Common Systems = Common Sense
It distinguishes great companies from the mediocre. It’s strikingly simple in concept. Still, companies that haven’t been doing it regularly, however, find it extraordinarily difficult to begin and maintain. The practice is “commonization.” Commonization greatly reduces unnecessary proliferation and replication of work. While requiring more initial effort, commonization accelerates and streamlines latter routine work. Common information systems, in particular, greatly help a company act as one company.
“Commonization” refers to seeking a common solution for multiple groups that share the same requirements in a particular area. In the auto industry, top executives have long stressed common vehicle parts, common platforms and common vehicle architectures. Unnecessarily making 70 different seat belt designs when three will do is an example of an enormous waste of resources. Similarly, having dozens of different business processes where one could suffice also greatly adds inefficiencies. For instance, Toyota’s insistence on a single Toyota Production System worldwide has given it an enormous competitive advantage. Commonization makes sense across the board. For instance, General Motors in 1992 had 27 separate purchasing centers. Thanks to a strong commonization push, it reduced them to just one purchasing center by 1997. Such reductions where championed by GM’s then CEO, Jack Smith, whose mantra was “Run common, run lean.”
Information systems (I.S.) plays a major role in facilitating commonization. The group can establish common applications, databases, hardware, and portals. For instance, Ford Motor Company had significant problems with rogue spending in the 1990’s. Its I.S. unit responded by developing a single, worldwide, supplier database. The automaker then mandated that only suppliers in this database would be paid. Doing so gave Ford management much better insight into its spending and how to take out supplier costs.
Johnson Controls (JCI) has standardized on four major application areas. These cover the bulk of its information needs. The principal application areas (and vendors) are:
- Manufacturing (MFG/PRO from QAD)
- Product development (MatrixOne)
- Finance (Hyperion)
- Human resources (PeopleSoft)
At the same time, JCI implemented a common data repository and search engine for its single, company-wide portal. These efforts pay off in the small tasks that monopolize the everyday life for a worker. After this effort, for instance, JCI’s product developers spend 70% less time looking for a CAD model than they previously did, said Sue Kampe, VP and chief information officer (CIO) at JCI. Prior to communization, a particular CAD model could reside in any of dozens of locations, accessible only by particular programs. This was true even for a component created just last year. A product engineer’s inability to find a particular part could lead to creating a totally new component, which is not only extraordinarily costly, but time-consuming. In today’s wobbly economy, manufacturers simply cannot tolerate such duplication.
If the benefits from commonization are obvious, why doesn’t everyone do it? The answers are politics, a narrow-minded optimization mentality and downright weak corporate management. Companies without a long tradition of commonization are Balkanized. The various groups that could share a common approach already each have their own pet systems. A worker’s career and, indeed, identity could be closely tied to one of these unnecessarily redundant systems. Their deep experience and knowledge of those systems make these workers seemingly very valuable to their firms—as long as those systems remain operational. Not surprisingly, such workers often bitterly resist the move to a common system that simultaneously eliminates the current system that they now “own.”
Other factors impede commonization. Multiple groups could potentially share a common system but not have identical needs. For instance, one enterprise resource planning (ERP) system could be very strong for discrete manufacturing operations. Another ERP system could be preferable for batch manufacturing. If a company has both discrete and batch operations, one company-wide ERP system would be sub-optimal at the manufacturing level. The benefits accrued at the corporate level by having just one ERP system, however, may outweigh minor inconveniences at the manufacturing level. I.S. management must play a major proactive role for commonization to succeed. If management here is weak or the company is excessively decentralized, I.S. commonization will not happen.