You would think IBM might learn from their own experience.

Organizational structure is strategic by nature.  How an organization is arranged influences other strategy, such as marketing and product development, and thus a whole host of daily activities and tactical initiatives.  Who your boss is and what her objectives are determine what you will do, such as replacing her coffee with decaf in order to encourage afternoon naps and thus allow you to get some real work done.

IBM almost went bankrupt due to their organizational structure, as their former president and resident cookie monster Lou Gerstner confided in his book Who Says Elephants Can’t Dance? Given their early and dominant lead in mainframe computers, IBM developed an organization structure that made the mainframe the center of their universe.  Everything IBM did evolved to support sales of mainframes.  When the minicomputer revolution ignited and UNIX (the original computer virus) escaped its petri dish, IBM’s intellectual inbreeding disallowed agility.  Their organization structure –software was subservient to hardware — made even suggesting alternatives heresy.  Though Gerstner did many things to disable IBM’s dementia, the single most important one was to decentralize and make all IBM divisions profit centers.

Nothing like sinking or swimming on your own to make you dog paddle furiously.

Recent IBM news should thus be disheartening to investors.  Organizational shuffling has merged software and hardware units together, with software being the overlord (and if a forced marriage was wise, software should be the top given market trends toward commoditization of hardware).  Two services divisions — technology and business — have also become conjoined twins, with tech services being the bigger brother in that particular carnival sideshow. In a statement almost copied from every start-up’s venture capital pitch, IBM’s CEO said “There are logical synergies across our services units, including the increasing value of leveraging our intellectual property in business process management and transformation projects for our clients.”

Stockholder should be scared when a CEO’s press release is littered with cliché buzzwords like ‘synergies’, ‘leveraging’ and ‘transformation.’

Granted, these internal mergers are not completely reflective of the old IBM.  There is no plainly visible central altar on which every blue suit must sacrifice chickens and Lenovo laptops.  Yet CEO Sam Palmisano has started the slow march away from decentralized entrepreneurialism to central planning … and we all know how well that worked for the Soviet Union (“Who?” asks the intern with authentic confusion).  Decentralization worked for IBM as a means to prevent going belly-up, and it worked for Hewlett-Packard before the Fiorina error, so Uncle Sam Palmisano’s decisions appears demented.

How an organization is arranged influences other strategy, such as marketing and product development, and thus a whole host of daily activities and tactical initiatives.  IBM’s recently released zEnterprise system, though not the hallmark of utter innovation, shows that hardware still matters.  Putting IBM’s systems and technology group under bit twiddlers means that hardware innovation will be limited to the needs of IBM’s software (instead of the entire software universe, which despite IBM’s thinking, is larger than Armonk city limits).  Product design is reflective of product marketing, and if hardware product marketing labors for people who peddle DB2, Lotus and Cognos, then there may well be very little IBM hardware innovation.

Which is when HP, Oracle and Dell will pounce.

The marketing lesson herein is really a management lesson.  People don’t do what you tell them, they do what you watch.  Every bureaucracy is designed to watch what employees do with some end goal in mind.  IBM’s consolidations causes employees to be watched over an artificial hierarchy of product group priorities instead of markets.  When markets matter less than internal organization, then the end will be near … again.