I recently heard a very telling story about Lou Gerstner, the executive largely credited with turning IBM around in the mid-1990s. When he joined IBM he commissioned a benchmarking study on payroll systems used by large enterprises. When the group presented their findings, they told Lou they had some good news and some bad news. The good news was that the firm had the best payroll system they had ever come across. The bad news was that the firm also had the 4th, 8th, 23rd and 28th payroll systems. The message was clear and Lou took aggressive action.

My experience tells me this overlap and redundency is a frequent and widespread occurance in large firms where there are substantial business units that either had the autonomy to build infrastructure to support their activities or they were acquired with little integration with the “mother ship” of the acquirer. I’ve seen this silo-driven culture saturate many critical enterprise efforts ranging from simple web site management to corporate strategic planning. Why is this so prevalent and what is being done to prevent making the same mistakes again and again?

Seems to me this is clearly a question of governance (or lack of governance) and the expense of poor/weak governance grows steeper with each new application and business unit that gets created rather than brought into the fold. There don’t seem to be many hard-liners who are willing to say “this is when the madness stops” followed by swift and meaningful action that includes going after legacy efforts.

I guess I just wish there were more big game hunters in corporate America. There’s plenty of productivity and cost savings to be had if you can fell a few elephants. I’ve often seen many who are willing to pounce on fledgling efforts to make them join the flock but few want to try going after the bigger legacy items. The hard and unglamorous decisions to effectively govern internal processes and systems (that aren’t necessarily broken) seem to be inevitably delayed until the firm is in dire trouble. Why wait until its nearly too late?