Download PDF
Dataiku > Case Studies > Smart Pricing in Retail
Dataiku Logo

Smart Pricing in Retail

Technology Category
  • Analytics & Modeling - Predictive Analytics
Applicable Industries
  • Retail
Applicable Functions
  • Sales & Marketing
Services
  • Data Science Services
The Challenge
A leading retailer in Europe with more than 3,500 stores and an e-commerce component was losing money due to being undercut by competitors on price. They also found that their customer base tended to wait until the end of seasons for huge markdowns and would only then buy certain seasonal products, which skewed their predictions for how to stock items in the future and perpetuated the pricing issue. In addition, they struggled to efficiently change prices and keep them consistent across stores and online - often, this resulted in inconsistent pricing, especially when individual store managers made their own decisions on sales. The retailer wanted to improve their pricing strategy by understanding what drove customer purchasing decisions for specific products and what prices would resonate best, easily understanding the price offered by all competitors in real time, and updating pricing consistently across stores and online.
About The Customer
The customer is a leading retailer in Europe with more than 3,500 stores as well as an e-commerce component that offers home delivery services. With hundreds of thousands of employees and customers spanning multiple countries, this retailer stays on the cutting-edge of big data technologies to remain competitive in a growing market.
The Solution
The retailer introduced Dataiku Data Science Studio (DSS) into their data team’s processes to incorporate predictive analytics at scale. They worked with Dataiku to produce a final data project that considers competitors’ pricing and uses it in a predictive model to determine, for specific products, whether the overall business can support aggressive price-based competition for that product. The solution leverages Dataiku’s REST API to adjust pricing in production automatically based on a specific set of predefined features. It uses real-time monitoring of models in production to ensure pricing model performance isn’t drifting and that pricing changes in production over time are well documented. The solution also includes a robust pricing dashboard based off of the predictive pricing model that alerts and allows physical stores to react to recommended pricing changes or online pricing changes.
Operational Impact
  • Reduced excess inventory by setting prices that allow products to sell better and predicting what price will be effective at what time.
  • Increased the productivity of brick-and-mortar store managers, who now use the pricing dashboard to automatically inform data-driven pricing changes.
  • Reduced the number of discrepancies between online and in-store pricing (using customer service complaints about those discrepancies as a proxy) by 65 percent.
Quantitative Benefit
  • 65% reduction in discrepancies between online and in-store pricing.
  • 10% increase in sales by using the model to determine optimal pricing at the right time.

Related Case Studies.

Contact us

Let's talk!

* Required
* Required
* Required
* Invalid email address
By submitting this form, you agree that IoT ONE may contact you with insights and marketing messaging.
No thanks, I don't want to receive any marketing emails from IoT ONE.
Submit

Thank you for your message!
We will contact you soon.