Customer Analytics
Retailers have always wanted to understand who is shopping from them. Today there are companies which help retail majors with customer information. However most of them are unable to make the best use of this. Primarily, most databases fall short in enabling interactive analytics on such high volumes of data.
At one of our customer sites, they wanted to enable real time querying on over 250 TB of customer
data. It used to take them days to create pre-aggregations, run queries and build dashboards and even then it didn’t allow for granular analysis. Moreover because it took such a long time to process and analyze that the dashboards were created monthly while the data was being updated live, making most of the analysis stale and redundant.
Key Business Questions to be answered
  • What is the penetration of any merchandize in terms of customer reach and contribution to the Total Box sales?
  • What is the weekly trend for the category and customer segment for the last week, last month, last quarter and last year or any other ad-hoc time range as required?
  • How is the selected merchandize performing in comparison with the the category, division or the SBU the item belongs to?
  • What is the penetration and reach in a designated target customer market by geography, region or store?

typing

How we enabled Customer Analytics at scale

img

  • Mashup customer data with point of sale data to create a unified view of sales for customer analytics
  • Power custom UI tailored to enable business users perform ad-hoc slice and dice of queries in seconds
  • Leverage existing on-premise or cloud Hadoop data lakes to power customer analytics at scale
  • Create a highly reliable, secure and fault tolerant system while minimizing data movement across clusters
  • Shorten development cycle of such massive analytics projects to less than 4 weeks by leveraging NitroDB
Customer Analytics
Retailers have always wanted to understand who is shopping from them. Today there are companies which help retail majors with customer information. However most of them are unable to make the best use of this. Primarily, most databases fall short in enabling interactive analytics on such high volumes of data.
At one of our customer sites, they wanted to enable real time querying on over 250 TB of customer
data. It used to take them days to create pre-aggregations, run queries and build dashboards and even then it didn’t allow for granular analysis. Moreover because it took such a long time to process and analyze that the dashboards were created monthly while the data was being updated live, making most of the analysis stale and redundant.
Key Business Questions to be answered
  • What is the penetration of any merchandize in terms of customer reach and contribution to the Total Box sales.
  • What is the weekly trend for the category and customer segment for the last week, last month, last quarter and last year or any other ad-hoc time range as required?
  • How is the selected merchandize performing in comparison with the the category, division or the SBU the item belongs to?
  • What is the penetration and reach in a designated target customer market by geography, region or store?
How we enabled Customer Analytics at scale
  • Mashup customer data with point of sale data to create a unified view of sales for customer analytics
  • Power custom UI tailored to enable business users perform ad-hoc slice and dice of queries in seconds
  • Leverage existing on-premise or cloud Hadoop data lakes to power customer analytics at scale
  • Create a highly reliable, secure and fault tolerant system while minimizing data movement across clusters
  • Shorten development cycle of such massive analytics projects to less than 4 weeks by leveraging NitroDB