Systems conversions are by their nature complex initiatives that often involve more than just the information technology assets of an organization. Conversion initiatives can drive transformational change not only to the systems but also the hardware, business processes, and people who rely on those systems for achieving their business goals.
There are many drivers for initiating a system conversion, including IT goals such as:
- Sunsetting a legacy system that relies on outdated software or hardware in order to move the business to a state-of-the-art platform
- Consolidating data in fewer databases and applications to improve data quality, maintenance, and IT and operating efficiencies
- Porting of data from in-house systems to vendor-maintained applications
Business drivers behind conversions are numerous as well and can include cost reduction targets, the requirement for increased throughput to support business growth, or to carve out an entity’s systems and data in preparation for a spin-off or merger with another company.
As there are many drivers for successful conversions, there are many pitfalls as well. Due to the complicated nature of conversions, it is not uncommon for these initiatives to run significantly over budget and miss important deadlines. Managers need a structured approach to converting and to choose their partners carefully when executing conversion initiatives.
Monticello Consulting Group understands how the elements of a successful conversion fit together and works closely with its clients to ensure conversions are executed to achieve superior outcomes that add value to the organization.
Structuring the Conversion Approach
Conversions typically start with the comparison of the data from the source system to the data of the target system. There are typically two types of data to analyze: transactional data (e.g. customer orders) versus static data (e.g. customer detail that does not change often such as name and address). Many conversions involve both data types. Regardless of the type of data to be converted, common activities in conversions include:
- Working closely with the business representatives who maintain and use the data that needs to be converted.
- Identifying system differences in storing/representing, managing, and processing the data.
- Data Cleansing. Data cleansing aims to make the conversion run more efficiently and the initial go-live run smooth, by the early elimination of invalid data.
- Data Mapping. This includes design of the data extraction methods, enrichment and mapping rules, the tools for supporting this and tools for loading the data into the target infrastructure. In recent years Monticello Consulting has partnered with clients converting data from relational databases to object oriented models, adding an extra layer of complexity to the conversion process.
One of the most important planning decisions for management is scheduling the conversion. Conversion in the production environment could disrupt the business at a critical point in its operations, putting the business at a disadvantage with respect to its competitors. If elapsed downtime required for the conversion exceeds the planned conversion period, there could be significant loss of revenues and/or reputation impact with its customer base. Early benchmarking of the conversion time required should be performed prior to defining the specific period for the live conversion.
One way to ensure that the conversion process will fall within proscribed timelines is to thoroughly test the conversion. An appropriate period of time should be allocated for the testing of conversion processes. This can be used to optimize the conversion schedule and to identify areas of poor data quality. Several mock conversions should be performed to ensure that it is possible to convert within the agreed period. Monticello Consulting has an extensive Testing Governance Framework for managing complex conversions. Please reach out to us for a walkthrough of our testing framework and best practices in this area.
In addition to thorough front-to-back testing of all the systems, the conversion plan/runbook should account for conversion volumes, historical data needs and timing requirements that may influence the conversion approach. The plan should make manual conversion procedures clear and concise especially where the use of automated tools is not possible.
The conversion process should include the appropriate controls and error/exception reporting. A typical control that Monticello Consulting has implemented for this purpose is a reconciliation tool and process for maintaining the integrity of the data between the source and target systems. This control aims to reject all records containing a serious error as soon as possible during the conversion process. Correction facilities should be provided during the conversion as well, and where not possible, the conversion runbook should include a viable rollback strategy at each important milestone in the plan.
Complex Conversions are a Reality
Not all conversions are executed in “Big Bang” fashion, for example, in a single conversion exercise during normal system downtime (e.g. nights and weekends). In addition, with e-commerce businesses, it is not acceptable to take the site down for the conversion of data on the backend. After an initial or partial conversion there is often the requirement for a subset of transactional data to remain on the source or legacy system while the converted population, in addition to new transactions, will flow through the target platform. We call this type of conversion having a “bifurcated environment.” Here is a high-level system architecture supporting bifurcation:
There are many business and technology drivers for operating in a bifurcated environment for both transitory and extended periods. Two examples include:
- The need to have new business transactions originate from the target platform before it is ready to accommodate legacy transactions. Digging deeper, often project-related obstacles had delayed or complicated the conversion off legacy systems and management had decided to go live with new transactions on the target system in advance of the conversion. In this example, time-to-market pressures for supporting new business can force IT managers to implement a bifurcated platform.
- The transaction lifecycle does not justify the conversion. Where the cost of conversion is too great and/or the lifecycle of the transaction is not excessive or onerous, management may decide to allow a subset population of legacy transactions to expire and to roll off the legacy platform vis-à-vis converting to the new system.
Because of the high cost of operating in a bifurcated environment, for most conversions bifurcation is a temporary state. In a dual (or more) platform environment, there is typically extreme pressure to sunset the legacy system in order to reduce IT costs and to provide the business with a single state-of-the-art system to conduct new business and manage legacy transactions. For some conversions, however, operating in the bifurcated state could continue indefinitely; for example, after the divesture of a business unit and the need to separate theparent company’s systems from the divested entity.
Contact Monticello Consulting today to discuss how we can help with your conversion initiatives.