C5i Case Studies Helped a B2B Retailer analyze potential customers at street level to increase sales conversion ratio

Edit This Case Study Record

	Helped a B2B Retailer analyze potential customers at street level to increase sales conversion ratio C5i

Helped a B2B Retailer analyze potential customers at street level to increase sales conversion ratio

C5i

Technology Category	Analytics & Modeling - Data Mining Analytics & Modeling - Predictive Analytics
Applicable Industries	Retail
Applicable Functions	Sales & Marketing
Use Cases	Demand Planning & Forecasting Predictive Replenishment
Services	Data Science Services
Challenge	The trading and distribution company wanted to increase its market share by adding new customers to its portfolio. The client did not have a comprehensive list of the customers which they were not servicing. The company aimed to increase their customer base by identifying potential customers on a street level basis and obtain a mix of existing and potential customers to identify and target the segments which would be most profitable. Read More
About Customer	The client is a B2B Retailer company operating in the retail industry. The company is involved in trading and distribution and is looking to increase its market share. The company's goal is to add new customers to its portfolio, but it lacks a comprehensive list of customers it is not currently servicing. The company aims to increase its customer base by identifying potential customers at the street level. The company also wants to obtain a mix of existing and potential customers to identify and target the most profitable segments. Read More
Solution	The solution involved using Web Crawling/Web Scraping technique to get the exhaustive list of customers available in that region. Variables such as menu price, longitude & latitude, customer group, chain/non-chain and derived variables (distance from the nearest existing customer) were used for further analysis. A highly interactive web app was built using R for easy viewing of existing and non-customers along with magnifier features like zoom in & zoom out. The concentration of both existing and non-customers was analyzed to identify low or high coverage districts and/or streets. A model was built using SVM Classification technique to find the probability score of non customers and categorize them into ‘high’, ‘medium’ and ‘low’ potential customers. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Data Mining

Analytics & Modeling - Predictive Analytics

Applicable Industries

Retail

Applicable Functions

Sales & Marketing

Use Cases

Demand Planning & Forecasting

Predictive Replenishment

Services

Data Science Services

Challenge

The trading and distribution company wanted to increase its market share by adding new customers to its portfolio. The client did not have a comprehensive list of the customers which they were not servicing. The company aimed to increase their customer base by identifying potential customers on a street level basis and obtain a mix of existing and potential customers to identify and target the segments which would be most profitable.

About Customer

The client is a B2B Retailer company operating in the retail industry. The company is involved in trading and distribution and is looking to increase its market share. The company's goal is to add new customers to its portfolio, but it lacks a comprehensive list of customers it is not currently servicing. The company aims to increase its customer base by identifying potential customers at the street level. The company also wants to obtain a mix of existing and potential customers to identify and target the most profitable segments.

Solution

The solution involved using Web Crawling/Web Scraping technique to get the exhaustive list of customers available in that region. Variables such as menu price, longitude & latitude, customer group, chain/non-chain and derived variables (distance from the nearest existing customer) were used for further analysis. A highly interactive web app was built using R for easy viewing of existing and non-customers along with magnifier features like zoom in & zoom out. The concentration of both existing and non-customers was analyzed to identify low or high coverage districts and/or streets. A model was built using SVM Classification technique to find the probability score of non customers and categorize them into ‘high’, ‘medium’ and ‘low’ potential customers.

Impact #1	The client was able to visualize the potential non customers at district/street level to identify areas with higher potential for proper resource allocation.
Impact #2	Identified customer segments which have the highest potential and then target the potential noncustomers with products already identified from the product mix clustering.

Impact #1

The client was able to visualize the potential non customers at district/street level to identify areas with higher potential for proper resource allocation.

Impact #2

Identified customer segments which have the highest potential and then target the potential noncustomers with products already identified from the product mix clustering.

Benefit #1	Street view analysis helped to identify 56% of non customers who should be targeted first as their probability of conversion was highest.

Benefit #1

Street view analysis helped to identify 56% of non customers who should be targeted first as their probability of conversion was highest.

Download PDF Version

Overview

Helped a B2B Retailer analyze potential customers at street level to increase sales conversion ratio

Operational Impact

Quantitative Benefit