By Mark Verbeck
Managing Director, Senior Analyst

New numbers from the Bureau of Economic Analysis suggest that the software industry might be entering its first recession

FIGURE 1
Aggregate Change in Hardware & Software Spending
See Image

(defined by two consecutive quarters of negative sales growth) since the current category was constituted in 1989. Business software spending has increased in each of the 49 quarters tracked by the survey, until the 0.4 percent drop in the first quarter of 2001.

This is in stark contrast with the business hardware category, which contracted in 15 of those same quarters -- almost one of every four quarters. What is behind this change in the software industry, and what can it tell us about what we can expect in the future?

There are two main reasons for software spending's resiliency to date. The first is the perpetual project backlog. Many corporate IT departments have years of backlog for new systems. In this environment, it is difficult to cut out projects. Second, IT professionals are fairly immune to cutbacks because they know so much about the existing systems, and they are difficult to replace. Because these resources are available -- and drawing salaries -- it makes sense to ensure they are productive in good times and in bad.

Although it is true that the software industry lacked the downside volatility of its hardware counterpart, it also lacked the upside. The best quarter for software spending in the last 12 years was a 7.4 percent increase, whereas hardware spending cleared this hurdle four times, with a top growth rate of 11.4 percent in the second quarter of 1995. Clearly, software spending is not limited by budgets. Our analysis and research suggest two other factors. The first is IT resources: even with consultants and vendor support, any software implementation requires a significant commitment of internal resources -- a demand that commonly triggers internal battles for the precious technology resources needed to make a project happen. The overall shortage of skilled technology workers only exacerbates this problem.

The second factor limiting software spending is how well an organization is able to adopt change. Internal limitations on changes in process, functions, or compatibility with existing systems often limit large-scale software projects. Indeed, it is the software's inherent ability to make business processes more efficient that produces significant change -- otherwise companies wouldn't purchase the software. Many of the notable software project disasters have involved the simultaneous adoption of many new systems.

Three long-term trends have altered the landscape to create an environment where software spending could actually be reduced. The first is the tremendous growth of the consulting industry. In 1989, the IT consulting market was nascent. By 1999, it had grown to a $34.6 billion business.

FIGURE 2
Hardware & Software Spending as a Percentage of Private Fixed Investment
See Image

The second trend is the growing portion of companies' capital budgets that software consumes. In 1989, software represented 10 percent of total equipment spending -- equal in size to hardware spending. Today, software represents nearly 24 percent of the total -- double that of hardware spending. The final trend is the backlog of maintenance and upgrade projects that accumulated during the e-business rush. This development queue provided a ready technology to-do list for IT professionals without necessitating the purchase of any new software. As a result, beneficiaries of the e-business rush like business-to-business (B2B) companies Ariba¹ and Commerce One, as well as eCRM (e-customer relationship management) companies Kana Communications and Blue Martini Software, have been particularly hard hit in the pullback.

Hardware & Software Spending as a Percentage of Private Fixed Investment

We expect IT departments to work through their maintenance backlogs quickly, and they will likely begin looking for the next projects to implement. This doesn't mean that spending growth will return immediately, but demand for supply-chain improvement, collaboration, improved customer relations, and new Internet-enabled processes will eventually fuel the industry's rebound.

For this reason, we expect companies like i2 Technologies¹, Manugistics, MatrixOne, Agile Software, Siebel Systems, Pivotal, and some emerging private firms to lead the way out. We do not expect, however, that software spending will return to the steady uptrend of the past. As enterprises devote larger and larger shares of their capital budgets to increasingly sophisticated software, we expect software spending to become less predictable and more volatile.