One of the most difficult things I have experienced in working with computer systems is to know when to have the honest discussion, and decide to cut your losses, and move in a different direction. Many times we begin a project with some degree of risk, simply because there are things you can’t know ahead of time.
Sometimes you make fundamental decisions that are wrong. When those problems don’t arise quickly, the cost of changing direction changes at an escalating pace to the point where it is difficult to move in a new or different direction.
Here’s an example. I worked for a large farm cooperative a couple decades ago. We were using very expensive hardware and software running on a mainframe. We had an assessment of our infrastructure. The assessment recommended we move to a midrange computer and software. It would only take a year to recover the cost of the conversion…so financially, that was a simple decision.
It took us a year to acquire the new hardware and software, and complete the implementation. We were at risk of going out of business because of the poor performance of the midrange computer system in our configuration and load. It just couldn’t keep up like our mainframe. Sure, we were saving a lot of money; but we were losing customers fast because of poor customer service due to the slow performance.
There was no way we could have predicted the performance factor based on our load. We frankly took a risk, based on the recommendation from the assessment, that all would be well. To our great relief a faster version of the midrange computer was released just days after our implementation. An upgrade to the new server resulted in a very good solution.
Computer systems can be complicated. Sometimes it takes years to find that your foundational decisions were wrong. I have a few observations:
- Don’t put off checking your assumptions. The faster you make adjustments the better your outcome will be.
- Don’t let pride be a factor in your decisions to move on or move forward.
- If possible, build tools to assess areas where your knowledge is inadequate in order to reduce the cost of inaccurate assumptions.
Cheers,
Ben