Skip to main content

How does a financial services firm improve sales targeting to predict its clients' desires to invest? Machine learning was the answer for PFC. Find out why.


Long-time client Primary Financial Company (PFC) manages an investment program for institutional investors to invest substantial funds in federally insured CDs. The company:

  • Monitors, tracks, collects, and disburses principal and interest on nearly 40,000 CDs
  • Manages over $7 billion in assets
  • Supports relationships with 5,000 financial institutions and institutional investors

PFC wanted to improve sales targeting to predict CD issuers’ funding needs and institutions’ desires to invest. It partnered with Fusion on a pivotal initiative to explore how advanced analytics/machine learning could drive data-driven, predictive outcomes.



Organizations with the expertise to leverage machine learning will significantly widen the gap between themselves and competitors who aren’t able to move beyond traditional analytics tools. Fusion’s data science and application development teams work together to unlock business insights. We allow our clients’ data to tell the story without introducing bias to the underlying predictions. This hybrid approach allows us to apply agility in the process to rapidly operationalize insights into tools and platforms you use every day.

PFC understood the value proposition of machine learning and partnered with Fusion to explore the following machine learning models:

  • Identify best issuers for sales solicitation, including former, current, and prospective issuers
  • Provide rate guidance to investors and rate/term guidance for CD issuers
  • Target investors by likelihood of close

The process included four main steps: data acquisition, transformation, model development, and predictive analytics.

All relevant private and public data sources were identified and acquired to gain more information on current and prospective customers. PFC and Fusion then collaborated to determine meaningful and available factors.

Sources identified for PFC data acquisition: Private include insurance history, trade history, and third-party analytical data. Public include federal reserve bank (FRB), Federal Deposit Insurance Corporation (FDIC), Office of Thrift Supervision (OTS), National Credit Union Administration (NCUA), and Credit Union National Association (CUNA). These data points were identified as relevant and acquired by PFC to gain insights into current and prospective customers. They were integrated into the existing investment-platform issuer data to offer a higher distribution of information to pattern-match for the prospect.

Next, the data was transformed so these factors would be consistent and accurate. With a solid foundation, Fusion developed machine learning models that would learn and identify patterns, then recognize those patterns when seen again to apply lessons to predict outcomes.

The first phase was highly successful, producing the following benefits, with more to come in the next phase.

  • Equipped PFC sales team with a qualified list of issuers to target based on previous profiles of customers. This enables efficient use of PFC’s finite sales and marketing budget
  • Provided a deeper understanding of where CD issuers need to price their instruments and the rate at which investors are likely to purchase. Understanding this spread allows PFC to potentially achieve larger-scale trading business
  • Provided a simple way for PFC’s co-brokers to market the right product at the right time to the portfolio of investors


While this initiative is entering a second phase, already PFC and our Fusion team will be ablae to ascertain with over 80% accuracy and 70% precision the likelihood of a particular investor to buy a given investment.

Based on the analytical profiles of financial institutions, PFC identified prospects to target, even with limited marketing and sales resources. The company also provided its trading desk with a list of potential investors and their expected likelihood to buy the specific product.

A key factor that led to success was PFC’s clear understanding of the target accuracy of the predictive models. By defining the required utility of these models, PFC could realize business value without the churn of endless model tuning. PFC’s deep collaboration with our Fusion team was another vital component to the overall success. Their team could quickly understand errant data and numbers and drive the value of Fusion’s models higher.

Identified prospects to target
Identified prospects to target
Created recommendations for effective market making
Created recommendations for effective market making
Identified potential investors and likelihood to buy products
Identified potential investors and likelihood to buy products

Ready to talk?

Let us know how we can help you out, and one of our experts will be in touch right away.