Expert Insights

Marketing efficacy remains at the forefront of business growth and success. In our increasingly digital era, the degree of such marketing efficacy is inextricably tied to technology. Digital marketing campaigns often prove the most effective in improving targeted marketing and client engagement, especially given a vast array of tools such as email and social media platforms. Despite these resources, companies across industries and specialties often face difficulties in executing targeted marketing, especially when it comes to identifying a responsive demographic. 

Dubotz worked on these difficulties with a Mortgage Servicing firm. Every month, this lender considers a large group of people with mortgages to potentially be the recipients of refinance offers. Subject matter experts would manually create hand-written rules to narrow this large pool of potential recipients to a smaller number of borrowers who would then be sent offers. Of the thousands of offers sent monthly, typically less than 10% of the recipients subsequently initiate a loan application to refinance with the client.  

Numerous businesses even far beyond the scope of finance have experienced a similar narrative as expending excess resources in a struggle to identify the most viable demographic for targeted marketing. With tools such as RPA Automation and Machine Learning, however, businesses can greatly increase efficiency of their marketing processes and increase customer engagement. 

To increase the client’s respondent yield, Dubotz created and trained a Machine Learning model, that identifies borrowers most likely to refinance each month. The Machine Learning model is initiated by the RPA Bot which retrieves the data from the CRM system and runs the logistics of sending out the targeted Marketing Offers based on Output generated by the model. The model was trained over a few months by keeping a human in the loop to fine tune it. The application allows for a more precise targeted marketing approach as developers created a model that assigns a rank to every borrower according to the individual borrower’s probability to respond to an offer in the mail. The model considers the client’s aggregated data on the mortgage borrowers as well as external data such as monthly federal interest rates, ultimately allowing the client to prioritize their marketing targets on this ranking in order to engage with customers who are more likely to refinance.  

Six months after implementing the application, the client sent out 300,000 fewer offers while simultaneously receiving a 10% increase in respondents each month. This increase in respondents and decrease in required manpower and number of printed mail offers epitomizes the potential of automation to revolutionize marketing, especially when used in conjunction with predictive analytics. By analyzing past data to predict future customer behavior and acting upon those predictions with Machine Learning and RPA Automation, businesses can devote their marketing resources to avenues with more guaranteed returns. 

View the Full Case Study this Article was based on, here.