We often get requests from clients who want to integrate data from a custom system, a website of another source into their accounting system. A common term used by our clients in describing the desired integration is a “one-step process.” The vision is a single click of a button and the requested data is taken from the source system and posted through to the accounting system.
However, in over 90% of the integration work we do, a second step is added, resulting in a two-step process. In general terms, it works this way. Once the data is extracted from the source system, a verification step is taken. Next, when the data is verified, it is pushed to the accounting system as a second step.
The reasons for the two-step process are quite easy to understand.
Source data may not be complete
Often data in a source system is not completely accurate. The data may be entered by employees or customers and may not match the data in the accounting system completely. For example, customer names or even customer numbers in the source system may not be the same as in the accounting system.
Pushing this data into the accounting system in a single step will result in errors or inaccurate data.
Very large corporations with massive data processing requirements have sophisticated and expensive systems to handle data integration between systems. The volume in those companies is simply too high to have much in the way of personal control over specific transactions or customer accounts.
We work in the small to mid-sized enterprise environment. In this environment, there is still a requirement for a personal touch. An extra level of control is required to ensure that a customer’s personal preferences are respected. This requires more control over the data transfer process – more control than can be offered in a one-step approach.
Control over transactions transferred between systems is best handled with the verification step mentioned above. The company can use this step to verify, not only the accuracy of the data transferred but that special customer’s transactions are recorded the way that customer prefers.
Again considering the example of very large enterprises, highly sophisticated systems can be developed and deployed because the budgets for such projects are large. Since the volume of transactions is very high and the benefits of two-way integration more apparent, larger budgets can be justified for very large corporations.
Integration budgets for small to mid-sized enterprises are generally quite small and the volume of transactions is a fraction of what large corporations deal with. A very effective way to control the development budget for an integration project is to introduce the verification step where some manual intervention is enabled.
When we think about dancing we may picture Fred Astaire and Ginger Rogers. If you have the time and budget to develop that level of sophisticated dancing with your data integration projects, go for it. If you’re like most of us and have limited time and funds, go for the two-step approach. You’ll get an effective system and who knows? You may learn to love country music!