A Closer Look At GM's Tech Spending Cuts

By Paul A. Strassmann  |  Posted 2006-02-07

General Motors CIO Ralph Szygenda reportedly said in 2004 that the auto giant achieved at least a 25% cut in I.T. spending over an eight-year period. His annual budget declined from more than $4 billion in 1997 to about $3 billion in 2004.

From 1997 to 2004, while sales were growing 13%, GM reduced its I.T.-spending-to-sales ratio—a frequently cited efficiency benchmark. The ratio went from 2.4% in 1997 to about 1.6% in 2004, a 33% improvement.

By the numbers, these look like superb achievements. But understanding CIO claims about differences in technology spending is very important, regardless of whether expenditures are up or down. Nowadays, CIOs are frequently asked to explain information-technology budget changes as business conditions change. And this can be done only by examining indicators that reveal how a firm's economic environment has been altered—and those indicators go beyond I.T. spending.

And GM is not the same company it was in 1997. It is much smaller and highly unprofitable.

A rise in outsourcing of work—and not just technology operations—has been a significant trend at GM over the past 20 years. In fact, GM was getting set to award some 40 new computer services contracts as of late January.

The estimated worth of GM's purchases from suppliers for automobile parts and sub-assemblies increased from $114.4 billion per year to $143.1 billion per year. In all, GM has shifted $28.7 billion of costs from GM employees to others outside the company.

This chart below shows what's happened to employment at GM and its increased reliance on outsourcing—displayed as the "outsourcing ratio" and defined as purchases from outsiders—read "outsourced"—divided by sales (to calculate this ratio, see "Outsourcing: What Ratio Is Right?" March 2005, p. 80).

From 1997 to 2004—which coincides with huge reductions in I.T.—GM's total employment shrank from 608,000 to 324,000, or 46.7%. Simultaneously, the outsourcing ratio for the entire GM enterprise increased from 67.7% to 75.0%.

In an interview with Accenture's Outlook magazine, Szygenda noted that business measurements, not I.T. indicators, should be used to judge whether GM has the best business processes. So, I examined business measurements—inventory-to-sales, transaction and R&D costs—to see if I.T. has made GM business processes more effective.

Comparing inventory to sales offers a tell-tale indicator of value-chain management—where I.T. has made the greatest contributions. In the case of GM, there has been a 92.7% increase in inventory corresponding to only a 5.7% increase in sales over the last five years. Comparisons with Toyota and Honda show that their inventories grew only as fast as their sales.

Consider what's happened to business transaction costs for sales, administrative and general overhead. Here, I.T. efficiencies ought to show up because outsourcing should be shifting transaction costs from GM to the suppliers' overhead. The bad news: GM's transaction costs grew by an explosive 73.8% from 1997 to 2004, instead of declining as GM employment was reduced by 46.7%.

GM's employment reduction was accompanied by a cut in R&D spending from $8.2 billion a year in 1997 to $6.5 billion in 2004, or 21%. That relieved Szygenda of supporting a costly constituency. At the same time, expenses for scientific and engineering computing plummeted.

So, the claim that GM reduced I.T. spending in eight years by at least 25% is certainly commendable. Other factors notwithstanding, it's my view that the decrease apparently still does not meet GM's main business need, which is a reduction in transaction costs—where the productivity of I.T. could be demonstrated whenever a firm increases outsourcing.

Paul A. Strassmann (paul@strassmann.com)will continue to track the automotive industry to see how I.T. efficiencies affect profitability.