The Network Computer has been around for quite some time. So has been its most
notable cheerleader–Scott McNealy, President and CEO, Sun Microsystems. Never the one to
mince any words, The Original Brat comes out with analogies all his own. Here in his
inimitable style, in an exclusive to DATAQUEST, he drives home the point for a networked
future. Straight from the horse’s mouth, then:
I always say
to the CIO in every company I meet, including my own, that we are hopelessly behind in
taking advantage of the networked future. But how do you take advantage of a revolution?
My view is not unlike my attitude to favorite sport, hockey. It makes more sense to focus
on where the puck is going, instead of where the puck’s been.
Do The Math
I believe there are two returns on investment you might want to think about over the next
six years. First is the millions of dollars that you will probably spend to make a simple
date change to the new millennium. Second is the investment in turning PCs into real
corporate network citizens. One is zero return on investment, the other is massively
negative. The network computer is real and it has nothing to do with upfront cost.
Analysts have produced studies that show the five-year cost
of a PC is actually increasing, from $ 40,000 to $ 60,000. Some CIOs say their true costs
are more like $ 16,000 a year-or $ 80,000 over five years. This isn’t the hardware costs.
Administration costs dwarf the purchase cost. Why? Because every time one company goes out
and designs a bigger chip, someone else goes out and upgrades some supersetware to use it
all up. The industry then spends billions of dollars promoting it to the consumer. What
happens to you? Your employees hear about the latest software, and they go out and buy it.
Then they have to upgrade to the new CPU and get a bigger disk. Suddenly they have spent $
4,000, and you didn’t even know it. And then they do something in the new version and
their friend can’t read it, so that person is also forced to upgrade…and in the process
they have probably added no value to the company, your shareholders, or the customer. Try
to take it away and you have to get into a war of religious proportions. “Don’t take
my Mac away, don’t take my PC away!”
In my view, the two main elements driving corporate buying
decisions today are fear and anger. Fear that the supplier is lying to you about what his
products can do. Anger that when you install them you find out you were right.
Don’t get caught in the trap. Concentrate on what’s really
happening at the moment. The ‘Network is the Computer’ is happening.
Get On The Network, Get On Network Time
My advice is: Don’t upgrade the supersetware, downgrade to subsetware on a network
computer or a Java-based client. And invest in the network to make it work. The
availability of bandwidth and web technology is moving much faster than anyone can
possibly reorganize their companies. Think about the gigabit-miles of cable being buried
out there right now, and all the wireless technologies, and the cable modem companies that
already have a franchise into millions of front doors. Watch the new companies springing
up, and the existing ones rapidly getting into electronic commerce. The network is
The network even gets to you when you fly on an airplane.
Would you have believed that you could use a phone, fax, and do email on a plane even a
few years ago? Soon, hotels will be putting low-cost network computers in your room. We
all have to speed up the corporate body-clock to get on Internet time.
Internet years are just like dog years, seven to every one.
Build your employees a good network and they will use it. It’s like the US interstate
highway system. If you build the network, people will use it to operate more efficiently.
You don’t even have to mandate what or when or why they should send something on the
highway. People will just see the resource and go grab it.
Anticipate The Electronic Challenge
The digital industries will soon look much different than they do today-computing,
telecommunications, consumer electronics, entertainment. These industries are going
through a period of astonishing change, and it’s hardly started. I can’t see a company
today whose franchise isn’t being challenged and continually reengineered. The Internet is
already rapidly disintermediating the software business, taking the power away from the
publisher and into the hands of authors. In the old days, you had to be William Randolf
Hearst with huge printing presses, delivery trucks and paper mills to publish to the
world. Now you have an Andreessen who writes Mosaic, puts it out on the Internet for free,
and then creates a company that’s worth $ 200 million overnight. That’s zero cost of goods
sold, zero publishing cost to get your product out there.
And now the Java application industry is growing at a
phenomenal rate with venture capitalists creating startups all over the Valley. To use the
publishing industry as another example, in the network model, the advertiser even goes
straight to the author. The current rules of brand don’t matter online-Andreessen didn’t
create an ad campaign to make Mosaic a leader. This model is forcing a lot of the
middlemen and information brokers to think long and hard about what will happen as the
encryption and bandwidth problems get sorted out. And inside companies too, some CIOs are
already starting to use their intranet to disintermediate the brokers and the bureaucrats.
These technologies create incredible opportunities to drive cycle time out of your
business and improve communication.
This distribution model is zero cost to goods sold, zero
cost to get products online. So my advice is just go out and do it. You’ll never step on
the gas and hit something in front of you. You will only get run over from behind. Don’t
be the bug on the windshield of the car behind you. Be the windshield.
Stay Open, Retain Freedom of Choice
I often ask customers, what is your alternate sourcing strategy? Or have you chosen, based
on your office suite, to go with one vendor forever? Rather than locking yourself in,
decide to stay open so that you keep your freedom of choice. Force your vendors to compete
on the best implementation, not the specification. It’s already happening on the Internet.
Be careful when you visit a site that says: "Looks best under XYZ browser."
That’s fundamentally the wrong thing to do. It’s immediately limiting innovation.
your software is developed with a common denominator in mind. Keep your intranet working
on a minimum of three browsers. Don’t write any software that won’t work on Netscape,
Internet Explorer, and a Java device. Stop hairball computing, where you keep all your
data and applications on a few big mainframes, and then keep buying a bigger mainframe
hairball accelerator. Stop the hairball on the desktop where it takes 10 million lines of
code on a five million transistor chip just to do word processing. Spread data and
information out across the network where everybody can get access to it, with no single
source of failure. The hairball is extremely captivating.
In my opinion, a wordprocessor just needs five
functions-backspace, delete, cut and paste, save, and print. Anything more has absolutely
nothing to do with increasing value to shareholders or customers, or getting the job done.
Why give everyone of your employees a 32-bit, 32 MB, multitasking, multithreaded,
configurable operating system on their desk? That’s creating a lot of wasted disk space in
your company. These are just personal activity generators and not productivity tools for
I believe that the mission of the CIO should be to partner with the business unit to take
cycle-time out of existing business practices. Publish all your information on your
internal web. Don’t send it to employees, on paper or even email, because by the time they
can print it out and get through it, it’s already old. Give the user safe and instant
access to the network from a personal web page, from any machine, using any OS, at any
time, with dial-tone reliability. That’s a utility model of computing.
Think of the phone. If it’s not dial tone by the time the
phone hits your ear, you’re already angry. How would you communicate if I took the phone
out of your home? If I took it out of your office, would you stay in business? The digital
industries are converging on this utility model. Data tone or web tone will become as
commonplace as dial tone. What’s important about this model is that telephone companies
don’t give you a switch in your home and tell you to go administer it, program it, backup
and configure it. They give you the telephone handset. Give your users the computing
equivalent of the handset. Then let them make calls and have your corporate MIS department
handle the rest of it.
Write your content to and your applications in a secure
networked language, like Java. Put the subsetware applet on a server somewhere so that it
can be downloaded over the network just by doing a mouse-click on a browser. And make sure
it runs on any type of hardware…that’s Windows, Mac, Unix, or a router switch, a
printer, copier, cellphone…in your car, your TV, or your game machine. Stop re-writing
software applications across each platform you use, which adds zero value to your company.
Write your applications only once in Java and leverage all your technology in one go.
The network is the future. Personally, 20 years from now, I
think my own kids will ask me, What was that computer and why was it so big and ugly? I
expect to have to describe to blank stares why this strange box was needed to do something
they will take for granted on a network-through their pager, watch, ring, a
kiosk…whatever incarnation the device takes.