Page 13 - FIGI - Use of telecommunications data for digital financial inclusion
P. 13

3.2.1   Subscriber identification data             calling and data plans and the frequency and value
            At the outset, operators typically collect and hold   of top-ups. Subscriber information also includes a
            information about the subscriber, particularly where   purchase history for a variety of additional products
            they are required to do so by know-your-customer   and services, such  as mobile devices and accesso-
            (KYC) regulation. Subscriber information will include   ries, mobile apps, and other services.
            a subscriber’s full name, address, nationality, tele-
            phone number,  and possibly an email address, a    3.2.3   Device data
            government-issued  identification number  (such  as   The telecommunications network will record the
            a passport or National ID number), and biometric   brand, model and operating system of the device
            information such as a picture, fingerprint, or copy of   a customer is using. This offers insight as to a
            a government-issued photo ID.                      customer’s level of consumer spending capabili-
                                                               ty or disposable income. Where linked to identity
            3.2.2   Order and billing data                     data, it becomes possible to link devices to the same
            Just as network equipment generates data, applica-  customer. While cross-device identity data is often
            tions used to manage telecommunications opera-     obtained using email addresses, social media logins
            tors’ business operations generate data. This includes   and validated linked accounts from different devices,
            sales data, payments data, trouble ticket data, churn   it can also be collected using data signals such as
            data, order and fulfilment data, and billing data.   matching locations, IP addresses, types of browsers,
            For post-paid accounts, operators have records of   and similarities of the operating systems. Cross-de-
            customer payment information and potentially other   vice data enables cross-device tracking, fuller data
            data such as bank or other transaction account infor-  about the customer, and more targeted messaging
            mation. For pre-paid accounts, which are particularly   for instance for customer engagement purposes (see
            relevant for financial inclusion, operators have data   section 5.1 below).
            as to decisions made by subscribers with respect to



            4  USE OF TELECOMMUNICATIONS DATA IN DFS

            Telecommunications data is used in multiple ways,
            sometimes on its own as reviewed here. Where                         CRM DATA
            people are already users of digital financial services,         Socioeconomic features
            they will have begun to establish a history of finan-  •  Age
                                                                  •  Gender
            cial transactions. This may begin with use of mobile   •  Estimated customer income
            money, in which case a significant amount of direct   •  High risk ZIP code
            financial behavioural history may be combined with    •  Regional area code
            the telecommunications data. Where the customer       •  Device brand  Product features
            has used digital credit, they will even have a credit   •  Device operating system
            history with the lender in question.                  •  Device type
               The combination of telecommunications data with    •  Line type
            such financial data is rich. The more the customer    •  Line status
                                                                  •  Line quantity
            builds a credit history with the lender in question, the   •  Late payments
            more weight will be given to that credit history, while   •  Month elapsed since activation
            the importance of telecommunications data in ana-     Source: Pedro, J. S., Proserpio, D., & Oliver, N. (2015).
            lysing creditworthiness will recede.                  MobiScore: Towards Universal Credit Scoring from
                                                                  Mobile Phone Data. In User Modeling, Adaptation
                                                                  and Personalization (pp. 195–207). Springer, Cham.
                                                                          .
                                                                                .
                                                                  https:// doi org/ 10 1007/ 978 -3 -319 -20267 -9 _16











                                                             Use of telecommunications data for digital financial inclusion  11
   8   9   10   11   12   13   14   15   16   17   18