Telefonica believes that it has a data mining opportunity to equal OTT rivals
While Google has the search graph and Facebook has the social graph, Telefonica Digital says that it has the voice graph. At a Telefonica Innovation briefing at MWC 2013, the operator spoke about a project that uses a linguistic analysis engine to mine and analyze call data. Among other things, the engine can be used to track call location and duration, which is useful for measuring the impact of something such as an event in a certain area.
However, a far more controversial aspect of the project is the analysis of personal calls. Telefonicaâ€™s linguistic analysis engine can extract information about a callerâ€™s personality and preferences based on the tone of their voice. The operator claims that it has a team (which includes psychologists) that can read personality traits with 80% accuracy based on the results of the call analysis. This level of insight would clearly be of interest to advertisers and organizations in sectors such as tourism.
Telefonica is keen to stress that it does not look beyond a userâ€™s tone of voice, that its approach is based on opt-in consent from the user, and that the whole project is purely internal at this point. However, Ovum believes that a commercial launch for the project is likely to be challenging. Telefonicaâ€™s linguistic analysis engine provides a deep level of granularity, even without listening to actual conversations, and we think that consumers will be very uncomfortable and resistant to this idea. Consumers are increasingly concerned about how their personal data is exploited, and this is reflected in the mounting regulatory scrutiny of the issue. All of these issues are explored further in Ovumâ€™s recent report Personal Data Futures: The Disrupted Ecosystem.