Slide 1: Value for Money Exceptions Report
Extensive economic research and analysis conducted using the Ministry of Housing, Communities, and Local Government (MHCLG) datasets on local authority revenue expenditure and financing. This work led to the creation of an exception report, which identified specific functions within a local authority that were significantly more expensive than those of peer authorities (CIPFA NN group) and regional and national averages. As a result, the findings prompted further forensic analysis of expenditures and scrutiny of provisions.
Slide 2: Hypothesis testing / Pareto Analysis
This piece of data science financial insight was developed for a Human Resources and corporate management team of a large organisation facing challenges related to staff sickness, which was causing financial strain and lost productivity. Pareto* analysis, also known as the 80/20 rule, revealed that just under 20% of the staff accounted for most of the repeated sickness absences. The analysis identified which specific occupation types and functions within the organisation were most affected. The work also highlighted the importance of targeting interventions to address the underlying issues contributing to staff absenteeism.
*Pareto analysis is a technique employed to identify the most significant factors within a dataset or process, based on the observation that a small number of causes typically lead to a large percentage of the effects.
Slide 3: HE Student Cohorts Analysis / Visualisation
This visualisation was created for a university that was struggling to turn its rich internal data into actionable insights. It represents the culmination of work I undertook analysing student outcomes across undergraduate and postgraduate qualifications. While several visualisations were produced, this one had the most significant impact and prompted a change in how the university handled early student registrations. The work highlighted the enormous financial implications of students who start courses but do not complete them, both for the institution and for the individuals involved. Although not shown here, specific courses and pathways exhibited particularly high attrition rates, which I also visualised as part of this work.
Slide 4: Family Cost Tracker
The dashboard was developed by connecting to a dynamic dataset that is updated daily to track the costs associated with hosting vulnerable families who receive significant support from a local authority. This dashboard brings the data to life, enabling users to filter and analyse it effectively. It helps track cumulative costs and estimate costs for new families entering the support program. This capability is highly beneficial for operational purposes, enabling the team to stay within budget and providing early warnings when costs suddenly spike beyond expectations.
Slide 5: Prediction Tool for Future Financial Costs of Adult Social Care provisions
I project-managed and co-developed the technical aspects of a tool to help social care finance managers predict future costs associated with adult social care provision. This tool uses a Markov Chain transition matrix behind the scenes to predict the probabilities of clients transitioning from one care state to another over time. For example, individuals currently receiving social care who meet specific criteria may soon require a more expensive level of care, such as nursing care. This intelligence and insight can mitigate the effects of unexpected sudden spikes in care costs. End users can input various parameters, including average costs for a range of care packages, anticipated inflation rates, and the number of clients and their current provisions. The tool then generates forecasts that estimate costs over the next 2, 5, or 10 years, facilitating more informed financial planning.
Slide 6: Unit Costs Dashboard
Dashboard developed to allow finance managers at a local authority to interrogate the unit costs of delivering different services and compare those unit costs to those of other authorities. The tool also allowed end users to change the calculation between unit costs based on delivery per head of population, per dwelling, or, for some metrics, by length of the highway network. The work was well received, allowing people to see where the authority may have been paying over the odds for contractors and providers or where staffing costs may have been too high compared to peers.
Slide 7: Service Restructuring Savings Tracker
A dashboard I created to provide a corporate transformation team with insights into the planned, predicted, and actual savings related to a two-year organisational restructuring program. This dashboard visually summarises the data for a monthly highlight report aimed at senior leaders. The information is displayed by department and overall, and it includes the costs of redundancy payments, expected future savings, expenses related to the consultation process, and the net gains or losses over five years.
David Cook 