To be sure, there are an ever-growing number of EDM methodologies that are marketed by software vendors and consulting companies alike. Most of the methodologies are perfectly capable of providing an approach to developing a comprehensive EDM solution. However, the solution to any problem is only as good as the clarity of definition of the problem allows it to be. If companies do not understand their needs before their solution is implemented, the solution is unlikely to solve the problem.
This month and next month, in an effort to facilitate a better understanding of the true, enterprise-wide nature of data management, I will discuss some of those implications and provide you with some strategies you should consider to help guide your EDM project toward fulfilling the lofty expectations that are surely set for it.
In this column, I'm going to focus on analyzing information needs, results metrics and information sources; key areas to focus on to measure results; and challenges that must be overcome on the path to an effective implementation. Next month, I'll discuss some technical innovations that are making EDM implementations easier, business functions and processes most impacted by an EDM initiative, and the components of a robust EDM solution.
Figure 1 represents which components of an EDM solution must be analyzed to truly understand data management implications across the enterprise.
Figure 1: Key EDM Analysis Areas
Let's start by looking at the enterprise implications of data itself. Data that is critical to understanding, managing and growing the business is resident in all corporate information systems such as data warehouses/marts, enterprise resource planning (ERP) and customer relationship management (CRM) systems, and myriad other systems that serve as sources of record for financial and operational data. Much data is also located in stand-alone departmental databases and applications and even in spreadsheets used for analysis. It is absolutely essential to determine what information users need to do their jobs and where (and in what form) that information resides.
Speaking of information users, it is also important to truly understand the needs of the user community - both in terms of what information they need and how they use information in the course of doing their jobs. Information needs and usage requirements will differ markedly based on job function. Managers will have different needs than analysts, who will have different needs than customer-facing workers. All of these potentially contradictory needs must be accounted for in the design of the EDM strategy and architecture as well as their implementation.
Once the scope of enterprise data and the needs of the user community have been defined, the next step is to understand how the data is used to monitor and manage business operations. Is the data used for research purposes? Modeling? Operational/financial analysis? Performance management? All these uses are different, and they must be accounted for when designing the EDM strategy and architecture. It is not enough to know about them, you must understand them and accommodate them.
In the actual design of the EDM strategy and architecture, there are three key areas that will play an important role in determining how effective your EDM project is at implementation and beyond. They are design, modeling, and delivery. Your concern for the design area should first center on how the actual EDM solution will be deployed - in other words how (and how well) the different software packages and technologies will be installed and integrated to provide reliable, consistent, structured data.
Figure 2 represents the key focus areas for EDM design.
Figure 2: Key EDM Focus Areas
You should also determine the comprehensiveness of any analytics component that has been designed. Will it meet the needs of all users? Is it flexible? Can it grow as the company grows? All these questions should be answered before the EDM solution is deployed.
The next focus area is the modeling aspect of the solution. Modeling capabilities are a subset of an overall analytics solution, but I'm breaking it out as a separate focus area because effective modeling/analytics is almost impossible without an effective EDM solution. It is essential in the design of the EDM strategy and architecture that you consider all present and any anticipated future needs that your company may have for modeling capabilities.
For example, do you have a neural network and predictive modeling initiative in place (or are you thinking about one)? Do your analytics capabilities employ association and affinity grouping or clustering and classification models? These are fairly sophisticated modeling techniques that rely on accurate information to make accurate predictions. It is vital to have the right information in the most accurate form available to supply these functions.
The final focus area, data delivery, is perhaps the most critical, because this is where all your work will be on public display. How, and how well, users are able to access the data they need and the format in which that data is delivered to them will play a large role in determining how the EDM solution is perceived throughout the company. If people can't access the data they need, in a user-friendly manner, your EDM solution will be perceived as ineffectual at best.
Be the first to comment on this post using the section below.