CHALLENGE

When attempting to forecast business results, my client was confronted with the complexity of the real estate market. As a result, predictions were often missed by a substantial margin. I was tasked with developing a model to help forecast the bottom-line on the basis of external market factors.

SOLUTION

I first performed a meta-analysis of existing studies to distill the real estate market’s main drivers, then extracted a number of years worth of data from various public sources. I blended this data with internal client data to assemble a complex dataset. Using the dataset, I developed and trained logistic and multilinear regressions model. The model is retrained as more results become available, and gets more accurate as more data points are added.

RESULTS

The model is able to predict business results with 99.9% accuracy. The model has enabled:

• Increased channel diversification, further improving acquisition

• The ability to scenario plan with confidence

• More accurate setting of acquisition targets