Cynthia Rettig has hit the nail on the head in her article “The Trouble with Enterprise Software” (http://sloanreview.mit.edu/smr/issue/2007/fall/01/)! This is based on my personal experience with multiple SAP implementations. Whenever an organisation commits to the levels of investment required to implement corporate politics will come into play. Executive stakeholders in the project will spin the ERP implementation into a success even when it the monolithic system stifles an organisation's ability to change. And that's the crux of the problem. It's not that ERP systems don't create efficiencies, it's the efficiencies are based on a point in time. ERP systems often cement an organisation into a particular way of doing business. As the business environment changes, the cost of changing business processes within the ERP systems outweighs the potential benefits of market opportunities on a case by case basis. This is particularly true for speculative new products and new market opportunities where the revenue growth is uncertain. The high cost of changing these systems often makes the business case for new opportunities inviable. As a result, line of business managers may be forced to use smaller systems to launch new opportunities thus adding to the complexity of enterprise IT environments.
Enterprise SOA will not be realised because the solutions developed by most vendors are too heavyweight and cumbersome. The best examples of SOA in action are web mashups. Organisations striving to achieve agility by implementing heavyweight middle tier stacks are barking up the wrong tree.
Why is it that one doesn’t picture an ocean liner, or a 747, or anything that is large and complex when we envisage “agility”. We always picture lean, lightweight, hyper… Companies should look to small, lightweight, simple services to deliver agile business processes. Leave the legacy processes to the ERP systems but build future nimbleness on lightweight web centric architectures. in action are web mashups.