You've got to be very careful not to have a CFO bias around expense control, particularly in this space, where most CFOs don't really understand the drivers or the cost structures. The other side of that is that you need CIOs who are business-savvy. If they have trouble articulating the business rationale and value of IT, they will not do well as a C-level equal of a CFO. They won't win any arguments.

At a recent seminar that CFO hosted, some finance executives were complaining that IT folks tend to offer "faith-based solutions" — grandiose projects designed to cure all kinds of problems but lacking good metrics to buttress the ROI proposition. Is that your experience as well?
I think that happens very often, but I will offer a rationale that perhaps most people forget. The accounting profession is hundreds of years old. It has a lot of metrics that are very mature and well understood. There's been a lot of time to work out the kinks. If you want to figure out how a certain type of expense behaves, you just have go to the GAAP accounting.

IT as a profession is barely 30 years old. Think about the growing pains: there were no standards for how to approach and deliver IT services until ITIL [the Information Technology Infrastructure Library] came along [in 1989]. So you had every variation on a theme possible.

IT is still very immature and struggling to figure out basic norms and how to measure things with consistency. The metrics are all over the place, and some things that you'd like to measure, such as CPU consumption on a server, you can't.

On the other hand, finance may appear to IT as being inflexible — they want to know exactly how much things are going to cost and don't want to take any risks, which may inhibit IT from achieving optimal performance. Is that a fair viewpoint?
Finance's job is to be skeptical. Business leaders forever have had to convince finance people to invest company assets in their schemes. But you want the business leaders to take some risks to grow the business or improve profitability. And IT should be able to do the same.

When IT people have trouble explaining something that they want to do, they sometimes derail off into technobabble. And sometimes we let them get away with it, because it sounds mystical and magical — we're not sure what they're talking about, but they're going to make gold out of lead, and we're all for it.

You mentioned earlier that understanding data storage was important. Why?
When you buy a new computer today, it has a 500 gigabyte hard drive. There's an external hard drive that has a terabyte of storage, which is 1,000 gigabytes. Well, it's not uncommon for a Fortune 500 company, particularly a financial-services firm, to have a petabyte or more of storage. That's 1,000 terabytes, and it's an enormous amount of data; you could digitize the entire Library of Congress on 23 terabytes.

My question is, what the heck are you going to do with all that data? Because it's beyond the comprehensible analysis of any human being, and probably beyond that of any large-size computer.

Now, look at what's happened with storage costs. In 1975, one gigabyte of storage cost hundreds of thousands of dollars. What drove the Year 2000 problem was a decision to have two digits for the year, because at the time it would have cost twice as much to have four digits. That problem ended up being fixed at a cost of tens of billions of dollars.

Well, the opposite is happening now. Storage is so cheap that finance isn't paying attention to how much is being bought. Storage costs are falling 70% a year, while consumption is going up 100%. That's a very interesting dynamic, and there are economic repercussions. At some point, the cost of doing something with all that data — cleaning it, fixing it, getting rid of it — is going to be enormous.

This is where finance and IT aren't communicating effectively about the long-term implications of investments.

Do you personally get involved in evaluating technologies and vendors and making decisions on what to purchase?
Yes, very much so. When I want to understand what's a good technology to introduce into my environment — how well it will integrate, what systems it will impact, how long it will take [to implement] — I talk to IT guys. When I want to figure out whether the vendor has given me a good deal and whether the industry is heading in this direction, I have to have a financial hat on. So tech, finance, and end-users all have to be in the room for vendor evaluation.

What else about IT finance is important?
A big item is chargeback models: how we assign the costs of IT investments to the business units in a method that's viewed as fair — I don't know that anything is ever actually fair — and appropriate and accurate.

I've never worked at a company where anybody was happy about how the chargebacks worked.
Nobody's ever happy. Let's be clear: if your goal for a chargeback model is to make people happy, you should be a politician, not a finance guy. The goal is to be fair and equitable, and to be able to explain to people why they're getting those charges. They don't have to like it. Do you like your tax bill? Whether you like it or not is not the point of taxes.

• « previous
• 1
• 2
• 3
• next »