An information silo is a management system incapable of mutual exchange with other similar systems. Information silo is a derogatory expression that can be used to describe a lack of operational reciprocity. Derived variants are silo thinking, silo vision and silo mentality.
In an Information Silo, the data is not properly shared. It remains within each system or subsystem. So, the data silos can be an obstacle for businesses willing to use data mining to further use their data.
The expression is typically used for introverting systems where information communication is vertical. Critics of silos argue that managers serve as information gatekeepers that make coordination and communication between departments more difficult and make seamless interoperability with external parties impractical. They argue that silos tend to limit productivity in virtually all organizations, that they pose a greater risk of security lapses and privacy breaches, and that they create frustration among consumers who increasingly expect to receive access and complete information sought.
---
Why data silos exist?
The term silo effect can be used in business and organizational contexts to refer to a lack of communication and common goals between departments in an organization. It is the opposite of system thinking in an organization. The silo effect takes its name from the farm’s storage silo, each silo contains only one type of grain.
These three stages are traditionally controlled by different departments i.e. Marketing, IT/Customer Service/Call centres, Customer Loyalty etc. These departments are driven by separate technology budgets and organizational structure. So, a Chief Marketing Officer would order his technology for marketing CRMs and databases and Head of call centre operations/Customer Service will control data stores under his/her control. As a result, customer data from different stages of customer journey gets locked up in siloed data stores across technical and organizational silos within the company.
How to fix data silos
First, technical silos need to be addressed i.e. single unified data store across the company which can house all types of data being generated in an organization. Data Warehouses solve this problem to a certain extent but their fall short of handling unstructured data at scale.
Second, authority and control over data i.e. who gets to do what on data. Typically data producer decides who can be the consumer of data and needs what type of access to data.
Cloud Data Lakes have evolved over a period to address the data silo issues as they can store structured data and unstructured data at any scale. This is because we can dump all data without having to structure it into tables at the time of the write. Data lakes can also address the governance, security and privacy issues associated with data. Data producers can exercise authority and control over their data.
