Evolving IT Systems - Key Concepts

Posted by Steve Cormier

Find me on:

1/14/14 11:00 AM

Many organizations who want to migrate to a new, comprehensive operational system try a ‘big bang’ approach. This approach to business transformation usually proves unsuccessful. Most ‘big bang’ projects fail for multiple reasons such as:

  • The gap between the evolved complexity of existing systems and the new system (out of the box) is too large.
  • However innately sophisticated it is (e.g. SAP), the new system almost always lacks the matured, business-specific capabilities of existing systems (if put into production, the new system would develop these capabilities over time). 
  • Timing of migration is a very difficult and complex topic of its own. How to create a workable migration plan is usually not well understood, particularly with respect to comprehensive reporting for areas such as sales, marketing and finance.
  • The stress of learning a new large scale system while maintaining the current one overwhelms personnel resources. This can cause errors in current production, the correcting of which pulls resources off the new system development causing further migration delays. 

The solution?

  • Develop a thorough understanding of what the future comprehensive system has to support from a business perspective. This includes such thorny issues as security and multi-role actor participation (e.g. a single individual or org unit that holds multiple organizational positions).

    If the true challenge of meeting these requirements is not realistically understood, the systems implemented will be too simplistic, fall short and fail. Or, they may fall short and a subsequent frenetic attempt to make them work will result in kludges that yield a system little better than the original in terms of complexity and flexibility. 
  • With an understanding of the targeted future, begin to build scaffolding on the current systems that moves them toward fulfilling the desired future state. 

    This will consist of data examination and transformation activities that will also deliver business value by supporting structures such as data marts/warehouses. These structures can more quickly deliver value to the business community, since they are read-only and less volatile. Their delivery will lend credibility to the effort overall. It is important that these initial efforts be narrow in scope, targeting highly valuable deliverables.
  • The creation of a metadata infrastructure that can support a comprehensive understanding of the organization’s data will act as a bridge from the current state to the future model. The organization can move to an integrated, operational platform with its eyes open as to the capabilities that have to be fulfilled and with a realistic migration plan that can be achieved.
  • Once the organization has a better understanding of the structure they want to achieve, and the one they have presently, they can build the new operational infrastructure (which may involve extending a packaged solution) with far more confidence. Move functionality to the new system in pieces carefully bridged to the older systems. The idea is for the new system to slowly consume the older one.

The main message here is; evolution not revolution. 

Topics: Digital Business Evolution