Applying analytics to ensure best practice resource optimisation and heightened societal return on investment.
On occasion, prisons have to operate at over capacity, with more prisoners than available beds. While the correlation between inmate numbers and prison officers is normally linear, Corrections had observed that this changed when a prison was full - with more staff time required for things like finding temporary accommodation and co-ordinating prison transfers. Although this could be managed using overtime, it was costly and not sustainable. Concerned that the appropriate staff funding was in place to manage occasional over capacity situations, Corrections tasked Datamine with creating a robust business case for Treasury that accurately estimated the extra funding required to cover projected increases in prisoner numbers.
Datamine developed a data-driven approach to ensure a sound business case. The solution correlated payroll data, hours worked, inmate numbers, facility capacities and Ministry of Justice forecasts to predict future staff costs associated with an increase in prisoner numbers.
The analysis revealed that staff costs increased from less than $380 per prisoner per week when occupancy was below 99% to over $420 per prisoner when occupancy rose to over 102%. When the prisons are nearly full, a 3% increase in the number of prisoners results in a 13% increase in staff costs. Quantifying these costs allowed a clear case to be put to Treasury.
To help Corrections continue to manage the increasing prisoner numbers, and provide an early warning of expected increases, Datamine also delivered an easy-to-use forecasting tool. Using this application Corrections now has a six-month forecast of expected prisoner numbers which assists in operational planning.
With overall responsibility for providing no-fault personal injury cover, a government-owned insurer wanted to develop a more customer-centric approach to the way it engaged with small businesses and asked Datamine to clearly establish two things:
- What was most important to small business clients in relation to the delivery of the insurer's services?
- How was the insurer performing in meeting those needs?
Being able to measure these variables and then identify distinct groups of customers with similar needs, the company hoped to gain actionable intelligence that would enable it to develop services that would increase overall customer satisfaction.
By combining the insurer's existing small business customer data with BRC Research survey data relating to a sample of those customers, Datamine was able to develop a small business customer segmentation model for the company that answered both questions.
Datamine analysis delivered the following rich information:
- Primary and secondary client drivers – in terms of what aspects were most important to customer groups in their dealings with the company (by extrapolating the survey data across all customers).
- Business customer demographics – including industry, number of employees, turnover, and company age.
- Transactional satisfaction data – with regard to previous interactions and outcomes around dealings with the insurer.
ESR is the sole provider of forensic science services to the New Zealand Police, and frequently undertake work for other Government agencies and commercial partners. Within ESR, the Forensic Science Unit (FSU) analyses human tissue, crime-scene trace evidence, bodily samples and any other evidential material, with their comprehensive knowledge of the presence and interpretation of DNA utilised across the country and around the world.
Offence Against Person (OAP) cases are assigned to the FSU with a single accompanying time-based Service Level Agreement (SLA). This assumes that all cases are uniform in their complexity, when in reality there are multiple inputs of varying sizes, quality and types per case, over a period of time. In addition there are external drivers that lead to ad hoc prioritisation of cases.
ESR tasked Datamine with providing a comprehensive understanding of cases, including patterns of inputs, timeliness and case profiles. ESR plans to use the results to improve efficiency of the FSU, and enable the informed setting of SLAs.
Using anonymised operational OAP metadata, Datamine defined and analysed key measurements and variables for TAT (turnaround time) of cases across a defined period of time. ESR had recently updated their internal processes, thus the goal of the analysis was to quantify whether the updated processes had led to an improvement in TAT, as well as identifying additional areas for improvement.
Datamine’s discovery analysis identified patterns and lags between time events and pathways in each value chain. For example:
- the number of business days from first exhibit received and last report sent,
- numbers of containers, analysis per case, exhibits and sub-exhibits per case
Key interpretations reported the number of cases completed within the prescribed SLA and the sort of cases where the SLA was infeasible.
Recommendations made to ESR include:
- Setting individual SLAs for cases based on case-specific variables such as the number of exhibits in the case
- Aiming to minimise time between exhibits received and analysis start, in order to reduce the average TAT
Datamine’s analysis has been used as a bench mark for other case types and has enabled discussions between ESR and the police to address operational lags in the process.
ESR’s focus is now on understanding the interrelationships between cases and the impact they have on each other with the goal of being able to predict the time a case will take depending on the inputs to the case in any given time so that realistic expectations can be set.
With the introduction of an international airport, a small tourist city had become a gateway to its region and local government leaders wanted to use this opportunity to create economic growth in the region – particularly in its retail sector. To do so it needed to understand the top-line differences between the retail sectors of itself and other tourist cities. Anecdotally, the Council believed it didn’t have the right ‘retail mix’ to service the new tourists, and by identifying both opportunities and threats, appropriate action could be taken to better target the region’s retail offering.
Using Westpac Business Insight data, Datamine sized the market and spending habits of residents of each destination and analysed the current retail offering in all the requested tourist cities - highlighting the critical differences between them.
For example, Datamine identified that while mainstream clothing & footwear retailers dominated in the client city, there was a noticeable lack of specialist stores present – a key driver in the ‘out of town shopping trip’ - which saw 11% of the city’s residents regularly driving 60 kilometres for retail therapy in another town.
Datamine also noted that the top three restaurants in a similarly sized tourist town represented about the same share of ‘top 10 food destination sales’ as the three big fast food chains in in the client city, signalling an opportunity for additional food revenue if city improved its ‘tourist food destination’ offering.
The District Council understood what it needed to do to increase spend in the city by locals and tourists alike, and began working towards attracting specific retailers to fill the gap in its retail offering and fuel economic growth in the region.
Fire and Emergency responds to a wide variety of emergencies every day. To ensure the right resources are allocated to its stations Fire and Emergency conducts community risk assessments — evaluating the potential for fires occurring — and their likely intensity. In recent times, however, Fire and Emergency found its risk assessment methodology wasn’t reflecting the reality of its emergency calls. Fires and hazardous materials were certainly part of a community’s risk profile, but so were natural disasters, medical events, motor vehicle accidents and rescues— scenarios that were not being factored into the existing methodology.
To assist Fire and Emergency in optimising its resource allocation, Datamine created risk profiles for the much wider range of emergency scenarios actually attended. By combining census data with information sourced from 10 years of actual emergency call outs Datamine was able to create a model that predicts future incidents — and their severity.
Understanding the impact of a broader range of variables on a community’s risk profile provides Fire and Emergency decision makers with the ability to compare capital spending on otherwise unrelated items. Would $1M be better spent on thermal imaging cameras, for example, or gas detection sensors? And what spend mix would provide the optimum result? With the risk profiles created by Datamine, Fire and Emergency is now better positioned to provide objective information when discussing resource issues with communities, unions, and other stakeholders.