Two words have the power to make some financial institutions (FIs) cringe: unstructured data. By definition, unstructured data refers to any data that isn’t housed in a database or other data structure. This can be textual or non-textual. Things like social media feeds, slide decks, spreadsheets and videos are all considered unstructured data.
The reluctance to utilize unstructured data may come from the perception that it is low-quality, inconsistent or perhaps dated. With 80 percent of business data taking this form, it seems unwise to simply discount it altogether. In fact, unstructured data can prove valuable when FIs are making business decisions.
To embrace the power unstructured data holds, FIs can start by building a strong data provenance model. This model can log the different data sources and score them in terms of their reliability. With this method, FIs create transparency in their data gathering and provide context for determining when to use unstructured data.
Unstructured data opens up a number of possibilities for FIs, including:
- Consumer service – Exploring data gleaned from social media sites and other platforms consumers frequent can reveal pain points and sentiments. This information can then be utilized to better serve consumers.
- Market research – Unstructured data pulled from consumer-generated content can help paint a picture of what consumer groups want or need from their FIs.
- Risk management – Targeted keyword searches can reveal risk and troubling patterns in unstructured data sets. FIs can then use these results to adjust their security and fraud prevention strategies.
Ultimately, FIs should have a clear vision of what questions they are looking to answer with data analytics. The data, whether structured or not, should be chosen carefully and with purpose based on the FIs’ end goals.