How to satisfy business, IT needs in a customer data conversion project

Accurately moving customer data from a legacy environment to source, on time and within budget is the clear objective of any data conversion project, but how to achieve it is not quite so clear, particularly when you are dealing with customer data.

Aubrey van Aswegen, MD of Knowledge Integration Dynamics (KID), offers some solutions on getting the data out of legacy environments and into useful business analysis data stores.

Customer data conversion projects present complex challenges, as they must satisfy the needs of the business and the IT shop involved in the project.

The goals of both sides to this conversion coin are:

* Delivering predictable time frames;
* Budget projections and resource requirements; and
* Accurate customer data.

In terms of achieving business goals, conversion projects should be incrementally handled, with pauses for analysing conversion rules and results. This ensures technical staff can engage business staff at every critical decision point in the project, ensuring that none is bypassed.

Without this, there are many potential pitfalls that can cause project failure, such as poor source system metadata and lower than anticipated source system quality. Problems such as these can scupper conversion project efforts and must be nipped in the bud.

Two critical steps are having the right consultant and properly investigating the legacy data. With these two components the business and IT goals can be met and the conversion project result in success.

Having a senior consultant working on the project ensures the experience they have from working on similar projects, in scope and size, result in their giving the best advice on procedural steps, taking into account whether or not data from the legacy environment will be used in operational or analytical formats. Getting the steps right, depending on the projected data format, ensures projects deliver against the business goals and within time, budget and scope. It also clearly delineates the technical resources, skills and tools that are required to make it successful.

In trying to achieve these goals, the organisation is faced with a number of issues. Technically, discovering the data is the first step in the process, and it can prove tricky because there tends to be little documentation in legacy systems.

The second technical obstacle to overcome is data hierarchy. Again, due to poor documentation, the data hierarchy tends to be well concealed.

A decade ago, when business intelligence (BI) was the catchphrase of the day and only the "hottest" IT shops were implementing BI solutions, tools and skills were immature. As a result many projects failed to deliver any benefits or returns, but the discipline has evolved and now skills and technologies have matured to the point where extracting data from legacy environments is no longer the technical problem it once was.

Today there are software tools that will discover the data and extract the hierarchies, and even cleanse the data to ensure a more consistent quality that the business can usefully analyse.

However, these tools do not make conversion projects any less complex; they merely assist in the process, and project success still relies heavily on marrying the technical and business needs. This detailed data analysis also assists the project by ensuring, at every step in the process, that business needs and technical needs can in fact converge.

With all the information that is gleaned from this process, the conversion plan can be revised and accurately detailed to ensure success. Following this, data samples can be taken and put through a data conversion pilot project that ascertains whether or not the plan does in fact have a good chance of succeeding. Once again, following this step, it is time to revise the plan before it is executed.

The actual conversion, even considering thorough planning and preparation has been performed as outlined above, is always the acid test. But by working together, business and IT users are almost guaranteeing a successful project.