This is one of our favorite projects to do for customers: Forecast future product order volume based on price changes now.
Forecasting Price-Volume Changes
It's a common business question: If I increase the price of my product, will my sales volume of that product change? If so, how will it change? Are there any price thresholds that cause volume to dip dramatically?
This type of analysis is one of our favorites. The impact is great and gives decision makers the data-driven information for one of the most rudimentary questions in business: what can I charge for my product?
In these solutions, we tend to use machine learning regression models for explainable entry level efforts with great results. For more advanced groups we can utilize deep learning or neural nets, but we have found explainability is an important feature early on in the maturity of this analysis.
In the effort we combine sales data with other data sets such as weather and market data. We build key features: price dimensions, current/previous volumes, the ratio of price and volume changes, and product details (to name a few). This will allow us to predict future volume changes within a defined time period.
The deployment of this solution ranges based on its maturity and business confidence level. We traditionally start by building a tool that allows financial analysts to run different scenarios and forecasts to see how price impacts volume.
Once confidence is built and the model matures, we build lightweight applications that have the ability to automate pricing. Not everyone gets to this level of automation and it is only targeted on a case-by-case basis.
The results and benefits for price-volume forecasting are vast. Our customers have seen revenue and profit increases due to the model outperforming previous methodologies of pricing.
Are you interested in price-volume forecasting with data science at your organization? Set up a 30-minute free consultation with one of our AI experts HERE.