot of skepticism plaguing, a
lot of concerns plaguing the IPTV industry in general, and most particular Microsoft.
And those concerns range from will service providers even do business with us because
we represent a potential threat to them, not just a potential partnership? Will we be able
to deliver software that is carrier grade? I come out of a carrier environment, I
understand carrier grade software, that's what I've lived most of my life in one form or
another, and Microsoft hasn't traditionally operated in the backbone infrastructure of a
carrier, and Mediaroom is the first time that we've really been in the backbone
infrastructure. It's a huge strategic win for us, and it pulls through a lot of other crown
jewels for the company like SQL Server, and MOM, and Windows CE, and so on. So
that was a big area of concern that a lot of the market had, could we operate there
successfully? So far, I think we've done a good job of proving that we can deliver carrier
And then there's tons of concern just around will it work in general. Can you do
something like high definition television over an IP network. If you go back three years
ago, there were more skeptics than not that if it was even possible technically, because
video is very demanding, and IP networks are designed to be bursting. One of the things
that I think has proven out incredibly well, not just to us, but in the market in general
around the world, is that you can do fantastic TV. In fact, some of the best quality video
in television that you can see anywhere in the world right now is being delivered over
And then the last area, as you say, well, that probably wasn't the last, there were a lot of
areas of concern if you go back even a year ago, but scaling up was definitely something
that everybody talked a lot about. The big concern at the time is, could you do
deployment totaling in the millions of users? Then when the market got a little bit more
sophisticated about how to ask that question, realizing that the way you build out these
networks is typically video head-ins that go into communities, and into cities, but you
need to be open to the tens, hundreds of thousands of users even in a particular service
group. I think the deployments that we've seen so far, and the internal metrics that we're
hitting have led us to have a lot of confidence that scaling up isn't really the big problem
that we face right now.
Scaling down as a business, different issue, we could face a lot of different kinds of
problems there, not just technical, but a pure business process, and how you support
small carrier. But the whole technical side of scaling up is not something that we
personally, we as a business, see as a major hurdle for us over the next couple of years.
Something could always come from left field. Once you get out more than a couple of
years, some things could change. But our competence level on the scaling question is
very high right now.
BRENDAN BARNICLE: What's the issue with scaling down?
CHRISTINE HECKART: A bunch of issues with scaling down. Some of them are, I'll
call them technical, but they're really more economic. You know, the way we built our
system was really for the tier one providers around the world. It's built with an
architecture that the reason why it scales well is because it optimizes for each of the
functions in the network, and it allows for optimization modularly around each of those
functions. That works great when you're scaling up.
When you're scaling down, it's kind of a different problem. When you're scaling down,
what you're trying to do is get every one of the functions onto maybe a single server, say,
in an extreme environment. We do that today in test labs. You can buy something from
us called a personal server, if you're going to go to a trade show, or if you're a set-top box
vendor, and you want to test set-top boxes, you get all the software suites on one server.
That's scaling down. But can you really deploy a service on that? No, that's never
designed to be infrastructure. So as you scale down, there are technical challenges. How
do you scale down and keep the performance high, but really start to integrate a lot of
different functions that today would reside, a software suite, on different server sets.
How do you get those really collapsed into a very small set of servers?
And then probably the bigger challenge is more business process. We're a very small
group, we're 600 people. We're dedicated entirely to the TV business. We're very laser
focused on that. But we still have a few -- we measure our sales force in the few dozens.
When we do a customer implementation, it's a pretty high touch environment still. Each
customer has their own network configuration, their own set of issues that they deal with.
We put a team there that they pay for. It's a very high touch kind of engagement model.
Good margins as a result. You know, great scaling up as a result. But we haven't really
figured out how do you cookie cutter that, and how do you scale the business down in a
way that allows for a low-touch model, and hitting a service provider that might see
BRENDAN BARNICLE: Christine, why TV? Why does Microsoft want to be in this
market? How does it fit strategically with things like Office and Windows?
CHRISTINE HECKART: That's like a whole conference in and of itself. So I'll give
you a two-minute answer. TV is a really important part of Microsoft's vision for
connected entertainment. There are 1.6 billion TVs in the world. Software is the next
evolution of the TV, and we see TV as really going all on demand except for maybe
sports and breaking news. And we really see software being the transformative next step
in the evolution of TV. That's borne out by a lot of the research that we do into where
consumer trends are going, and into the ways that people want to consume their content
across multiple kinds of devices. So, it's a super important part of connected
It doesn't stop with the TV. What Microsoft wants to do, and what we aspire to with
Mediaroom, is to really have whatever your favorite content is available on whatever
device you want to view it on, and whatever particularly time you want to view, and to be
able to move it around reasonably casually, and without having to have a degree in IT be
able to have your media with you wherever you are. So it's a super important part of
Microsoft's TV vision as well as the overall Microsoft vision, and service providers as
well. Like I said, this is the first time we've been in the infrastructure, so for Microsoft
the Microsoft TV division has two kind of important strategic spearheads. One is, we're
getting into service provider infrastructure. The other is we're getting onto the TV into
the living room, and out of the den, and off the PC.
BRENDAN BARNICLE: And so what does it do for a plethora of other products, you
mentioned earlier SQL and MOM, and some of the other things, how do those products
tie into the strategy?
CHRISTINE HECKART: Every time -- we don't base our strategy around that. We
don't sit around every day thinking, how can we sell more SQL Server, but we built the
product on top of Microsoft products, so every time we sell a license, we're pulling
through a whole lot of other Microsoft infrastructure, and it's infrastructure that has never
been in the central office, and in the back home network before. So it's a real good
proving ground, hopefully to give service providers confidence that they can base their
networks on server infrastructure. Google, and MSN, and Yahoo, and others that are
hugely scaled service businesses do that today, but the service providers never have.
They've always had very purpose built hardware, and it's pretty expensive, and this is
potentially kind of a new era in showing a different way to build a very scalable, very
reliable network, but using low cost, off the shelf hardware components.
BRENDAN BARNICLE: And is it -- can you give us the update on what the revenue
model would look like coming out of this business ultimately?
CHRISTINE HECKART: We don't, within E&D, we don't sub-forecast the individual
businesses outside of Xbox. But I can tell you a little bit about generally how the market
is being viewed by other analysts, the ranges, and how we make money. So if you go out
and look, you'll see ranges for IPTV opportunities, in the 30 million to 100 million
households. I would say the truth over the next two, three, four years is about in the
middle of that range. It could be better --
BRENDAN BARNICLE: (Off mike.)
CHRISTINE HECKART: Subscribers. Homes passed, they're already in the -- AT&T
has four million home passed right now on their way to eight. So, you can see lots and
lots of homes passed. How many total IPTV subscribers will there be worldwide, you get
very aggressive and bullish analysts on that, and you get very conservative I would say
we've seen that there's a lot of challenges, especially in the near term, for the service
provider as a whole to transition their businesses to being entertainment companies.
AT&T is doing that, and I would say game over for them. They have become -- if you go
walk the halls at AT&T, if you talk to them, they think, they act, they are modeled around
an entertainment business now as part of a quad-play. There are other providers around
the world that are going to have a much more difficult time and some of them will
probably make that transition successfully, and some won't. The way we make money is,
we sell the service provider, but it's a B2B2C model. We only make money if they ramp
subscribers. They pay us on a per household -- they license our software on a per
household basis, generally, and then they can have multiple set-top boxes per household
if they're buying a high-end SKU. So it's in our best interest to really see them be
successful in ramping their subscriber counts over the next several years.
BRENDAN BARNICLE: Any questions? Sure.
QUESTION: What is Microsoft’s share of set top box OS today, and outside AT&T
who are the other big Microsoft customers?
CHRISTINE HECKART: So we track, but we don't publicly share what we believe the
market share figures are. We look for you guys to do that. So I would encourage you all
to do it, because I would love to see what you think about the numbers. If you look at the
top tier providers around the world, the tier one providers, we have 23 of them. There are
400 to 600 service providers around the world. There are probably more than that, but
after about 400 it drops way off my radar screen. We're focused on the 20 percent that
are 80 percent of the market for sure.
We've got 23 of them today, they include AT&T, British Telecom, Deutsche Telecom,
Swisscom, Telcom Italia, the T-Online properties, Hungary, and Slovak, and all the
different places that they are, Tele Denmark, TDC, South Africa Telecom, Reliant in
India, which we believe is a pretty substantial opportunity, Chunghwa in Taiwan, and
there's a handful of those 23 that aren't publicly announced yet. Some of them are -- they
are almost always the incumbent telco in their market. There are two places where that's
not true, like some of the T-Online properties which are ISP challengers, but generally we
pick up the incumbent telco.
QUESTION: Just one other follow-on which is, of the sort of set top box makers who
have you guys had the closest ties to?
CHRISTINE HECKART: Yes, we have six - -we have a pretty broad ecosystem, in
general, it covers all -- from silicon all the way through encoder servers, set top boxes, et
cetera. In the set top box area we've got six partners today, I'll be challenged to
remember them all, but Thompson on an Intel platform, not the soft platform yet.
The rest of them are on the soft platform, and that includes Cisco, both Linksys, which is
the old KiSS platform out of Europe, and Scientific Atlanta, both of which are going to a
Cisco brand over time. Motorola, Phillips, which is BT, Pirelli, and I'm blanking on one.
But, generally the way our model works is that we will encourage a service provider
that's signing up to pick from the current plate of available set top boxes, because that
gets them into the market quickly. Then they'll each have some favorite in their market,
and so then we'll put them in line.
We've got a standard SOC, it's a Sigma SOC today, and we're working on getting another
one into the market. That's been a challenge for us, and something we've got to get over
come, so that we can help drive the cost point down, and the SOC is definitely one
important part of the -- it's a system on a chip, yes, sorry, a piece of silicon.
QUESTION: Is the problem with getting another one just that no one else -- (inaudible)
CHRISTINE HECKART: No, it's not. I don't know what the problem is, and I own
that area. It is -- there is just a lifecycle of getting the chip stamped out, and then getting
the software ported on it, and then getting it optimized. So we really drive the system
hard, and so we've got SOCs out there, and they do okay, but they really can't hit the
performance metrics that we're looking at being required to do the things that we want to
do, which are some pretty fancy tricks on the TV. And the Sigma performs wonderfully.
So we have a good solution in the market, that's good news. We need to get other
solutions in the market to really drive down the price overall in the ecosystem, and that's
an important -- set top boxes are one of the biggest -- probably the biggest element of
SAC, and SAC is one of the -- subscriber acquisition cost, and that's one of the biggest
elements of the overall business case, outside of just the network. So one of the roles that
we play, and one of the reasons that we do have such a broad ecosystem is to try to drive
costs out of the ecosystem, as well as try to get innovation through the ecosystem, and
use the innovation that they create for us.
QUESTION: (Off mike.)
CHRISTINE HECKART: Yes, the bill of materials right now in IPTV is still quite a
bit lower in the tens, at least in the low double-digits of percentages lower than what you
typically see in cable. The set top boxes are less expensive, because you don't have
tuners, it's all based on software. So we do a good job of having a pretty low bill of
materials now, and we got forward pricing on the chips. But, it's something we
constantly measure ourselves against, we constantly want to improve on, and service
providers want us to help them drive.
BRENDAN BARNICLE: Other questions?
QUESTION: (Off mike.)
CHRISTINE HECKART: Yes, T-Online France is, I think, still our biggest
deployment, and France is still the most competitive IPTV market in the world. Triple
Play there -- there are seven different providers, Triple Play goes for 30 euro. We hope
the rest of the world doesn't follow the French model. In fact, there is consolidation
going on in that market today. So I think we're going to see that kind of change over
AT&T is on -- they're probably second now, and they will absolutely be the biggest by
the end of the year. They are on a huge clip, really confident in their ramp rates, and
you're likely to see them be the biggest. They -- I don't believe they disclose, nor do we
disclose for any of our customers, their specific numbers on penetration. AT&T has
talked about the fact that either 80 or 85 percent of their uptake has been for their high
end package. So they're very pleased with, not only the percentage they're taking of
homes, and you can kind of do the math on that yourself, but the actual packages being
bought. They're not just going in and winning on price. They really are selling a very
rich package to these subscribers.
Now, Swisscom, a little bit different on that, and I'll just share this quickly. Swisscom is
a really small market. They're ADSL-based, so they push us in all kinds of ways that
others don't, which is to drive down the amount of bandwidth you need to deliver TV
over. They really stretch the limits on that. But, one of the interesting things we're
finding in Switzerland is they're getting a great baud uptake, and they're getting a great
baud uptake on their long tail, the bottom 250-300 titles are driving about a third of their
revenues. And if you translate that into profits, it's a much higher percentage of profit,
because that long tail content you don't pay a lot for, and yet you can sell it for a pretty
healthy amount on a rental basis. So they're really happy, and we see them as being a
good model of how you can increase RPUs using baud, and using the long tail, if you can
make it easy to find, and easy to rent, which Mediaroom does.
QUESTION: (Off mike.)
CHRISTINE HECKART: Great question. We track primarily three things that we
measure ourselves against, our own revenue, revenue targets for Robbie, margin that we
need to hit for Robbie, and then the thing that we track most closely, because it drives all
of that, and you should track, are subscribers. And what's the ramp rate on subscriber
And to answer your question about where we are, I would say from the original business
case that was put together a couple of years ago, I'd say we're about a year behind. If you
look at our numbers and what we originally had reported up to Robbie, Bill, and Steve as
our aspirations, if you go back, say, three years ago on a strat review cycle, we ended up
being a little bit later to the market than what we'd anticipated. We took a little longer,
and the market took a little longer to develop some of the key technology components.
So it was a little bit longer before we got our customers into the market.
So we're tracking pretty closely on, in some cases a little bit above, the ramp rates that we
expected, but we're about a year behind, and that's because we were a year late into the
QUESTION: (Off mike.)
CHRISTINE HECKART: I get the nature of your question. I'd slice that two ways.
You tend to see the deployments happen in urban areas. That's not entirely true, if you
look around the world, but it's a good generality, it's a good rule of thumb. So just by it's
very nature you find the network build outs happening in the urban areas, and then you're
correct that the closer you are to a serving wire center, your DSLAM, the higher the
bandwidth will be. Now, whether or not that ends up being important depends a lot on
the particular service and configuration.
If you're in Europe and you only have one TV, and it's standard definition, you don't need
more than three meg. Most of the providers there, if you go into France, even the ISPs in
France, are reliably getting 15 to 45 meg depending on where they're at. So three meg,
that's pretty easy throughout most of the world. If you're here in the U.S., you have an
average of three set top boxes, AT&T is getting 25 to 40 meg, and the variance is based
on how far away you are from the central office. But, does it work better, as long as you
have the bandwidth necessary it works the same.
QUESTION: (Off mike.)
CHRISTINE HECKART: So I think those are low for Verizon. I think those numbers
are pretty drastically low, but I could be --
QUESTION: (Off mike.)
CHRISTINE HECKART: I think it's higher than what you're saying, and I'm not going
to give their numbers for them. You can do talk to them. In general you'll find that if
you can utilize your copper backbone, even with a VDSL upgrade, or maybe ADSL 2,
you'll generally be anywhere from 20 percent of the cost, so 80 percent reduction to 50
percent reduction on a comparable fiber to the home. Now, with that said, I personally
don't take religion around fiber to the home, or fiber to the curb. I think every provider
will have a combination of both. Every provider knows they'll have a combination of
both. If you make that into a religious battle you're just being pretty short sighted.
The trick will be, have a single service, be agnostic about how it's delivered. And with
Mediaroom we're completely agnostic. We don't care what, you can have tin cans, and
strings, it doesn't matter, as long as you can hit the bandwidth and performance metrics,
we really don't care. In fact, even today in Europe we have a lot of places where we're
doing a combination DTT preview over the air, with IP back channel, and that's the BT's
entire infrastructure. And they don't do anything else.
BRENDAN BARNICLE: Thanks, Christine. Thanks, Joe. I know you guys are around
to do some one on one, but thanks for joining us. We really appreciate it.
CHRISTINE HECKART: Thank you.
JOE BINZ: Thanks for hosting us Brendan.