Strategic Forecasting for Infrastructure
Situation
A county on the East Coast experienced significant population growth from 2010 to 2020 with shifting demographics that created challenges in school planning. As the county expanded, the need for new schools became critical, but predicting where to build and how to allocate resources was difficult. With school construction costs ranging from $35M to $75M, the county needed a precise forecasting model to guide its decisions.
Challenges:
The county struggled with planning new school construction during rapid population growth and forecasting future demand.
Their existing cohort survival model predicted enrollments well for one year (90% accuracy) but wasn’t sufficient for five-year forecasts.
Previous vendors failed to deliver reliable models, leaving the county without support for long-term capital planning.
Solutions Overview
Conducted an 8-week pilot to assess whether data models could predict population growth and guide school planning.
Collaborated with the county’s IT, school board, and planning departments to develop and refine accurate models.
Created a live dashboard to provide officials with real-time data for decision-making.
Results
Innovative Forecasting Model
Delivered a groundbreaking machine learning model for hyper-localized five-year population growth forecasting, enabling smarter investments in school facilities worth hundreds of millions.
Recognition and Expansion
Earned a top ten award for most innovative counties and expanded the project to include planning for parks and libraries.
Adoption and Impact
Secured adoption of the forecasting plan by the school board and county, establishing a foundation for long-term strategic planning.