Nerds of Steel

A senior steel industry manager who had just been offered a promotion to head office was reflecting on his choice. Looking for guidance, he asked a colleague for advice. ‘Just add value,’ was the terse reply. His implication of course was that corporate ‘overhead’ usually doesn’t. His deeper implication was that real value is only added in the operating units. In the new world of the steel industry, that bias could be costing steel companies growth and performance.

Steel industry employees have long been suspicious of corporate headquarters - with good reason.

To simplify, the steel industry has had three basic organizational types.

1. Large integrated companies (Big Steel) that roamed every major developed nation and some of the developing nations too.
2. Mini-mills, fueled by European technology and North American entrepreneurial spirit, that scavenged the periphery of the industry until their superior productivity and lower costs gained share in the core markets of their much larger rivals.
3. The ‘re-constituted mills’, those failing companies that were salvaged by new owners, a new labor contract, skillful operational management and a few careful capital expenditures.

Of these, only the first group had any corporate headquarters scale. They competed in multiple products. They researched and developed process and product. They served the largest and most demanding customers in the market and in Europe and Asia at least, they also took care of distributing much of their own production. They were large complex organizations. For much of their existence their industrial dominance in benign economic conditions made that complexity tolerable. But it ultimately made them easy victims to the blinding simplicity of the mini-mills and to the burden of running high fixed costs in an industry facing slow growth and constantly falling prices. For the latter part of their existence, big steel organizations looked bad at almost everything.

But steel companies are now growing again, both organically and via acquisition, which is forcing executives to contemplate anew the shape and structure of their organizations. They know the old big won’t work and the highly effective operating model of mini-mills won’t ’scale’. A decentralized structure doesn’t help manage the new competitive challenges which require a clear view from the corporate office and plenty of managerial capacity to assess acquisitions and new capital projects, to design new ways of managing, among other things, global purchasing, post-acquisition integration, coordination of commercial and technical service across divisions and so on. And the skill with which the organization achieves these things will be a greater determinant of competitive success than differences between competitor business units.

But coordination costs money. Organization designs, as in architecture, usually try to keep to a minimum the parts of the structure that are there to hold the other parts up. The architectural exceptions arise in bigger buildings. In those cases, it’s necessary to commit a large portion of the floor space to the structural and service components. Typically that means about 35% of the floor space in a tall (> 20 stories) building compared to about 20% for a smaller building.

In just the same way, larger organizations can expect to spend a greater proportion of their management and systems resources to maintaining the operation of the organization itself. A scratch of some of the data seems to support such a view. A comparison of the % of corporate sales devoted to sales, general and administrative costs (SG&A) with tons shipped for a collection of 15 steelmakers in 2006 (see chart) suggests that the bigger the business, the bigger the percentage of sales spent on SG&A.

More intriguing still is that when comparing only those companies shipping less than 20MT there is an even stronger correlation.

Why do the largest companies obscure this relationship? Ignoring the possible distortions of different accounting regimes or corporate SG&A allocations, two things seem to cloud the view. First ArcelorMittal has an SG&A percentage of about 9%, less than some of the much smaller companies. In other words, there appears to be an economy of scale to SG&A achieved somewhere above 20MT. SG&A doesn’t continue to grow as a percentage of sales forever.

Second, the relatively small amount (3-5%) of overhead at the other larger companies in the analysis (Nucor, US Steel of the USA and POSCO of South Korea) has a dampening effect. The possible explanations for these companies’ low SG&A rates are:

a) They are more efficient with their SG&A and have superior organizational models, and/or

b) They operate predominantly in a single national economy and enjoy administrative efficiencies as a result

But it is more tempting to view the situation as an under-investment in SG&A. It’s just possible that these producers are not being frugal, but penny pinching. It may be time, given the extent of the corporate strategic agenda, to stop looking at all SG&A as a cost to be minimized and consider it more of an asset to be put to work. It might be less a case of the bigger you are, the more you spend on SG&A; and more a case of the more you spend on SG&A, the bigger you can get.

This is not a license to be profligate with organizational overhead. But it means that not all overhead should be considered waste and that large companies that skimp on productive overhead could find it difficult to grow as fast or even to purchase supplies, recruit and manage people and projects or operate as efficiently as ArcelorMittal. The best corporate organization will win. And if that is the case, there’s no longer any need for talented managers to be concerned about the invitation to join the ranks at corporate headquarters. They’ll find plenty of ways to add value there.