Commercial sales are influenced by little details.
Sales data contains hidden links and connections that associate and group products. Exposing these product relationships informs marketing strategies and referral systems. Use Graphext to model the value of customer communities with RFM analysis or expose associations between items through product basket analysis.
Projects
Finding undervalued used cars
This project models data from 50,000 used car listings to find undervalued vehicles. Using function engineering to calculate the influence of factors such as age, make, mileage, power and color on the value of a car, this project can highlight hidden opportunities to buy used cars.
E-commerce and RFM analysis
This project segments the customers of an e-commerce site using topicality, frequency and money spent. The clustering model groups customers according to their interactions with the site, allowing the identification of specific customer sub-communities.
Loans in Germany and credit rating
This project reverse engineers the characteristics of 1000 German loans to understand the most influential characteristics for lenders and banks when considering the risk profile of an application.
Analysis of Disneyland customer reviews
Extraction of language features from 30,000 Disneyland customer reviews.
