I have lost count of the number of cost reduction initiatives that I have seen or been involved in, that have been declared a success, and yet never reduced cost by a single penny.

It's great that you have managed to save x many hours, that you have increased your efficiency or that you have avoided increasing your spend, but please, please, please do not confuse these with reducing costs.

It just like when my wife comes home and tells me she has saved $50 because they shoes she'd bought were 50 percent off.

If the shoes were not included in our original budget, and at the original price, then there is no cost saving, and more than likely our costs have gone up, now down.

You cannot spend your way to cost reductions!

I'll never forget sitting in a board meeting listening to our EVP of IT Infrastructure giving a presentation on how he had saved the company $20m from our IT Hardware budget. It was a superb presentation, lots of charts, some with numbers some with great pictures and he spoke very eloquently.

At then end, he asked if there were any questions. Feeling that I hadn't quite understood, or that I had missed something I gingerly raised my hand.

"Yes Gordon, what's your question," he said smiling.

"So where can I see this money you have saved?" I asked.

The room went silent, chilled even, and I could feel every single eye on me. The presenter had a look that I thought might turn me to stone if he persisted with it.

The deep silence was punctured by our CEO saying, "What an excellent questions, thank you, Gordon, I was just thinking the same thing myself."

As it turns out, practically all of the claimed savings were just cost avoidance measure, and there was little to no real savings from the planned budget at all.

As my reward for challenging the EVP of IT Infrastructure, I was given the task of saving $20m the following year. Seeing as how I was such an expert, and even better, it was allocated as the only target needed to achieve my bonus.

I must admit there seemed to be a smile of revenge as I was given the task or poisoned chalice as my boss called it.

When it comes to cost reduction in IT, there are only four real contributors people, services, hardware, and software (including licenses).

So the only way to reduce costs was to spend less than planned on these any or all of these four items. Anything else would be either an efficiency gain, which is nice but don't necessarily reduce cost or are cost avoidance which again doesn't reduce the planned costs which was the goal.

The company had been through a number of aggressive acquisitions, and with that, we had incorporated all of the acquired companies hardware and software resulting in over 4000 servers in our onsite data center this looked like a great place to start to look for cost saving opportunities.

Given that our business was IT intensive, the main focus of the infrastructure teams was ensuring that the systems remained stable, and any server that didn't cause any problems was left unchecked due to time pressures.

A quick study of the utilization of the servers showed that a little over 700 were utilized less than 1 percent. When I asked why this was, I was told that it was because they were back-ups and it was normal that they would be underused, especially if there was no problem with the main servers.

The number seemed a little high to me, so I had the team do a little digging. What we found was that 450 of these servers were not supporting any systems at all and they were in fact just idle, occupying data center floor space and consuming power.

Here was a chance to save some real costs, especially as most of the servers were being leased, and could be sent back to their supplier.

Of the 450 unused servers, 400 were costing us money we didn't need to spend, and just by returning those we were able to cut our budget by $12m.

The reduction in servers also meant that we could reduce our software licenses saving another $4m.

There was an unbudgeted proposal to extend the data center as we were short on capacity, but by removing 450 servers, we this was mitigated, and we didn't need to spend the $8m estimated. We didn't add this to the saving though as it was a cost avoidance, not a true cost saving.

As the team started to understand that the only way to save money was to reduce our planned spend, additional initiatives came up where we could co-locate applications on servers which had spare capacity, and then allow us to return what we now turned into unused servers.

With the reduction in equipment, it also meant that we didn't need the same headcount as we had before, and we were able to release some of our contractors which again allowed us to cut our planned budget.

By focusing on real cost drivers people, hardware, services and software and looking at reducing our real consumption of those, we were able to successfully cut our costs by $20m.

Cost avoidance and efficiency gains are great, but real cost savings come from focusing on spending less on those things that you planned to consume.