Redundant, erroneous, lost, and poorly coded data. Every company has these kinds of problems with its vital information residing in different origens. Although combining the information from different data bases is an easy task, it could obstruct some ERP, CRM, and the management of supply chain projects.
Although this is not a new problem, it is very noticeable in high-profile projects, vital for the business, such as; installing Enterprise Resource Planning (ERP) software, Customer Relationship Management (CRM), or the supply management chain. Some analysts say a majority of companies tend to focus on the commercial process, and not on the way, the logic, or the consistency of the resulting data. When a company acts like this, the IT department becomes frustrated, and abandons some projects to pursue cleaning, reconcilation, and integration of data from different storage or “silos” around the company.
For example, different selling systems, inventory or manufacturing in a clothing store could follow any given article under different names. A central data base (if existing) could name the articles using terms as “extra large” or “XL,” even though both are refering to the same item. At the same time, we can add what is referred to as the “back room,” very common in many homes. If enough boxes containing seasonal clothing, ornaments, family documents, and other things pile up, soon enough it will end up with many different articles difficult to manage. That can also happen in companies. Multiple operating units, manufacturing facilities, resources, and installations could manage different applications from different sales persons to carry out sales, human resources, etc. This combination of different data creates a pile of unclassified, and unsorted information.
Under one roof
When the time to integrate the information arrives, the question that many IT managers are raising in their companies is whether to get rid of the five or six applications that capture data, and installing a new one, or to collect the data again, clean, reconcile and organize it. Which is undoubtedly hard and complex work.
In most companies, each system has its own set of internal codes. Cleaning the storage data from these host systems is not usually an acceptable option. Considering doing so requires too much time, and is not cost effective. The most common solution is to find a third party tool to perform an adaptation of the data from the different systems, and to combine them in a data warehouse, so none of the business units lose control of their own data, and gradually integrate other applications.
Still, when a company decides to substitute multiple applications for a new one, to solve the data chaos, the process is never easy. Many of the major sellers of CRM and ERP products, for example, offer applications suites made by their own applications, plus others that they have acquired. As a consequence the products contained in these groups have not been created together, and therefore may not transmit data in both directions easily. In recent years, the ERP system’s providers are trying to solve the problem by providing data models, and datawarehouses with their application suites.
Something IT managers have understood when integrating all the data in one place or warehouse, is that the errors, and incidents from the past are now clearly visible. That is why justifying the cost of data integration projects is not difficult. The logic that is usually applied when this topic comes up with managers of a business’ units is simple. The argument that the data is what makes us what we are as a company usually works. Because, if clean and good data is supplied, better decisions would be easier to make.
Plan to adjust information
Before cleaning any kind of data, the IT department must establish a plan to localize, and to relocate data to be able to decide how to manage it once it is available. Those who have gone through this offer the following tips:
1.- Determine what kind of information needs to be captured. For that, a mapping, or locating of critical data committee for the business should be created.
2.- Find mapping software that can collect data from different sources. Including old existing applications, PC files, HTML files, unstructured data sources, and an enterprise business level like ERP.
3.- Start with a high profitability project. The best advice is the first integration project to be carried out in a business unit should generate high profits for the company. Sometimes you might think it would be easier to get started with a small company unit. However, you could not easily justify the cost to obtain the manager’s approval.
4.- Create a mapping, cleanning, and collecting process that could be repeated time, and time again. The aim is that everytime you wish to get an actual view, or image of certain data, it would be possible to generate a report, since data integration is not something that is done in one operation, and companies have to deal with it constantly.