What is holding back telcos from better leveraging the vast volumes of data they already collect, but don't always analyze, to improve their services/operations and the customer experience?
Until the last few years, telcos have not needed to leverage their vast datasets to improve their operational efficiency. High subscriber growth and healthy margins meant operators were hardly concerned about efficient processes, says Justin van der Lande, principal analyst at Analysys Mason.
"The ugly truth is that they didn't need to. Only a short while ago operators were publicly owned monopolies and the demand for services outstripped the ability to supply. As markets mature and as growth has slowed, service experience and operational costs become the key concerns of operators."
He said technology has also played an important part, with falling storage and computational costs delivered via the likes of Hadoop, which offers the ability to work on massive datasets at a very low cost.
Analytics tools and data infrastructure now provide the foundation for improving processes, but he noted there is still a relatively low take-up. This is because implementation costs are still high. "Although the underlying costs of data and analytics data infrastructure have fallen, the analysis of data, the creation of data models and implementing changes to established processes are all customized for each operator and every process."
Tolaga chief research officer Phil Marshall argues the problem is that most of the data is stranded in operational silos.
For example, detailed network performance information is constrained primarily to network operations and used for the purposes of managing the network. He says similar silos exist for data from other sources such as billing and customer care.
"While this is starting to change, the horizontal integration and correlation of information across the silos is difficult, because each silo tends to have its own unique demands, fiefdoms and operational priorities. The technologies are available but depend on cross-organizational support. Typically the only way for this to happen is if it is driven by the CxOs," Marshall says.