In service provider environments, data is distributed over a large variety of expert systems that are designed for specific OSS functions, including network management systems (NMS) and inventory management systems (IMS). Although those systems are sometimes used for network business intelligence (NBI) needs, because alternatives often do not exist, they lack critical features of a business intelligence solution.

NMS are specialist systems that are
designed for the operational (configuration, provisioning, etc.) needs of a supplier
and technology specific network. Besides
requiring expert skills, a NMS is a mission
critical system. It does not tolerate extensive
query, reporting & analysis operations, typical
of NBI, to be performed.

IMS do offer some characteristics of NBI but
lack the immediacy and visual exploration
capabilities of a NBI. Most IMS recognize this
fact by offering interfaces and APIs to external business intelligence applications.

In addition, a NBI needs to extract data from
different systems because :

• The data might not be available in a single system (e.g. Alarm data might need to be extracted from an alarm correlator);
• The data in a given system might lack the required data quality. (e.g. subsets of data in an IMS might not be properly reconciled with the network data).

Finally a NBI solution analyses existing data and should not allow the user to create new data, because it inevitably leads to data quality issues. Low data quality is among the top reasons for revenue assurance fallouts at service providers. Manual data entry into a business intelligence solution must therefore be avoided. It also means that most of the data loaded into a NBI application consists of bulk data and that a NBI application does not need to support high volumes of transactional data flows. This allows the choice of building technologies that offer a better price/performance ratio than technologies typically used in NMS or IMS.