Remember the glory days of IT spending, the late 1990s? Obviously, they’re gone, killed off by a combination of a post-Year 2000 slump, the bursting of the Nasdaq bubble and Sept. 11. Recently, there’s been some good news – sort of. According to the latest data, the loss of IT jobs has slowed to the lowest level since 2000. Losing less isn’t the same as gaining more, but it’s at least movement in the right direction, and it makes one wonder what’s behind it.
It’s likely that a part of the story is orderly economic growth. If IT was overstaffed in the past and the market then corrected, it’s logical that as the economy grows, IT needs will grow with it, and a correction won’t be needed. Business transactions tend to grow at about the level of the gross domestic product (GDP), which has increased more than a third since 2000, and IT spending tends to roughly follow GDP.
Roughly, but not exactly. In the post-bubble collapse, the ratio between GDP and IT spending hit a post-World War II low. This ratio shows clear cyclical behavior, and previous lows like the one we saw in 2002 were followed by a period of significant growth, lasting as long as six or seven years. Our 2002 low came just a few years after a peak in strategic IT spending that nearly matched the post-WWII high reached in the late 1960s with the introduction of IBM’s mainframe System/360.
Suppose we really do have cycles in IT spending. What does the current job situation tell us? In past upward cycles there was a distinct, three stage growth process. First, the improvement in conditions from the bottom point caused companies to pick up spending plans they had deferred. This created a small bubble of tactical reinvestment in past IT paradigms that lasted two to four years, followed by an improvement in the job market. Why? Because the next stage in the cycle is a more strategic investment in some new IT paradigm, and that leads to the third stage, companies’ ramping up workforces to make the IT changes this stage demands.
From the evidence of the previous stages of strategic-spending, it seems likely that new IT paradigms deal with a new way of relating computing to workers, and what fits the bill in today’s market is service-oriented architecture (SOA). By creating more flexible ways of building applications, SOA lets users tune applications to workers almost individually, optimizing productivity. That’s important because productivity enhancement is the main reason to spend on IT in the first place.
If all this is correct, we should see signs of the oncoming strategic stage in other areas, and venture capital research shows that network and software investments are coming back into vogue. We should also see leading-edge successes in SOA, which clearly has been happening. What’s next, then?
First, timing. At about this point in the most recent cycles, there has been a spending downturn for one to three quarters that seems to be linked to the transition between the already mentioned tactical spending on the previous IT paradigm and new, strategic spending. The current IT indicators seem to point to the notion that we’ll have this hiccup for the remainder of 2006, at which time we would expect to see orderly growth to the peak of the cycle, around the end of the decade.
Second, impact. Job loss will continue to slow, and job growth probably will resume late in the summer of next year. Capital spending on IT equipment aimed at SOA will expand through 2007, with software, servers and then networks benefiting. It’s likely that 2007 still will be a tad soft for networks, because CIOs will be focused on the IT side. Incumbent vendors will benefit most in the early part of the cycle (to about 2008), and start-ups will come into their own later on. All this is consistent with what’s happened in the past.
The industry’s long nuclear winter may be finally coming to an end.
