HP | Data Analytics & Propensity
The end customer: Central Government’s Cabinet Office
HP worked with Hito to transform the Cabinet Offices operations into a more streamlined, efficient business that could function more like a modern, smart organization.
Concern that the organisation was beginning to fall behind the private sector in terms of its capabilities for improving recovery rates of overdue debts – and doing so in the most cost effective manner possible.
A pilot programme invited technology and service providers, via a series of hands-on projects, to come in and showcase their capabilities. Of particular interest were analytics, and the role they could play in helping reduce unnecessary costs while also increasing collections.
Hewlett Packard (HP) was asked to take part in the pilot and selected Hito as the ideal provider for showcasing their capabilities. As part of HP’s portfolio offering, the Hito Edge solution offered the data analysis and propensity function, while another HP partner delivered the resources needed to undertake the recoveries.
Hito’s approach to analytics started with the development of bespoke, statistically derived, propensity models. To do this Hito used a combination of data provided by the client as well as other sources.
This approach was made possible by the Hito Edge analytics team’s in-depth knowledge of a variety of debt types and customer attributes – insights that collectively helped inform the likelihood of successful collections, and where activity could be reduced and first contact success rates enhanced.
In addition, Hito’s team alongside the client’s end-users were able to design and build a propensity solution. Armed with the outputs, and wrapping around them external data e.g. that held by credit reference agencies, Hito was able to undertake statistically derived propensity (SDP) modelling – data that is then re-modelled for each debt type, as well as by environment and geographical location.
Hito’s ensures their SDP modelling makes it easy for clients to interpret the outputs. This is confirmed through the use of a linear regression based approach, where the customers ‘DNA’ can be extracted for each and every case – ensuring the rationale for any course of action is always fully understood.
The models for the pilot were created within Hito Edge, which uses built-in, real-time propensity scores, which alongside their criterion help drive the correct decisions and strategy for collections. This helped cut out any unnecessary activities and costs while increasing the rate of collections. The entire solution ensures each employee is able to work to their full potential throughout each working day.
The outcome of the pilot project was successful, and lead to the Cabinet Office’s decision to create a route for central government to use the private sector. This resulted in a joint venture between the Cabinet Office and TDX Group.
HP had brilliant result with the pilot project, using Hito Edge they predicted 35,000 cases that were ranked into 10 streams – with the lowest probable 10% of cases being in decile 1 and the best 10% in decile 10. The graph below shows the accuracy of the model from what was predicted as collectable: