knowledge centre

Analytics applications provide rich functionality and low-risk deployment to help drive new use cases

Justin van der Lande Principal Analyst, Research

CSPs should consider analytics applications for specialist use-case requirements, but need to ensure that the underlying platform and tools are compatible with their data infrastructure and enable customisation.

Analytics

Traditional analytics tools enable skilled users to incorporate new data sources, develop new models, present the processed information and potentially link this through to an automated process. More-common requirements or 'use cases' have associated templates or blueprints in order to reduce the work required to build each of the steps. These include adapters to common data sources, the development of appropriate data structures and analytics models that reflect common uses, such as churn. Analytics applications take this concept through to its logical extension by combining all of the key components together in a vendor-supported 'application' (see Figure 1). These 'analytics applications' provide rapid, reliable access to functionality for a communications service provider (CSP).

Figure 1: Analytics application components [Source: Analysys Mason, 2013]

Figure 1: Analytics application components [Source: Analysys Mason, 2013]

Analytics applications encapsulate industry knowledge

Analytics applications take industry best practices and encapsulate them into a pre-configured software package, without the need to involve expensive, scarce data scientists. As with all applications, several pre-requisites need to be in place before deployment is possible. Vendors that have adopted an application approach have created a 'platform' on which the applications must be deployed. This provides the APIs, data infrastructure and interfaces that the applications utilise.

However, the applications are not without their flaws, which may make them less suitable for some CSPs. The shortcomings include:

  • CSPs will need to have the flexibility to change their processes to match those created by the application
  • the new underlying data platform on which the analytics application sits may not be compatible with currently deployed data infrastructure
  • they lack the flexibility to incorporate an individual CSP's requirements
  • CSPs will have to rely on a vendor for updates to the application, and be comfortable with sharing this resource with other CSPs
  • CSPs that are trying to rationalise their analytics tools will be adding additional tools to their IT infrastructure and will need resources to support them.

Compelling business cases will overcome these issues in cases where the vendor can provide deep industry knowledge and the application can deliver value swiftly – particularly when the more-generic tools would take longer and the associated vendors do not have an understanding of the use case.

General analytics tools are always needed

CSPs will always need general analytics and business intelligence tools. They often use such tools to provide customer insight and support segmentation initiatives – both areas in which vendors can apply knowledge from other industry verticals, such as retail or finance. Large CSPs have made significant investments in staff and infrastructure, and need the flexibility to create and support their specific use cases. Complex analytics models have been developed and refined over the years specifically for their environments. CSPs often have multiple analytics tools deployed – even within a single department – so are accustomed to supporting complex and diverse analytics environments.

CSPs should consider analytics applications for specialist use-case requirements, but need to ensure that the underlying platform and tools are compatible with their data infrastructure and enable customisation.