The Role of OLAP in CRM
Customer relationship management (CRM) has revolutionized retail and the consumer life cycle in the past five years or so. These days, a comprehensive CRM solution will affect customers before, during and after their purchases, from the moment they become aware of the product or service for sale until a newer and better product or service comes along to replace it.
But CRM as a process would have remained relatively unrefined and impractical had it not been enabled by many different technologies, particularly related to databases. As a matter of fact, the dramatic impact of CRM on retail (especially e-commerce) took place soon after data warehousing tools and methodologies matured.
The data warehouse, which is essentially a computer system that archives chronological information based on analytical needs, can improve almost any aspect of CRM.
It can help a sales force manage leads, assist support staff with common problems or aid marketing professionals with breaking down consumer populations into demographic segments, to name just a few ways. (Data warehousing is hardly limited to CRM, though. It also can influence employee compensation decisions based on salary history, for example.)
Data warehousing, in turn, is enabled by Online Analytical Processing (OLAP), which optimizes reporting and organizing information for dimensional queries. This sounds complicated, but it isn’t.
Basically, it refers to the values produced by matrixed data — i.e. information placed in rows and columns. So what makes OLAP more sophisticated than, say, a spreadsheet?
The complexity of OLAP is in the way it’s organized. Its arrangements fall into two categories: Relational OLAP (ROLAP) and Multidimensional OLAP (MOLAP).
ROLAP is a data warehouse that typically stores information in mainstream relational database management systems (RBDMS) such as those offered by Oracle, Microsoft and MySQL. It also has tools added on that can facilitate simple queries. These solutions are often specialized, as they frequently are used for specific tasks such as niche marketing campaigns. They’re also fairly easy to configure.
Aside from relative ease, the main advantage of ROLAP is that it can store quite a bit of information. On the downside, it can be fairly slow and relies heavily on the SQL databases on which it runs.
On the other hand, MOLAP is known for its quick slice-and-dice functionality. Because of its multidimensional structure, it performs more complex calculations, which are built into the system from the get-go and performed repeatedly.
It cannot hold capacious amounts of data as ROLAP can, however. (The number of dimensions that different MOLAP tools can handle varies wildly, from the hundreds to fewer than 10.) Also, MOLAP generally involves many different proprietary technologies, which can make building these systems somewhat complicated.
There is a third solution emerging that is designed to combine the benefits of MOLAP and ROLAP. Predictably, it’s called Hybrid OLAP (HOLAP). In this schema, large quantities of detailed data are placed in the ROLAP for computing, and the summative information is then entered into the MOLAP for the final, 50,000-foot view.
Bottom line: If your organization is selling something, it’s probably already using a CRM solution supported by MOLAP, ROLAP or a combination of the two. If your company isn’t using one, then you need to get it up to speed — while you still can.