knowledge centre

Creating a digital service provider using a big data analytics platform

Justin van der Lande Principal Analyst, Research

This report analyses the changes that communications service providers (CSPs) need to make to their data analytics tools and environment when aspiring to become digital service providers (DSPs). The report considers the business drivers for CSPs becoming DSPs and the different aspects of the transformation needed. CSPs must be able to support new business cases in order to compete with web-scale providers and must therefore adopt the same big data technology.

The report provides insights into the different layers and functions that a big data analytics platform should have. This covers the underlying data storage technology such as Hadoop, that underpins the platform and segments the platform into three layers. The three layers include a shared logical data model, the tools to create and run the logic/algorithms and the publication layer to expose the results to users. Deployments of a big data platform can be achieved using a 'big-bang' or step-by-step approach, with more advantages achieved with the former because it has a clearer remit and processes are simpler.

The report looks at internal and external drivers for a big data analytics platform. The internal drivers are related to improving processes and making better use of resources, whereas the external drivers look at the generation of new services and data monetisation.

We include a detailed case study from China Unicom Sichuan province with an outline of its solution, the benefits that it has gained from its 'Big Data Center' platform and which services they are planning to provide soon.