​The question is not "CAN" a borrow pay back a loan, but rather "WILL" they.  We've solved it. 

  • We typically reduce lender defaults by 25-33% (those the lender would normally say "yes" to but should have said "no.")
  • We also dramatically improve decisoning of the turn-downs (those the lender would normally say "no" to but should have said "yes.")  Borrowers that are typically lost forever ---- We can extract +25% of those as success.
             User Story
  • Typically lenders in reasonably high volume scenarios see immediate revenue impact of $2 - $5,000,000 per month or more.  Money that’s just sitting there …through better decisioning.
  • With ONLY a single social media profile, nothing else, we can accurately predict or project a consumers FICO score range almost 80% of the time.  Again: 80% accuracy with just a single social media profile.
  • We can with 71% accuracy, using ONLY a single Facebook profile, predict if a borrower will be a transactor or revolver. That is, will they pay-off the loan early or let it fully amortize (keep in mind that this is not a 50/50 baseline as those types of consumers are not balanced in the population and include consumers who might present as Dormant. For example only 29% of the population are Transactors.  So, yes, 71% accuracy is extraordinary.
             User Story
  • A credit card company in their marketing initiative trying to get a million new accounts (because credit card companies make money from transactors/banks from revolvers) we proved we could provide a 9.1% increase in ONLY getting transactors --- which increased revenue for them 1.8mm per month.  Money that’s just sitting there …through better decisioning.