|
|
The Total Cost of Ownership (TCO) is
defined as the cost of procuring, deploying and maintaining
your management information systems. Several recent studies
have determined that the TCO per client is in the range of
$8,000 to $15,000 per year. While these figures are debatable,
there is consensus that the cost of maintaining a traditional
PC LAN environment is staggering.
Enter Thin-Client Technologies
Thin-Client technologies promise to dramatically lower the TCO
by reducing hardware, software and support costs through
centralized management and simplified desktop devices.
Unfortunately, TCO is a very elusive number that varies
dramatically based on several company specific factors. There
is some debate over what components should be used in the TCO
calculation, but the following list represents the major
components that should be considered.
- Hardware and software costs
- System, storage and network management
- General operations and maintenance
- Helpdesk
- Employee (self and peer) support
- Other hidden costs
Studies that have attributed 40% or
more of the TCO to end-user labor (end-user time spent
learning how to use their hardware and software, maintaining
there own systems and helping their co-workers) have been
accused of being conservative by some, yet an exaggeration by
others. We will come back to this debate in a moment.
Our perspective on TCO is slightly different as compared to
what we have seen in TCO studies and other publications.
Most people agree that the hard costs are easy to calculate,
and it is the soft costs associated with support and
efficiency that are difficult, if not impossible, to
determine.
The BLACK HOLE
At the center of the soft cost problem is the "Black Hole",
better known as Information Services in most organizations. We
use the Black Hole analogy because we have never ran into an
IS group that was over-staffed. In most cases, as IS adds
resources, they are immediately consumed by outstanding tasks.
The IS appetite for additional resources is typically
insatiable. In fact, we argue that you could double the
resources in almost any IS organization and still have more
than enough work to keep every one of them busy 40 or more
hours/week.
If this is true, then how do you determine the appropriate
staffing levels for your IS group? The answer to this question
is dependent on the senior level management philosophy within
the organization. Let's take a look at the common IS
management philosophies.
Philosophy #1 - Do Your Job
Some companies follow the philosophy that management
information systems are tools that allow employees to do their
jobs. The tools should be easy to use, and the employees
should spend as little time as possible learning how to use a
specific tool or supporting its operation. Instead, the
employee should maximize the time spent working on the tasks
for which he/she was hired. This requires an IS staff of
adequate size and capabilities to quickly respond to (and
solve) end user problems. This philosophy generally leads to
the creation of large IS departments that are also capable of
designing, deploying and maintaining the infrastructure and
quickly resolving system, network and end user problems.
Philosophy #2 - Your Computer Is Your Job
The other extreme is the management philosophy that computers
are becoming increasingly important in almost everyone's job,
and for this reason the employees should be constantly
striving to improve their computer knowledge. While you may
agree with this statement, the questionable part of the
philosophy is the belief that this knowledge is best achieved
by solving (or attempting to solve) most of their own
problems, thus spending less time on their primary job
function. In this case, a much smaller IS department is
required. When a user encounters a problem, they learn not to
expect the immediate support from IS, so they are forced to
try to solve the problem on their own or with the help of a
co-worker.
Philosophy #3 - Why Do We Need Support?
There is a third group that has also deployed small IS
departments because they don't understand the value that IS
adds to the organization. This group views IS an an overhead
expense that continues to grow even though the average cost of
computers has dropped by an order of magnitude in the past
decade. A common question asked by this group is "I don't
understand why we need (the size of) our current IS
department. What do those IS guys do all day anyway?".
Most organizations have a management philosophy that falls
somewhere in between these three models, but the
characteristics of one stands out over the others.
Effect of Management Philosophy on TCO
Looking back now to the problem of calculating the TCO, we now
understand why there is a debate over the cost of end user
labor. It should be clear that in the first management
philosophy (Do Your Job) the cost of end user labor should be
extremely low since there is a large IS staff readily
available to support the users. On the other hand, extreme
implementations of the second and third philosophies could
easily lead to end user labor exceeding 40% of the TCO.
In a traditional PC LAN environment you can lower the cost of
end user labor by adopting a management philosophy that allows
for improved IS support. In this case, the cost savings will
be at least partially offset by increased IS support costs.
Likewise, IS support costs can be lowered by pushing the
responsibilities back to the end users, resulting in increased
end user labor costs. The bottom line is that no matter which
philosophy you choose to implement, the support cost in a PC
LAN are significant and unavoidable. Shifting the support
between IS and the end users will do very little to lower the
TCO.
Will Thin-Client Technologies reduce the TCO?
YES. The only real question is by how much. All of the current
Thin-Client technologies (Network Computers, Windows
Terminals, WBTs, NetPCs) provide a reduction in the TCO
through metrics that are fairly easy to measure. Thin-Client
desktop devices have broken below the $500 price point and
provide a robust interface on a standardized platform that is
very easily maintained. Thin-Clients often require "fat
servers" and/or faster networks since many of the traditional
tasks have been offloaded to centralized devices for
processing and/or storage. Again however, the increased server
and network cost are easy to calculate, and most (if not all)
of these expenses can be offset through the costs savings
associated with the low-cost desktop devices.
The real TCO savings in Thin-Client environments comes from
the reduction in total support costs. Thin-Client environments
are designed to centralize and leverage management tasks and
reduce the need to visit the desktop. The desktop devices
themselves are highly standardized and easily replaced when
necessary with very little required configuration. Again, the
actual TCO savings will depend on the environment and the type
of Thin-Clients being deployed (NCs, Windows Terminals or
NetPCs).
If these TCO savings are not enough, could you imagine being
able to run any application on any type of client? Or better
yet, could you imagine your users achieving similar response
times from your critical applications regardless of location
(LAN, WAN, Internet, dial-up lines)? Moreover, could you even
begin to place a price tag on these capabilities?
Today's Thin-Client technologies are achieving many of these
goals, and they have the potential to go much further. As you
consider Thin-Client technologies for your organization, you
can go through all of the TCO calculations you want. No matter
what you determine, one thing is certain: someone will debate
you on the accuracy. Thin-Client is not just about reducing
TCO, it is an evolutionary change to the way we compute that
will:
- Reduce hardware and software costs
- Reduce maintenance and support costs
- Reduce LAN/WAN bandwidth requirements
- Reduced down time
- Improve application availability
- Improve scalability
- Improve performance
- Enhance security
Many of these factors are usually not included in TCO
calculations.
Deja vu?
Interestingly, there are also studies that have been published
recently that show that the TCO of a mainframe/dumb terminal
environment is significantly lower than PC LAN or
Client/Server environments. However, these studies have not
slowed the growth of the PC or breathed significant new life
into legacy systems. You (most likely) wouldn't trade in your
PC network for a mainframe because of the user friendly
interface and powerful desktop applications. The capabilities
of the PC outweigh the TCO argument, hands down!
Why then would you look at another new technology simply from
a TCO perspective? The answer is you shouldn't. Experts firmly
believe that Thin-Client technologies are going to make an
evolutionary change to the industry. They will not replace the
traditional PC to the extent that the PC replaced the
mainframe/terminal, but the change will be widespread and
dramatic. If you are hung up on the TCO calculations you are
missing the boat. It's time to open your mind to the true
potential of Thin-Client.
top
|
|
