In order to fulfill the potential of Big Data, mobile operators need to commercialize Big Data services at scale. This means that Big Data teams which have previously been internally focused need to transform into commercially-oriented and externally-focused business units.
In our experience of working with mobile operators in over 30 different countries around the world, organizations follow a common set of steps in getting to the point of launching an externally-focused Big Data organization:
The Basic Tooling and Data Lab phases usually occur within the IT department, with minimal impacts on an operator’s organization and culture.
The Internal Testing phase is more challenging, and starts to reveal a key issue: who should control the Big Data budget – Marketing or IT? Presently, Big Data teams tend to sit within IT, but Big Data’s profound impact on a mobile operator’s customer-facing activities mean that the formation of cross-functional teams is essential for success, and that the budget must be driven from commercial imperatives.
The Commercial Prototypes phase requires new sets of skills in order to design new Big Data products and deliver those to a range of corporate and individual customers. Some of these skills do not exist inside most operators, necessitating a new wave of recruitment and the formation of partnerships. At this stage, many operators consider spinning out the new Big Data division into a separate organization (although it has been seen that this is not necessarily the optimal structure).
The Data Factory phase (which – in our opinion – no operator has yet reached) will require additional skills in terms of scalable systems, integration with other data providers and deep partnerships with marketing and CRM providers, as well as strong collaboration with various web players and OTT service providers.
Benchmarking: how does your organization compare to others?
The chart below shows the detailed characteristics of each Big Data development phase, along with Redwing’s estimation of the proportion of the world’s operators in each phase.
At this point in time, we do not believe any operator has yet implemented a Data Factory, and we do not expect to see this happen until 2015.
(Click on chart for larger version)