Penny Wise Pound Foolish
For those of you who aren’t familiar with this expression, it means to be cautious (wise) with small amounts of money but wasteful (foolish) with larger amounts of money. Does it happen often? Absolutely! I have no doubt it happens in every field and speciality, but it still amazes me just how often it happens in the IT industry.
The catalyst to this post was the recent Sophisticated Cooling Apparatus post on The Daily WTF. The picture alone says a thousand words! As you see from the picture, you have a hardware setup worth thousands, tens of thousands. Not only that, but those machines had to be configured and setup which costs time and money. And I suspect they’re also running a lot of expensive custom software.
What’s truly amazing is that the biggest failure point is a $5 used fan that’s attached with a note to not remove or unplug it because the whole system will collapse. Huh?!? A massive system with large resources (not just hardware and software, but also people) is at the mercy of a basic cheap fan! It makes no sense at all.
Which lead me to search for other examples, and it didn’t take me long at all to find many other examples. It just so happens that The Daily WTF had another article rightly entitled Penny-Wise, Pound-Foolish. And that story was even more appropriate!
In it, a bank hires a $300/hour contractor to setup monitoring software to analyze traffic on $5,000,000 worth of servers. Obviously you’d think they want to get their money’s worth from the contractor, especially at that hourly rate. You’d definitely want to give him a great computer system to work with. You’d want to completely pave the way so that he can work as efficiently as possible. Or so you would think…
Unfortunately that’s not what happened. The person got assigned a completely under-performing system that could barely run Microsoft Office! Never mind actually writing any code. So here you have someone who’s at the mercy of an obsolete computer that can barely even function. What a waste of time! Considering that a newer machine could be had for under $2000, that’s less than the consultant’s daily cost. And remember, in this scenario most of his time is sitting there waiting for the computer to just respond to a command (it had 256MB of ram – not even enough to run Windows XP).
The consultant of course complained, asking for a more powerful box. He was of course denied. He got the all too common response of “let’s make due with what we have for now”. And he’s not the only one to experience this. I’ve known many developers who’ve brought in their own hardware (including myself). This is hardware they paid for themselves! It happens too often.
And by the way the above story is even worse than my summary, but I think you get the point.
So why does it happen? And especially why does it happen so often? Because in larger companies and governments it’s all about budgets. More specifically whose budgets. For example the cost of the upgraded computer for our lowly consultant would probably be coming from another budget, one where they didn’t want to spend their money on something that didn’t give them a direct ROI. There’s no bigger picture here. It’s not like at a small startup where every penny is highly valued. It doesn’t make sense, but unfortunately that’s how the game is played.
Another possibility is the process to acquire the upgraded computer. Often in big bureaucracies it’s easier to acquire new people, even very highly paid consultants, than it is to acquire simple and cheap hardware. I’ve personally seen it many times, and I’ve heard about it even more. I remember at one time spending almost the same amount of money debating the value of getting more advanced hardware as the hardware itself cost. And it was no ones fault, that’s just how it works.
And that’s why you end up with a contractor basically being paid to sit for $300/hour. The larger bureaucracies can absorb this cost because of their size.
However all is not lost. I have a simple solution to propose. Let’s assume that not everyone is an idiot required to fill in ID-10-T forms for every little request. Let’s assume people for the most part are good and want to do a good job. Let’s TRUST people.
Instead of requiring a large process to get a computer upgrade, purchase a smaller software application, etc., let’s assume they know what they’re doing and let them do it. Give them a discretionary budget to spend on things that will make their jobs more efficient. Let them maximize their performance.
Sure some people will abuse the system, but that’s nothing compared to the state we’re in now. Using the example of the $300/hour consultant, it would take a lot, and I mean a LOT of abuse to outweigh the benefits this type of trust system would give you. Even if the consultant decided to purchase a $10k computer, the company would still be ahead!
But it gets better. Remember that for each request sent out, it has to get the approval of several people. Even that $50 stick of ram needs to be approved by several people. And a process had to be put into place. In other words, getting the approval to purchase $50 of ram probably costs a LOT MORE than the $50 stick of ram. I’d bet it easily costs over $200.
Which means that if you compare the trust system to the current system, you could technically have 4 people completely abusing the system for every real request and still come out ahead! And I don’t believe for even an instance that 400% of the requests are from people trying to abuse the system.
But wait, it doesn’t end there, you also get another great benefit. If you trust people they will be more productive. Not only are you saving money, you’ll get better results. People who are trusted are more motivated. When you have a good team that gels well together, they can do great things. If you embody distrust and bureaucracy, well things come to a crawl and any and all motivation slowly dwindles to nothingness.
Trust your people. The results might surprise you.
