Master data management (MDM) for telecoms

Posted: June 2007 C: Internet Posted by: K. Panayotakis

The case for MDM in telecoms

MDM is a relatively new buzzword which is based on traditional information management principles, like: • The aim for the single version of truth • The need to consolidate information on a business entity in a single ‘reference source’ or master source. • The fact that replication of information may lead to reduced information quality

MDM reflects the need to better structure the information architecture, in order to support enhanced business intelligence and operations.

Introducing MDM is easily proposed, but hardly implemented in Organizations with a complex enterprise architecture, which is built to serve the operational needs rather than the business intelligence ones.

A major MDM decision relates to the 'atomic' level, at which business entities are to be handled. In a telecom operator these may be: customer, service. This means that all information on a specific customer or service instance should reside in a single reference data source.

The complexity of a telecom business

The complexity of a telecom business is very high, given the range of customer segments handled: • Residential Customers • Business Customers which are further divided into subsegments • Partners who resell the telco business’s products & services • Other telecom providers which have a wholesaler relationship, Customer segments which have very different needs and are targeted with different services / products and price plans.

Furthermore the different CTP’s that may be involved in the customer relationship: Call centers / Account managers / Shops / Web, and the numerous service request use cases which may differ per product family, complicate further the operations.

Moreover the complexity of a telecom business is very high, given operations (focusing on some of the most critical process areas) like: • Customer relationship management: a telecom operator with a complex product portfolio offered to a complex set of customer segments (residential & business) via a complex organisation of Customer touch points (CTPs), has to deal with the support of very complicated CRM operations on: customer inquiries, order management, sales management, billing & collections, problem & fault handling, QOS/SLA management. Handling different customer segments and channel segments can be a very challenging task. • Service fulfillment involving various departments and different telecom networks or network components • Telecom network resource management: a telecom operator has to manage efficiently complex telecom network resources, in order to optimize the telecom service offered: improve the quality of service, reduce the time to order fulfillment, reduce the number of network infeasibilities. The complexity of various telecom networks owned by a large operator (e.g. fixed PSTN, mobile telephone, data networks (leased lines, frame relay, IP, ATM, MetroEthernet)) and the involved resource-facing business units (respective technical departments) • Partner relationships management: a telecom operator is exchanging information on network usage (CDRs) with other operators and may be reselling products of other telco providers. Order management and billing systems of both parties, need to be synchronized.

Major Telcos generally manage thousands of discrete business processes, to run their operations. To automate a subset of critical processes, they typically operate tens or hundreds of BSS/OSS software applications. Moreover, to achieve end-to-end process automation, application integration is required. Consequently, process automation complexity is high.

Applying MDM in telecoms

It is rather easy to apply MDM in Organizations, using a single or a couple of reference databases.

However, when the Organization is using tens or hundreds of systems, the same information class may be replicated in various systems. For example Customer information can be found on the CRM / Billing / Service provisioning / CTI databases and many more systems which may not be fully integrated.

Even worse, information on the Customer may be dispersed in various systems, leading to the absence of a data source which has the whole available info on the Customer. For example: billing data residing on the billing system / installed base residing on the service provisioning / payment history residing on an accounting system etc.

Having limited info on the installed base in the CRM

In order to apply MDM, the enterprise architecture needs to develop a higher degree of integration. For example an order for a new service is forwarded from the CRM to the service provisioning system. After the service is installed, the installation info should be forwarded to the CRM in order to build the full customer profile in a reference source.

CDI is an approach to aggregate isolated customer data in order to build the master data source for the Customer entity.

Obstacles in MDM

Cases in which 'info silos' handling information pieces of the same business entity which cannot be linked together. This can be due to the lack of cross referencing codes (i.e. different customer ID's for the same Customer used by different business units or channels). Moreover anonymous activity on the web channel cannot be associated to any Prospect or Customer data.

MDM may not be enough

All master data on a business entity may be located in a single database, but the data structure (or info architecture) may not reflect the actual real business entity relationship. The common example found in a customer data structure: duplicate customer entries which reflect the same customer. (more on this on customer information architecture).

An interesting analysis on architectural approaches to develop MDM is given by a recent article of Warren Thornthwaite.