Are you finding it more difficult lately to construct a good business case? Infrastructure managers are probably screaming, "Lately? For years!" But now it's becoming more difficult for development teams, too.

Part of this was caused by the "big package" revolution of the last decade. Getting every department that uses an ERP package, for instance, to contribute its portion of the upgrade costs is a hassle for most IT managers, especially when they hear, "We didn't ask for this, so we're not paying for it."

Now, other factors are starting to appear. Service oriented architectures are a brilliant idea for advancing enterprise flexibility and speed to market. But they are a nightmare when it comes to tracing a stream of benefits in a department and tying them to the costs of common code. It's time to rethink the business case, because the application as we've known it is dying and being replaced by a mix of infrastructure and middle layers with a little application code on top. Rather than a monolith that can be justified -- and upgraded -- as a unit, we have many moving parts, all on their own upgrade cycles.

Better CIOs have seized upon portfolio management to help them deal with reinvestment questions. The best among them have moved even beyond that to looking at the portfolio and the business value it supports as a whole and then making investment decisions to manage the top line (by increasing value generated) and bottom line (by decreasing the cost to deliver that value).

Rather than have each project justify itself, savvy CIOs ensure that the complete portfolio -- investment and reinvestment -- maintains a desired rate of return.

Coupled closely with this notion is the idea that qualitative business value -- the gains you expect to be there but can't cleanly quantify -- counts. Customer gains yield a revenue impact, hence a top-line gain in value for the portfolio, regardless of the cause. Likewise, it's the overall profile of the infrastructure supporting the business that matters, not whether one part of it is experiencing a temporary cost bubble due to upgrades.

If this sounds to you as though the CIO is running IT just as the business is run, you're right. Organizations must invest in themselves just to hold their market positions. They must develop new products and take the risk that the investment won't pay off. They must periodically replace worn plants and equipment, refurbish customer points of presence and match competitive customer-service initiatives. This holistic view is precisely what modern IT systems demand.

Some of the problems the portfolio approach solves include:

• How do we justify changes we're making to support our enterprise architecture?
• Where's the business value in reducing complexity, especially when it'll take years?
• How do we replace the old version of that core package when the clients don't want to pay for it?
• How do we justify infrastructure upgrades when there aren't enough new applications coming onstream to pay for them?

It's also a good way to demonstrate you're ready to run a business, not just IT.

Bruce Stewart is a former CEO and onetime senior VP and director of executive services at Meta Group.