Success Through Partnerships: Sybase KK
An Interview with Yoshimi Ogawa, President of Sybase KK
In November 1994, Computing Japan met with Yoshimi Ogawa, President of
Sybase KK. Ogawa-san joined Sybase KK as its president and first employee
in August 1992. He has nearly a quarter-century of experience in foreign-
affiliate high-technology firms, including an assignment with IBM Japan to
IBM's Asia/Far East headquarters in New York. Prior to joining Sybase KK,
Ogawa-san was president of Prime Computer Japan.
interviewed by Wm. Auckerman
Computing Japan: When was Sybase Japan established?
Yoshimi Ogawa: Legally, Sybase KK, a subsidiary of Sybase Inc., was
established in March 1991. It was a paper company, zero employees. The
practical start was August 1992. That's when I joined Sybase KK; I was the
first employee.
CJ: Why did you leave Prime Computer Japan to join Sybase KK? What was it
about the company that attracted you?
Ogawa: That's an interesting question, because before I joined Sybase, I
didn't know anything about the company. I was president of Prime Computer
Japan, and through that company I was a member of the American Chamber of
Commerce in Japan. At least a couple of times a year, other ACCJ members
would ask me to consult for companies attempting to start a business in
Japan - companies like Sybase.
My parent company said that as long as the company was not a competitor
of Prime Computer, I could do so. So I used to consult with visiting
executives over a cup of coffee for a couple of hours at no fee. Sybase
was one of these. About the time I was talking with a Sybase executive
visiting Japan about what they were doing, Prime Computer Japan merged
with ComputerVision Japan, a CAD/CAM software company. CAD/CAM software
was not my expertise; it was a completely different field for me. That in
part prompted my decision to leave the company.
Also, when I had joined Prime Computer Japan in 1986, their annual sales
were about ·2.4 billion. I said that I would make it a ·10 billion
company; that was my target - not ten years of service or something like
that. Fortunately, five years after I joined, the company made ·10.5
billion. So I felt I had accomplished my objective.
Simultaneously, Sybase approached me. We talked, and it took a few
months, but finally I decided to join Sybase. Before that, I didn't know
about Sybase at all. I knew Oracle, though, because Prime Computer Japan
sold several types of Oracle software.
CJ: How would you describe Sybase, and what are the major products that
your company sells?
Ogawa: We call ourselves a client/server system architecture company. We
promote client/server systems to help the customer to be more competitive
today and in the future.
To promote client/server systems, we sell four program areas. One is a
relational database management system, which is probably more than 70% of
our business source income. Another one is application development tools -
our customers use these tools to develop their applications, whether
character-based, GUI-based, Windows-based, or multimedia-based. We have
several types of application tools.
A third area is connectivity products. Client/server is growing,
encompassing multiple departments; enterprise-level client/server is
becoming essential. Japan is still a mainframe-oriented country, though,
and it would be impossible to develop an enterprise client/server without
getting involved with mainframes, so we have been providing connectivity
programs to leverage legacy systems. This connectivity product area is
important.
The fourth area is system management software. These four areas are all
software; we do not have any hardware product.
CJ: What is your breakdown of sales by operating system: UNIX versus
Windows software?
Ogawa: In Japan, 99% is still UNIX-based: Sun SPARC, or HP, IBM, AT&T,
Sony, DEC, and NEC UNIX systems. We have announced a Windows NT version,
and a NetWare version and OS/2 are coming in the first quarter of 1995.
These probably will be taking 10% or maybe 15% in the next couple of
years.
We do not have any of our products running on mainframes; in that
respect, we are completely different from Oracle.
CJ: What is the market share currently held by Sybase KK?
Ogawa: The Sybase market share in Japan in 1991 was 7%. In 1992, it was
11%. And in 1993, 15%. We're hoping 1994 will be more than that. My target
in the near future is to capture at least one-third of the Japanese market
share.
CJ: Dataquest figures show that Oracle and Sybase are the only two
companies that seem to be increasing their market share for the RDBMS area
in Japan. Is that because you two have better products?
Ogawa: Particularly in Japan, two things are important to get market
share. One, as you said, is that we need to have competitive products. But
another important factor is support. Japanese customers are very serious
about support. Not only product quality, but also technical support after
the customer implements our product is vital. These are the two keys to
expanding business in Japan.
CJ: Sybase KK has been expanding rapidly over the past two years, I
believe.
Ogawa: Yes. At the end of 1992, we had a total of just 12 people. In 1993,
we jumped up to 52. As of October 1, 1994, we had 140 people, and by
January 1995 we will have probably about 180.
CJ: What is your employee breakdown?
Ogawa: The majority of our people are engineers. We have four basic types
of engineers. One we call pre-sales engineers; they help customers
determine their needs and then decide what solution is best. They do
benchmarks and demonstrations, and provide technical support before a
customer decides what kind of product to buy.
Another group of engineers provides professional support. Once a customer
decides to buy a Sybase product, they will consult with the customer
before installation about how to design the database, how to use it. They
give the customer advanced training, that kind of stuff.
A third type of engineer gives technical support. They provide the needed
daily support - if the customer makes a mistake, or the software has a
bug. I call these the "firemen."
A fourth type of engineer does corporate R&D support. They help with
Japananizing products - how to localize for the Japanese market and local
customers' needs.
CJ: Where are your employees located? Do you have branches around Japan?
Ogawa: Most are here in Tokyo now. But we opened an Osaka branch in
October 1994, and by the end of 1995 we will probably have Fukuoka and
Nagoya branches.
CJ: I know that you're often approached by companies who would like to be
business partners. How do you handle these types of requests? Do you have
criteria for who gets in and who doesn't?
Ogawa: Very few Japanese customers will buy a system by themselves: that
is, RDBMS from company A, hardware from company B, and a network from
company C. The majority of customers want to buy a solution instead.
Sometimes they don't care whether the hardware is this company or that
company. They say, "Here is my target; this is what I want to do. Give me
a solution."
Since Sybase does not sell hardware - only our software products, our
service ó it is very hard for us to bring a total solution to customers.
That's why we implement that portion of the business with partners.
A second reason for having partners is that while Japan is not so big
geographically, the size of the economy is huge - a couple million
companies, from the small to the large. If you want to cover direct sales
to that huge market, you have to add a lot of people. It's very hard to
do.
In the IT market, we have to train our people quickly and deeply - so we
have to sacrifice something. It's a timing issue. Having a partner, an
outside resource, is a good combination. They help us by having a good
reputation, or selling A hardware or B hardware.
CJ: What kind of partners are you developing?
Ogawa: There are three types. One is geographical, as I said. We find
partners who can help us cover Japan geographically very well.
The second type of partner is hardware based. Many of our partners sell
multi-vendor hardware, but they have their favorite target. One of our
biggest partners, Itochu Techno Science (CTC), sells lots of Sun SPARC
workstations; they are focusing on the Sun market. Other partners are
selling NEC, or IBM, AT&T, Sony, etc.
I tell my people, "Hardware is our river." This is the Sun river, this is
the IBM river, and so forth. The important thing is how to put our product
on that river and go to the market smoothly and productively. So on each
river, we have a favorite ship, which is our partners - CTC is a good ship
in the Sun SPARC river. So product line is the second type of partners.
The other type of partners is by industry. We have maybe 20 or 30
partners, but one partner's favorite industry might be finance, or
manufacturing; for another it is dealer distribution, or service. These
partners allow us good coverage by industry type. What we try to do is to
avoid unnecessary competition. If I have 10 partners who are selling only
the Sun workstation to the manufacturing industry, they may have to give a
heavy discount to compete. But ultimately, the customer won't be happy
because if the vendor can't make a good profit, the vendor won't give good
support after they have installed our product. We try to avoid that and
still cover all the industries.
CJ: Do you have any plans for acquisitions or joint ventures with Japanese
companies?
Ogawa: In Japan, we are not thinking about this significantly yet. Maybe
within the next couple of years. The client/server market in Japan is
expanding, but it is only just starting right now. Japanese customers tend
to be very conservative. They look at US trends, at European trends - then
they will implement the same direction in two or three years.
When our market starts growing rapidly, one year or two years later, I
may think about joint ventures. But not today, because even on the vendor
side, not to mention the customer side, very few people have expertise in
the client/server area. That is a Japanese market problem.
CJ: In this kind of environment, it must be very difficult to train users
in the technology, and especially the tools, that they need to use Sybase
products properly. What kind of training do you provide?
Ogawa: The majority of training is done through our partners. Each partner
has their own education courses: basic training courses, advanced courses,
and so forth. Our first target is to train the trainers about our
products.
We have established a professional services division who are doing
consultation and training courses. Sybase has been spending a lot of
money, a lot of energy, a lot of people on training not only our employees
but our partners. Although training is aimed at our partners first, if
necessary we'll do the end users too. The majority of end user training is
done by our partners now, though.
That is very important.
CJ: Since you are growing so fast, is it difficult to find good employees
here in Japan?
Ogawa: Three to five years ago, yes. Today, no, because the Japanese
economy has been struggling the past couple of years. The bubble economy
collapsed, and many companies are now cutting costs. So the very
aggressive people are looking for a new and challenging type of job. For a
company like Sybase KK, it's a very good opportunity to get good people.
CJ: Do you find that most of the people coming to you have a solid
background, or do they need a lot of training?
Ogawa: If we hire new graduates from the university, we need to train
them. But today, we are hiring career people; so just a minimum of
training is all that's needed. Of course, continuous training is necessary
for the whole company, because we are adding new functions, new products.
CJ: Speaking of new products: You've said that the Japanese market is only
now becoming aware of client/server and RDBMS solutions. How has this
affected your product lineup?
Ogawa: We have been lucky, because Japanese customers have waited to
implement client/server. Our competitor, Oracle, started in the Japanese
market two years before us, which is to their advantage and our
disadvantage. But the market was initially quiet regarding client/server
implementation - that has helped us.
We are promoting four product areas: relational database, connectivity,
application development tools, and system management. The relational
database management systems will continue to be a majority of our
business. Based on these sales, sometimes the customer will buy an
application development tool from us, but RDBMS is our core product.
CJ: What products are you working on?
Ogawa: Sybase announced an SQL server in 1987. In 1993, we announced a
major product called System 10. The main difference is that, before System
10, it was a client/server system. Now, it is an enterprise client/server
system - it covers the whole enterprise.
Client/server used to cover just selected people, or selected
departments, not the whole enterprise. Today, though, we expect a
client/server system to contribute to an integrated corporate system, to
increase overall productivity and cover all employees. We are also
starting an integrated enterprise client/server system that also covers
consumers - a company's end users.
To make an enterprise client/server, you need scalability, more security,
and greater performance. System 10 covers all of these. Also, since
client/server is now being implemented to employees who are not familiar
with information technology, you need an easy-to-use system. If you are an
expert in information technology, a character-based tool is good; but to
go to the receptionist, each salesman, each engineer - you need a
graphical user interface, or a multimedia-type system.
CJ: Do you develop any software in Japan, or do you just localize US
Sybase programs?
Ogawa: For the coming two years, we may not develop any new software in
Japan. We just modify or expand existing functions. Maybe in another three
to five years we'll cooperate with our partners and develop new software.
But for the coming two years, we will import software from the US, our
parent company, and modify that to fit the Japanese market.
Sybase's strategy - not only Japan, but the US ó is that we do not touch
the applications. We do not do any application business; that is our
partners' role. This is completely different than how our competitors
operate, and it is why partners are very happy to do business with us: we
do not steal their business in the future. The application business
belongs to our partners; we help them to make more productive, better
applications. That is the Sybase strategy.
CJ: There seem to be very few successful Japanese software companies; most
of the popular packaged products are localized US software. Do you have
any ideas about how Japan can regain its own in the software industry?
Ogawa: I think we have to change the education system. That will take
probably one decade, maybe two. Speaking from my own experience - I lived
in New York for a few years, and I sent my daughter to an American school.
I was very impressed, because the teachers I met at the elementary school
would say, "Your daughter is very good in mathematics." I could see the
scores, and other items like English were very bad. But the teacher
wouldn't focus on that, saying instead, "She's a human being, not
almighty. Look at her strong points. Let's talk about how to improve her
strengths." This is my view of the American education system.
But back in Japan, the teachers said, "Yes, your daughter's mathematics
score is good, but look at her language. Poor. Let's think about how to
improve your daughter's language score to bring it close to the average."
The Japanese education system tries to improve the average score. As I
experienced it, the American education philosophy was to work on the
strengths or good points, not to think about the average. That's what
makes a Bill Gates, or John Scully, or Bob Epstein (who co-founded our
company). We've got to change the Japanese education system.
Over the past 50 years, after World War II, probably improving the
average score was the best solution for us. That's why Japan has been
successful in producing quality electric appliances, or automobiles. But
from now on, it is software - and for that, you need inspiration, or new
ideas. Producing good software is different than producing an automobile
or camera. Much different. Until we can change the education system, I
don't think Japan can be successful in making software.
Even Ichitaro, the best-selling Japanese word processing software - if
they had started in the US, against American competition, I don't know if
it could have been a successful product.
CJ: Getting back to Sybase KK, what have been your major business
achievements?
Ogawa: One month after I joined Sybase, in September 1992, I was asked to
make a presentation about Japan for our board members. I still remember my
challenges. To make Sybase KK successful in the Japanese market was number
one. The second one was to bring Sybase in Japan to 15% to 20% of total
Sybase revenues. KK's level in 1993 was only 4% of total Sybase; for 1994,
it will be 5% to 6%.
Sybase Inc. is growing at over 60% annually; Sybase KK is growing at more
than 100%. So even growing from 4% to 6% has not been easy. But I want to
eventually grow this company 15% to 20% of total Sybase revenues, because
Japan's GNP in the world is about 17%. Hopefully, we will get at least a
35% market share in Japan.
CJ: Do you have any final words?
Ogawa: Yes, let me add one final comment. Over the past 50 years, Japan
has built a very successful production system. That's why today, when the
dollar goes to 98 yen - manufacturers are struggling, but they are
surviving. That proves that on the manufacturing side, the production
side, we have achieved very good productivity.
But if you look at the office side, the white collar side, productivity
is very poor. That's the challenge Japanese society is now facing over the
next decade. And client/server systems, from the information technology
side, can help us to become more productive. In that sense, Sybase can
contribute to the Japanese people. That's my intention.
Sybase: the interoperability leader
An interview with Skip Glass, Vice President of Product Marketing, Sybase Inc.
Prior to joining Sybase Inc., Mr. Glass was Northwest area manager for
IBM's client/server computing division. His previous positions at Sybase
include director of product marketing for interoperability products and
group director of product and platform marketing.
interviewed by Terrie Lloyd
Computing Japan: Looking at the market from a global perspective, I know
that Sybase has been adopted by many banks, but where are you headed in
Japan?
Skip Glass: Banking was our first segment, it's true, but we have a large
percentage in manufacturing, process oil and gas - we also have a large
presence in health care and retail. So our mix is fairly diverse. No one
industry is more than about 15% to 20% of our overall makeup, so we aren't
reliant on any one industry.
CJ: Let me ask about your Powersoft acquisition. How long before the
announcement did you approach Powersoft, and why did you target Powersoft
in particular?
Glass: We've been working with Powersoft for a number of years, but we
actually decided only three weeks before the announcement that we wanted
to do it. We looked at a variety of issues. We were about to come out with
Build Momentum, which was our front-end tool, and our best estimates
showed we might get $20 million of revenue from that tool in the first
year. With Powersoft, we think we may be able to get as much as a couple
hundred million. And they also had a low-end distribution channel, they
had a single-user based SQL server - it filled those few holes.
It also gave us size. We believe that the mainstream customers in
client/server aren't going to want to piecemeal the parts; they're going
to want an integrated set of solutions from a company, and size is going
to count. They're going to look and say, "Sybase is the seventh largest
software company in the world; they're safe." If you're going to compete
against the Oracles and IBMs and Microsofts of the world, size counts. You
just really need to bulk up. We took a look at the number of things we
needed to solve as a corporation, and said, "Bang, we can knock three of
them off the agenda with the acquisition of Powersoft."
CJ: Regarding your competitive position in the Japanese market, the
question seems to be how you are going to maintain momentum since it has
got down to a two-horse race between Sybase and Oracle. Oracle's model has
pretty much been to port the product to every platform they can think of,
and then train each of the small camps of users who support those hardware
platforms on how to use the product. So they have a very vigorous training
program. Sybase, on the other hand, seems to work more closely with its
vendors, moving some of the training and sales functions off on to the
vendors. Would you say that's a fair assessment of the two models?
Glass: I'm flattered you would suggest that it's a two-horse race.
I would say that our relationship with vendors is tighter than Oracle's,
because we depend upon them to a greater extent to partner with us in the
selling process. However, we do have a very strong sales force that we
have developed here in Japan. We have basically tripled in size in the
past year as it relates to our team here in Japan.
CJ: Overall, how do you differentiate yourself from your competitors?
Glass: If you look at a worldwide perspective, we're much bigger than
niche now. Our major three competitors would be Oracle, Microsoft, and
IBM. We're also a partner of Microsoft and IBM, so we compete against them
in some arenas and partner with them in others.
If you look versus Oracle, at the Oracle setup solutions or data movement
applications - when you buy their tools, you must work with the Oracle
database. Our tools, Powersoft, are independent of the Sybase database.
Our interoperability solutions are independent of our database and will
continue to operate with multiple data sources. So we remain an open
choice, while Oracle forces you into their architecture.
In addition, Oracle is in the application business, and they actually
compete against the application providers. It is our intent not to be in
the application business. Therefore, we believe that application providers
will want to run first on Sybase rather than compete against Oracle.
CJ: And why would a potential customer choose Sybase instead of
Microsoft?
Glass: Microsoft is not open. You have to operate on one operating system
(NT), one object paradigm (OLE), one systems management paradigm, one
messaging paradigm.... We support a broader set of choices. That doesn't
mean we won't sell a heck of a lot of our product on the NT. This would be
the standard Sybase SQL running on the NT; it's not a watered-down version
like Oracle has.
Microsoft historically tends to run 20% to 30% cheaper than us; they seem
to have a desire to do that. But let me tell you: Microsoft's SQL server
is based on a 4.2 version, a two-and-a-half-year-old version of the Sybase
product. If we can't continue to retain a lead with that....
We've got an interoperability lead. We have a systems management
solution, which is shipping, and Microsoft's is not. We're the leader in
replication, and we have a 10.0-based version of the server; they've got a
4.2 base. So I think a 20% to 30% price differential is well worth what
you're getting in incremental technology when you buy a Sybase solution.
