Using ML to Optimize Letter Campaigns
Machine Learning (ML) still has an aura of fiction surrounding its use. While ML has been used in the collections industry for most of this century, utilizing its unique potential to comb, sift, and manage large volumes of accounts has been mostly underutilized.
One of the most efficient applications of the power of ML is to build a model designed specifically for optimized letter campaigns—an often underestimated yet vital aspect of any collections organization.
Developing a New ML Model
Letter campaigns are an important part of the collections process. Many agencies simply need to give encouragement to delinquent account holders and while letter campaigns are required by law, they are also a great way to show friendly customer service while building your brand in the eyes of the consumer.
However, letter campaigns are often a misused concept. Catch-all campaigns beyond required notifications that send letters to every account holder, or just those you have addresses for, are likely to fail.
Utilizing ML, Equabli’s platform can build campaigns that target specific accounts and track years and years of data to understand, organize, and plan campaigns that look at the most likely-to-respond consumers. At the heart of every new campaign, however, isn’t the ML model we end up with; it is the data-gathering session.
Getting a Handle on the Data
Equabli first conducts a multi-week discovery session gathering all the relevant information about the current operational strategy and data management process to build the relevant data pipeline and collect all the key data points needed to build an account-level value-predicting ML model.
This step is vital. Without a deep understanding of the accounts, the company as a whole, or even the approach necessary for getting a letter campaign off the ground, both the company and the technology firm would be lost in data that seemed irrelevant.
Once a data pipeline is built, all the relevant desensitized account information can be loaded into a secure cloud-based data infrastructure and processed through its cutting-edge ML model development framework. This is where Equabli comes in handy for collection agencies, law firms, debt buyers, or lenders that have direct, written communications with consumers.
Developing a model from the ground up is already tricky. Developing a model without 100 years of combined experience is almost impossible. Equabli efficiently and effectively builds advanced fast-learning ML models using the latest innovative tools in the field of Machine Learning and Data Science.
Harnessing the Results
Like any letter campaign, the campaign is only as good as the company’s ability to utilize the information. After the letter campaign is designed, modeled, and running, Equabli works closely with the client to ensure that its ML models are both accurately reaching consumers and growing the client’s revenue.
After building a letter campaign designed to match a unique use case, it’s important to properly assess the customer segmentation needed and extract all the key features that help predict the willingness and capability of each customer under the letter collection strategy. The performance of most ML letter campaigns is highly significant, with the potential to provide an incremental lift of ~30% in annual gross collections in any given letter channel.
In back-testing, one of Equabli’s ML letter campaigns led to some powerful results. ~90-95% of gross collections were achieved with the top 60% of accounts identified to mail by the ML models. Live results post-model implementation have the potential to generate ~30% incremental gross collections.
Learn More Online
Equabli’s unique approach and deep collaboration with its clients post-development ensure these solutions are effectively implemented, tested, and properly used in the client’s collection strategy to realize and maximize the value of the model. This “last mile” operational transformation support is unique to clients given Equabli’s core abilities in the industry and comes from a core value of “customer obsession.” This level of support for Equabli’s clients is possible thanks to Equabli’s domain experience and understanding of the pain points and pitfalls of evolving into a data-driven decision culture.
Visit Equabli’s website to learn more about how their faster implementation, better decisioning, and sharper tools have positioned the organization as one of the central, integrated, and pivotal hubs of the debt recovery industry.
About Equabli
At Equabli, we’re experienced and ambitious fintech entrepreneurs tackling the delinquency and default stages of the consumer credit lifecycle. Equabli is the hub for debt management. Our one-stop debt management platform uses intelligent tools and technologies to simplify and streamline the debt recovery process and optimize outcomes for both lenders and borrowers. Equabli makes recoveries easier, gives valuable insights, and supports a healthier recoveries ecosystem for all.
