Forecasting service utilisation to guide funding and ordering
Funding and resource orders were based on utilisation estimates that were costly to get wrong. We brought scattered utilisation data together and built a model that improved forecast accuracy by at least 50%.
- more accurate utilisation forecasts
- 50%+more accurate utilisation forecasts
- resource ordering based on expected demand
- Targetedresource ordering based on expected demand
- cost savings
- Significantcost savings
The challenge
Hospitals and health services often allocate funding and order resources based on expected service utilisation. When that figure is an estimate, mistakes are expensive. Over-ordering ties up money in resources that go unused, while under-ordering leaves services stretched.
The data needed for a better forecast already existed, but it sat across many databases, was recorded in different ways, and was rarely analysed together.
Before and after
Before
- Utilisation estimated from partial or historical figures
- Relevant data spread across many separate databases
- Funding and ordering decisions carried significant uncertainty
- Money lost to over- and under-allocation
After
- Utilisation data analysed together in one view
- Forecasts at least 50% more accurate
- Resources ordered against expected demand
- Funding decisions backed by evidence, with significant savings
Our approach
- 01
Start with the decision
We looked at how utilisation figures were used in funding and ordering, so the forecast would answer the question that mattered.
- 02
Combine the data
We brought utilisation data from multiple databases into one consistent dataset.
- 03
Build and test the model
We developed a forecasting model and tested it against past utilisation before it was used for planning.
- 04
Put it into practice
We embedded the forecasts into the ordering process, so they shaped real decisions.
The outcome
Forecast accuracy improved by at least 50%. Resources are now ordered against expected demand rather than estimates, delivering significant savings.
Decision-makers also have more confidence in how funding is allocated, because the figures behind it are grounded in evidence.
