%20(1).png)
Introduction
As the world becomes increasingly data-driven, businesses are looking for new ways to gain insights and stay ahead of the competition. Alternative data, which refers to non-traditional data sources such as social media, satellite imagery, and credit card transactions, can provide valuable insights that are not available through traditional data sources. However, analysing alternative data can be challenging because it often lacks the structure and standardisation of traditional data sources. In this blog post, we will discuss five criteria typically used to analyse alternative data for scarcity, granularity, history, structure, and coverage.
Scarcity
Scarcity refers to the extent to which a particular type of data is available. Alternative data sources can be scarce, which means that they may not be available in large quantities or may only be available for specific time periods. To analyse scarcity, businesses must identify the key variables that they are interested in and determine the availability of data for those variables. For example, if a business is interested in analysing social media sentiment about a particular product, they may need to gather data from multiple social media platforms to ensure that they have a sufficient sample size.
Granularity
Granularity refers to the level of detail or precision of a particular type of data. Alternative data sources can vary in granularity, which means that they may provide different levels of detail about a particular variable. To analyse granularity, businesses must determine the appropriate level of detail for their analysis and identify the data sources that provide that level of detail. For example, if a business is interested in analysing consumer spending patterns, they may need to gather data at the transaction level, rather than at the aggregate level.
History
History refers to the availability of data over time. Alternative data sources may provide data for specific time periods or may only provide data for the present. To analyse history, businesses must identify the time periods that are relevant to their analysis and determine the availability of data for those time periods. For example, if a business is interested in analysing the impact of weather on consumer behaviour, they may need to gather data for multiple years to identify patterns and trends.
Structure
Structure refers to the organisation and format of a particular type of data. Alternative data sources can vary in structure, which means that they may require different approaches to analysis. To analyse structure, businesses must identify the appropriate data modelling techniques and tools for their analysis. For example, if a business is interested in analysing satellite imagery to identify changes in land use, they may need to use machine learning algorithms to classify and analyse the imagery.
Coverage
Coverage refers to the extent to which a particular type of data represents the population or phenomenon of interest. Alternative data sources may not provide complete coverage of a particular population or phenomenon, which means that they may not be representative of the entire population. To analyse coverage, businesses must determine the appropriate sampling techniques and assess the representativeness of their data. For example, if a business is interested in analysing credit card transactions to understand consumer spending patterns, they may need to ensure that their data represents a diverse population of consumers.
Conclusion
Alternative data sources can provide valuable insights that are not available through traditional data sources. However, analysing alternative data can be challenging because it often lacks the structure and standardisation of traditional data sources. By analysing alternative data for scarcity, granularity, history, structure, and coverage, businesses can gain insights and stay ahead of the competition.
It can be a challenging task to identify quality alternative data and thats where ESG Analytics comes in when it comes to alternative data in ESG - we process millions of documents every day, in order to find signals that can be used for decision making.
