The conventional approach to data organization within a business is largely correlated to the original operational intent. Whether the data is collected or created as a result of executing a business transaction or as a result of managing operational activities, in most cases the information is effectively a byproduct of the transaction or operation. In turn, the “management” of that data is limited to archiving – capturing the record of the event and maintaining it in case someone needs to examine it in the future.

However, in an organization that plans to adopt a customer-centric strategy, that same data seen as a byproduct of operational processing becomes a critical asset. The challenge, though, is that the typical approaches to organizing data for transaction processing are not necessarily suited to analytical processing, for a number of reasons, such as:

• The quality of the data is sufficient for the execution of the transaction, but may not be satisfactory for the analysis.
• The underlying models may not capture the desired data attributes that are needed to profile the customers.
• The level of precision in identifying customers uniquely is not relevant to the transaction processing and fulfillment processes.
• Demographic data to be used for analysis may not be needed to execute the transaction.
• Information about relationships, connectivity and conceptual hierarchies is not needed for transaction completion.

Customer profiles employed within a customer-centricity strategy rely on some key data artifacts organized in a way that supports the analyses that segment, characterize, classify and group customer data. That means that from the beginning, the analysis is going to require organization of customer data along specific facets, including:

• Entity and identity data, which first enables you to define a specific representative model for a customer and then select those attributes that can be used to uniquely differentiate any specific customer from all others;
• Attribution data, which are the variables that can ultimately be used as part of the analysis for clustering, segmentation and classification;
• Relationship data, comprising logical groupings (such as a “household”), hierarchies (such as a company’s org chart) and relationships (such as a social network);
• Behavior data, consisting of the collected sets of transactions, interactions and even the reactions that take place at the different customer touch points.