Just 26 publicly traded U.S. Companies posted revenues of more than $50 billion in 2004 (allowing for various fiscal year-ends) to get into the $50 billion club.
Between them they account for a lot of people (several million employees globally), a lot of business ($2.5 trillion in collective revenues, and only one of them failed to report a profit) and, potentially, a lot of spending on IT. Talk to their CIOs, however, and the impression you get is that these companies are not well served by the mainstream IT industry, whether it's for hardware, software or servicesbut especially in software and services.
These firms are all truly marquee names, everyone wants them on a customer qualification list, and they can generally afford to buy the best. So what's causing the disconnect?
Partly it's because they are so large that any meaningful project they might undertake would consume significant talent and resources from even the largest services provider. Partly it's because their needs are so highly specialized that "commodity solutions" seldom have much to offer them. Partly it's because they are so diverse (covering fifteen different industries, with no more than four in any major industry category) that they don't really form a coherent "marketplace" that could attract a common set of solutions. After all, it's a lot easier to build a human resources system for the "core" market of companies with less than 10,000 employees (or the mass market with less than 100) than it is to build one for the tiny handful of companies that have more than 500,000.
And those core market systems just don't scale all that well. What works well at the scale that gets you into the Fortune 1000, or even the Fortune 500, will show signs of serious strain when you get to the Fortune 30.
The kinds of extensive engineering efforts needed to scale up by an order of magnitude are hugely expensiveand simply aren't justified by the returns from the few customers that require such scale. That's why most, if not all, of the IT departments in the very largest businesses are essentially still custom development shops. Even where they base some part of their business automation on commercial off-the-shelf products (as they all do in some areas), they end up customizing these offerings to such an extent that they might as well have developed them on their own.
There is an irony here.
It can be argued that the whole commercial off-the-shelf ERP wave was sparked by SAP, which got its start developing systems for the world's biggest oil companies and process manufacturers. But it wasn't until the advent of R/3 and a retargeting by SAP and others toward the "mainstream" corporate marketplace that ERP really accelerated.
This "small market" problem exists elsewhere. Consider the hospitality and travel industry (with central reservation and capacity scheduling systems), telecommunications (with the OSS/BSS suite), multistate, multiline insurance (with policy management and underwriting systems), and utilities (with billing and customer service).
The players in these industries may not all be as large as the $50 billion club, but there are few enough of them in each industry segment that they must all develop or heavily customize the systems that run their businesses.
Such high levels of customization translate into very high total lifecycle costs: Not only is the initial development effort high, but the results have to be maintained over what is, usually, a long lifecycleand there is no one with whom to share the costs. In a very large-scale, low-margin business (think auto manufacturing, big-box retail or pharmaceutical distribution), this adds a measurable burden to operating costs and corporate performance. It adversely affects the bottom line, yet the companies have no choice but to invest. You can no longer run a business of this size without extensive automation.