Real-time Analytics for Social Protection
Introduction
The Social Protection industry aims to reduce poverty and vulnerability by promoting efficient labour markets, diminishing people’s exposure to economic and social risks such as unemployment, exclusion, sickness, disability and old age, and enhancing their capacity to manage these risks. Many Social Protection agencies are promoting a "digital first" service delivery model, through which citizen needs can be addressed in real-time and people can be empowered to help themselves using online and mobile technologies, rather than visiting a Government shopfront. Operating in real-time requires agencies to leverage up-to-the-minute citizen information to support highly responsive decision making and to predict the likely outcomes of intervention and other activities. In-memory platforms, such as SAP HANA, have the potential to enable this through real-time predictive, similarity and big data analytics capabilities. This paper presents four use cases for Real-time Analytics within the context of the Social Protection industry:
1. Real-time Customer Segmentation;
2. Real-time Customer Risk Profiling;
3. Real-time Caseworker Decision Support; and
4. Real-time Social Policy Development.
Use Case #1: Real-time Customer Segmentation
In today’s resource constrained environment it is necessary for Social Protection organisations to segment their customer base to ensure that the right people receive the right level of service at the right time. A rule of thumb is that up to 80% of customers could be candidates for straight-through processing, where the customer is empowered through digital technologies to help themselves. The remaining 20% may have complex/special needs, requiring case management. The ability to analyse and assess customer circumstance data in real-time as it is being entered presents the opportunity for Social Protection organisations to identify and refer the 20% for caseworker assistance, and to fast-track the remaining 80% through a more automated process.
Use Case #2: Real-time Customer Risk Profiling
Risk is another reason that a customer might be segmented from the 80%, and it exists in two forms: social/economic risk and fraud/compliance risk. Social/economic risk is the risk to the customer that a change in their circumstance could result in an otherwise stable situation deteriorating rapidly. In such cases, it is critical to intervene as early as possible (ideally before the event) based on early indicators. Fraud/compliance risk is the risk to the Government that the customer might falsify their identity/circumstance to claim benefits to which they are not entitled. Again, it is critical to initiate fraud/compliance activities as early as possible (ideally before the event) based on customer behaviour and social trends. In both cases, predictive analytics can assist with the real-time identification of high-risk customers for segmentation and management.
Use Case #3: Real-time Caseworker Decision Support
Today’s caseworkers are supported by a myriad of decision support tools, some that have the ability to determine a customer’s eligibility to benefits based on their submitted circumstance data. In order to be useful however, these systems require a minimum data set to be applied to a well-defined ruleset – they do not work so well at the point of first contact when many aspects of the customer’s circumstance are unknown, or in the case of discretionary programs. In these instances, similarity analytics can be applied to recommend benefits based on successful outcomes that have been achieved by other customers with a similar profile. This capability has particular value within an eMarketplace and in support of real-time outreach/campaigns to target customers/cohorts.
Use Case #4: Real-time Social Policy Development
Another approach to optimising benefit outcomes is to adjust social policy to target the most vulnerable members of society, or to drive particular behaviours within a cohort. Typically new benefit programs might be created, or existing eligibility criteria, entitlement thresholds and payment rates might be adjusted by way of legislative reform. However, it can sometimes be difficult to predict the impact of these reforms in advance of implementation, when there is a risk of unintended consequences such as a particularly vulnerable cohort being worse off under the new legislation. Program simulation can help to mitigate this risk by enabling the organisation to adjust the levers of social policy in a real-time sandpit environment, to predict the likely outcomes of legislative reform during the drafting phase. The same capability can also be applied by service delivery agencies to inform and influence future Government investment in social programs.
Conclusion
Real-time Analytics has the potential to transform Social Protection service delivery through real-time customer segmentation, customer risk profiling, caseworker decision support and social policy development. In-memory platforms, such as SAP HANA, enable Social Protection organisations to leverage all available enterprise data when needed, and without impacting the performance of operational systems, to deliver improved outcomes for citizens in real-time.
