HPMD Bullet: Don't Throw Iron At It

HPMD Bullets

BULLET #23
DON'T THROW IRON AT IT

I don't know what it is about data processing departments and hardware ("iron" in data processing lingo). We've all seen it happen. As soon as demands increase, they're quick to add another computer. The smart ones anticipate capacity and add more horsepower in advance of the demand. At worst, the newest-fastest state-of-the-art hardware is added to replace the old clunker we have. At least its got more MIP's per dollar. But it's more dollars. This is throwing iron at the problem.

The following are some thoughts on some other options to an iron mentality:

1. Prove exactly where the bottleneck is; and where it isn't. Are you certain?

2. Prove why and how the proposed solution will in fact solve the problem. Are you willing to stake your salary on it?

3. Prove it's the only solution, or at least the best one.

4. Look hard at the "in-between" solutions, the less elegant, more pragmatic, less costly ones. It may be ugly, but it does the job.

5. Start some competition on who can come up with the best ideas for wringing more horsepower from the system we've got.

6. Buy the "in-between" solution used rather than new. It's not as shiny, but it gets us where we need to go.

7. Try the less automated, smaller options first. Can the system can be run on a smaller machine? on a LAN? on a PC? on paper with pencil?

8. Know how long the purchase can be delayed and what the risks are. If cash is tight, a delay may buy more than just time. But know when the drop-dead date is, after which we've missed the boat.

9. Ask what would happen if we didn't make the purchase at all. What would we lose? What's the worst that could happen?

10. Look at buying less items. Buying in quantity may be cheaper per unit, but it's still more money.

11. Don't buy for next year. There will be better and more alternatives by then.

12. Don't buy now because it's "such a good deal". You may save 40%, but you've still spent the 60%.

13. Know the priority of the purchase relative to all other items or projects the department is doing. Are the top priorities being done first?

14. Finally, make sure the purchase gets you closer to the company's goals. If it doesn't contribute to the mission, can it be that important?

HPMD