New Media Pack Download
Grocery Retail Structure 2009Data Services – Test Case
Objective:
• To gauge ‘the customers’ penetration of key independent market sectors
Data Cleaning:
• Pre Clean – tidy of current Customer Data
Removal of duplicate records
• Ensuring data is in correct fields to enable data to be run against Postal Address File
All data is verified against the Postal Address File
• Any un-PAFed data was manually researched & cleaned
Data De-Dupe
• The Knowledge Store’s data universe was matched against that of ‘the customer’s’, including business sector
codes and business type
• The ‘the customer’s’ data was then de-duped against The Knowledge Store’s data universe
Data Enhancement
• Any grey area data was manually checked for correct business type and changed accordingly.
• Any demographics selected by ‘the customer’ were then be added from The Knowledge Store data universe
Sector Profiling
• Stores were then profiled
By grouping of store attributes i.e. sells alcohol, has a chiller etc.
• Stores were then ranked
By the overlaying of additional demographics such as proximity data (i.e. near a bank or university or A road)
• Determination of sector profiles giving sector size & sector market value
• To gauge ‘the customers’ penetration of key independent market sectors
Data Cleaning:
• Pre Clean – tidy of current Customer Data
Removal of duplicate records
• Ensuring data is in correct fields to enable data to be run against Postal Address File
All data is verified against the Postal Address File
• Any un-PAFed data was manually researched & cleaned
Data De-Dupe
• The Knowledge Store’s data universe was matched against that of ‘the customer’s’, including business sector
codes and business type
• The ‘the customer’s’ data was then de-duped against The Knowledge Store’s data universe
Data Enhancement
• Any grey area data was manually checked for correct business type and changed accordingly.
• Any demographics selected by ‘the customer’ were then be added from The Knowledge Store data universe
Sector Profiling
• Stores were then profiled
By grouping of store attributes i.e. sells alcohol, has a chiller etc.
• Stores were then ranked
By the overlaying of additional demographics such as proximity data (i.e. near a bank or university or A road)
• Determination of sector profiles giving sector size & sector market value
