Perils of diversification

Thursday was the first of my honors student presentations to their industry sponsors. This team — Kevin Dines, David Hsu, Clifford Jung and Vanessa Silveira — gave the concluding presentation of their three-month consulting project to researchers at IBM’s Almaden Research Center.

The student project was to find an eventual market for a research project by the PhD researchers. To do so, their project required interviewing government agencies and associated private firms such as real estate developers and urban planning consultants.

When for their project they contacted the various consultants on behalf of — large and small — the initial reaction was almost always “IBM is a potential competitor.” The scientists were amused at this, but if you put “urban planning consulting” into Google, you get an IBM-sponsored link:

I think IBMers would see the company as a supplier of IT and integration services to these consultants. However, as a multinational that increasingly supplies everything to everyone, it’s understandable that existing firms would worry that sharing information would enable further diversification or forward integration.

Oracle has spent the last 20 years buying some downstream customers and competing with others. Its market footprint in databases — approaching an monopsony — allows it to dictate terms or at least force these customer-partners to the table, even for big companies like HP, IBM or SAP.

Still, it makes sense that seeing everything as your potential market would cause everyone to see you as a potential competitor. This is not a risk of diversification we normally teach in strategy, although it’s commonly mentioned as a risk for upstream or downstream diversification.

Aside: as with my previous visits to ARC, the facility strikes me as a gorgeous site where the employees are as well treated as anywhere in any company in the world. As I tell my students, it’s always best to work for a high-margin company.

Note: This was ready to post Thursday afternoon, but posting was delayed due to Google’s 20-hour failure in running Blogger.