The Most Expensive Way to Save Money

Nobody decides to neglect their infrastructure.

That is worth saying plainly, because the way this usually gets talked about is unkind and not especially accurate. The picture people paint is of a careless organization ignoring obvious warnings. What actually happens is a series of small, reasonable decisions made by people trying to be responsible with money.

The switch is old but it is working fine. The server is out of warranty but it has never given us trouble. The firmware update can wait until things slow down. None of those are foolish thoughts. Every one of them is a person choosing not to spend money they do not have to spend today. In most parts of a business that instinct is exactly right, and it is why they were hired.

Infrastructure is the place where that instinct quietly stops working.

The problem with "it's not broken"

Most things in a business tell you when they are struggling. Revenue dips. Turnover rises. A customer complains. The signal arrives before the consequence.

Infrastructure does not do that. A twelve year old switch running a production floor gives you the same green lights on its last day that it gave you on its first. There is no gradual decline you can watch and plan around. It works, and then it does not, and the gap between those two states is about four seconds.

So when someone says "it's not broken," they are almost always telling the truth. That is the difficult part. The statement is accurate right up until the moment it becomes the most expensive sentence anyone in the building has ever said.

What you are actually trading

The decision to defer is not really a decision to save money. It is a decision to trade a small scheduled cost for a large unscheduled one.

Unscheduled is the word doing the work there. Planned maintenance happens on a Saturday morning, with the replacement part already on the shelf, a rollback plan written down, and someone rested and unhurried who has done it before. It costs what it costs and then it is finished.

Unplanned recovery happens on a Tuesday at 4pm during month end close, or the morning of a shipping deadline. Nobody has the part. The documentation is older than the person reading it. Two other systems turn out to have depended on the failed one in ways nobody wrote down. A two hour planned window becomes eleven hours, and those are not just IT hours. They are a floor of people who cannot work and orders that do not ship.

That is where the real number lives. Not in the hardware. In the hours, and in the fact that you did not get to choose when they happened.

Nobody has unlimited budget

Here is where a lot of IT advice goes wrong. It assumes that if you just explain the risk clearly enough, the money will appear. It does not, and pretending otherwise is not useful to anyone actually running a business.

You have a number. It is smaller than the list of things that could theoretically be improved. That is not a failure of planning, it is just what operating a company looks like.

So the honest question is not "how do we update everything." It is much narrower and much more useful: if you can only spend X this year, where does it go?

Answering that well requires knowing which aging system is genuinely fine for another three years and which one is a single point of failure that takes the whole operation down with it. Those two things often look identical on an inventory sheet. Telling them apart is the actual expertise, and it is most of what we do. Some equipment we will tell you to leave alone. Some we will tell you to move on this quarter, and we will explain why in terms of what happens to your business if it fails rather than in terms of version numbers.

That is the conversation we would rather have with you than the one at 4pm on a Tuesday.

If you have a number in mind for this year and you are not sure where it should go, let us look at what you are running and give you a straight answer. If you are in decent shape, we will tell you that too.