Financial services firms including banks, insurance providers and peer-to-peer (P2P) lenders are increasingly relying on predictive models built using artificial intelligence (AI) and machine learning (ML) to improve customer acquisition and retention, refine risk modeling and enhance loan application processing, among other areas.
For predictive models to be more accurate, financial services companies should supplement their internal data (first party data) with external data and features from third-party providers, and embrace a practice referred to as data enrichment, says consumer intelligence solutions provider Mobilewalla.
In a newly released guide, Mobilewalla delves into three data and feature enrichment strategies that financial services companies can leverage to grow their business and avoid risk, highlighting key areas where financial firms can use data from third-party providers to improve processes and better engage with customers.
The first strategy highlighted in the paper involves the use of householding and relationship networks to optimize predictive modeling, allowing these models to understand patterns in customer behavior, creditworthiness of prospects, as well as their qualifications for specific financial services products.
Householding refers to a set of consumers residing in a distinct (same) household. Financial services companies can leverage demographic and behavioural insights of these consumer cohorts and cater to their different needs.
Meanwhile, Relationship Networks define the ‘collective company’ of consumers which includes anonymised cohorts of users in the same vicinity that represent their social relationships and work engagements. It delivers insights that are broadly predictive of different consumer behaviours important to financial companies such as the propensity of defaulting on a loan or the likelihood of being a high-value customer.
But according to Mobilewalla, most organizations don’t typically have access to this data, noting that partnering with a third-party data provider is often necessary to fill in the missing features.
The second data and feature enrichment strategy highlighted in the paper is the use of cross-device mapping to match web engagement with audience data to identify anonymous website visitors and individuals being served ad impressions.
By combining third-party data enrichment with IP address mapping, financial services companies can attach distinct identities to anonymous prospects. This identity information can then be used to target audiences for marketing campaigns, or drive customers’ towards the solutions that would best fit their needs.
Finally, the third strategy revolves around the use of data enrichment to better understand customers and improve services. In this scenario, access to data from external sources provides financial services companies with the opportunity to deepen their understanding of customers, improve customer experience, and reach new audiences.
The last decade has witnessed a steady and significant increase in the number of Internet users and booming mobile usage. New customer behavior and new channels have resulted in an influx of new data which financial services companies are now leveraging to improve processes and better customer experience.
According to Insider Intelligence’s AI in Banking report, released in January 2021, 80% of banks are highly aware of the potential benefits of AI. In fact, 75% of respondents at banks with over US$100 billion in assets said they were currently implementing AI strategies. By 2023, banks are projected to save approximately US$447 billion by developing and implementing AI applications.
Featured image: Photo by Scott Graham on Unsplash
