A large US credit card issuer was not achieving its forecasted roll and delinquency rates. Underperforming vintages were restricting the company's ability to grow new accounts, and internal teams had exhausted their standard playbook of interventions without meaningful improvement. The collections operation was running on legacy strategies that had not been recalibrated in years, and recovery processes were disconnected from upstream account management decisions.
PAG was engaged to re-engineer the collections and recovery operation from the ground up. The engagement began with a Bucket 1 import strategy redesign, focusing on early-stage intervention before accounts rolled into deeper delinquency. Payment propensity models were developed to prioritize outreach to customers most likely to cure. Channel optimization shifted contact attempts toward the methods each customer segment was most responsive to. A vendor placement model was built to allocate third-party volume based on performance data rather than static contracts. The client's payment website was redesigned to reduce friction and increase self-service resolution.
The results were substantial across every metric. Website traffic increased by 50%, and online payment completion rates jumped from 30% to 60%, dramatically reducing the volume of calls needed to collect. Operating costs dropped by 35%. Payment suite optimization delivered a 20% improvement. Recovery rates improved by 11%. Customer complaints fell by 23%, and charge-offs were reduced by 7%.
The success of the collections transformation created confidence for a broader engagement. The client expanded PAG's role to encompass full portfolio management, including underwriting strategy, fraud operations, account management, and digital transformation. Each workstream was integrated through GOBLIN, ensuring that decisions made in one area were informed by data from every other area.
This case illustrates a pattern we see frequently: a focused, results-driven engagement in one operational area creates the trust and data infrastructure needed to drive enterprise-wide improvement. The key is starting with a measurable problem, delivering measurable results, and building from there.




