|
|
|||||||||||
Long-term, standards-based, acquisition plan for of total cost minimization.
The Holy Grail of the commercial computer department is standardization. Standardization simplifies and orders the chaos now commonplace in the world of network computing. Even small steps toward uniformity* or simplification of variety make a profound difference in long-term costs.
I. The Introduction/Discontinuance Cycle. Continual change and evolution have created an installed base of enormous variety. The infrastructure is continually fragmented by constant change.
A. Current main-stream sources of computing hardware have profit incentive to change products rapidly. The only way to continuously sell computer hardware is to inflate the specifications and discontinue models rather than selling them cheaper over time. If standard equipment were made available over time, it would cannibalize the sales of the newer, more expensive, product lines, with its appropriate reduction in revenue as prices fall. Every sale of the "old" model is at the expense of a potential new model.
B. The agenda of the distribution channel, from manufacture to delivery, is to maintain and grow sales. Every introduction of marginally improved models perpetuates the variety. Data Processing executives are forever struggling with the inevitable jumble of diverse equipment and software. The pace of introduction and dicontinuation quickens, and the complications exponate.II. Common Management Strategies. Many solutions that have been implemented to deal with this problem are half-measures, at best.
A. Many companies are currently implementing a "cascade" approach. Targeted acquisition based on specific user-initiated requirements, and ultimate redeployment in a domino fashion down the system. This can be very costly because each seat requires multiple de-installations, re-configurations, software and data transfers, physical moves, and reinstallations. The labor time involved and lost productivity costs, while difficult to measure, are quite staggering. This does not solve the long-term support cost issues inherent in the typical distributed computing model.
B. "But what if I fall behind the technology curve?" Applications under consideration cannot depend on technologies that doesn't exist yet. State-of-the-Art is impossible to maintain and folly to chase.III. Proposed Long-Term Hardware Solution:C. Leasing is just a way to finance the cycle, but may be an important (if expensive) tool. It can complicate the scene with a variety of equipment and configurations with different lease anniversaries.
D. Outsourcing only shifts the support burden to another entity. That is, the problem isn't going away.
Each step in the process is impotant to the process of creating and maintaining degrees of uniformity. A change in model number at the end of the manufacurer's or distributor's signals a shift to the next phase.
Phase I. Acquire a specific model type new from the manufacturer while supplies are available through normal distribution channels. Keep the configurations stable.
Phase II. As these models are discontinued, continue to acquire the same model type new as available through the surplus market. As new models are introduced, the older models, while still new, are available for purchase through the "gray" market. These are heavily discounted and may carry full warranty.
Phase III. Acquire the same model type refurbished to original specification. If it's identical to the ones you already have, it's any easy fit. Product life cycles get shorter each year. Dependable availability in the refurbished market is vital.
Phase IV. As appropriate, return to Phase I and repeat the process.
IV. The Justification:
Acquiring a standard machine type
and model of computer assets over time achieves four great goals:
a) It's just like the last
one, so you'll know its exact configuration, which means that
b) You can support it over
time, and
c) The hardware and its support
gets cheaper as you go, and
d) You are not at the whim
of the manufacturer's or distributor's discontinuance cycle.
The cost of supporting uniform networks goes down dramatically over time. When you're ready to establish the next standard, take the next jump. Leapfrog when you're ready. It's much easier to install, configure, replace, redeploy, sell, or upgrade machines that are all the same.
Because prices for commonly-used models decline significantly in price over a fairly predictable market-value life cycle, total cost over three years (or more) will be significantly lower for the following reasons:
i) Because of your standard platforms, support costs decrease dramatically over time as you climb the learning curve. (Examples: set up, software and device driver configurations, error resolution, diagnostics, etc.)At the end of the planned life cycle, begin a new cycle based on then-current platforms. This paradigm applies to PCs, printers, networking gear, communications products, monitors, etc.ii) As time goes on, you take advantage of price decreases and market value deterioration while maintaining network uniformity.
iii) Your purchases can be made with an eye on attrition of your current installed base without commitment to onerous safety stock levels usually necessary to assure network uniformity.
iv) Because of the timing of your acquisitions, the present value of cash flows will decrease your total cost to a significant degree.
v) Maintenance of the installed equipment becomes much less expensive as replacement parts become more abundant. (Examples: disk drives, power supplies, internal cards, etc.)
Adherence to standards over an asset's life cycle and can save an immense amount of money and time --the two assets in shortest supply. Plan ahead to achieve and maintain network standards with continuous, standard, uniform computer hardware.
*The process of initiating and maintaining uniformity requires active restriction of model specifications introduced into an organization. The concept of is an integral part of minimizing total cost while "freezing" current variety until a new standard is introduced by management. It is not always possible to have absolute identicality of all equipment; It is a long term goal in reduction of variety and staunch issestance on specific platform adherence.