Construction cost data provider Gordian has launched a big-data-driven service designed to predict construction project cost changes up to three years into the future.

Predictive Cost Data, available in RSMeans Data Online, combines more than 15 years of RSMeans data and other indices from government and private sources, then takes into account factors such as project location and construction type and applies proprietary algorithms to deliver what Gordian claims are cost estimates that will hold their accuracy for years. 

To validate the forecasting during the project development, Gordian ran back-testing using data from the last 10 years, finding the accuracy to be within 3%.

“Traditional forecast data, developed during a time before the availability of big data simply does not meet the needs of today’s construction professionals for accurate planning and budgeting,” says Noam Reininger, chief data officer at Gordian. “Construction costs are volatile, and traditional economic forecast methods can fall short in predicting market swings or sharp escalations.”

Predictive Cost Data was developed to address those issues. Gordian data scientists can now  calculate what they say are accurate cost forecasts thanks to advances in computing power, data visualization and updated statistical procedures, Reininger says.

“We’re able to find patterns and determine structural drivers of construction material and labor costs,” Reininger says. “Our predictive cost models are based exclusively on data-driven empirical evidence.”

“Owners, architects, engineers and other professionals responsible for planning and budgeting projects now have a proven and reliable source for future construction costs,” adds Reininger. “This unprecedented ability to utilize future costs at the material, labor and equipment level allows users to make data-driven decisions in the planning and budgeting stage with a higher degree of confidence.”