Decision Engine for Collection

DECISION ENGINE FOR COLLECTION & RECOVERY

CRIF offers an advanced collection management decisioning solution, enabling organizations to manage all decision-making touchpoints of the end-to-end collection process.

From portfolio data enrichment to segmentation through advanced collection scoring models and optimal strategy design, we help our clients improve their performance.

Contact Us

KEY AIMS OF COLLECTION STRATEGY

INCREASE RECOVERY RATE: AI-driven process leading to higher recovery rate
OPTIMIZE RECOVERY COSTS: Reduction of OPEX costs relating to recovered amount, improving process efficiency
ROLL RATE IMPROVEMENT: Reduce the Roll Rate for fewer cases “rolling” into default
FTE EFFICIENCIES: Increased process automation, reduced involvement of collection officers in the process, reduced rework rate
CONSTANT POLICY REVIEW: Constant review of policies, staff allocation, processes, budget execution, etc.
TAILORED APPROACH: Consumers expect a personalized approach but potentially with less human interaction.

DECISION ENGINE USE CASES ACROSS THE COLLECTION PROCESS

Early collection strategy
Score (collection) / Self-cure models
Contact strategies
Collection amount estimation
Management of collection agencies/insourcing
Restructuring & Settlement (late stage)
Recovery allocation strategies (Work/Place/Sell)

COLLECTION DECISION ENGINE CAPABILITIES

ANALYTICS & MACHINE LEARNING

Ability to leverage advanced analytics to obtain granular portfolio segmentation and create clusters with the same “persona” profile. Granular segmentation can guide the timing and type of collection action.

CONFIGURABLE DECISION PROCESS

Ability to make consistent decisions by leveraging both analytics and the control of subject matter experts over decision-making logics thanks to our no code designer.

MONITOR & IMPROVE

Ability to monitor the performance of decision-making processes using a dashboard in order to discover improvement needs, apply any changes and carry out what-if scenario analysis to understand the impacts.

Testing in production using a champion/challenger approach

HOW DOES DECISION ENGINE FOR COLLECTION WORK?

In the initial step, the decision engine is fed with data, which can come from multiple data sources including: internal data or external data sources such as credit bureau, business information, open banking, etc.

The decision engine performs data transformation and the calculation of policy rules and KPIs, further enriching the target data model.

In order to improve performance, once you have a complete dataset, our segmentation and scoring models allow the collection strategy to be identified, directing each segment toward the optimal course of action.

The collection strategy and score are determined at the outset of the process and periodically refreshed in order to adjust the customer assessment based on the customer’s behavior and responsiveness to the actions applied.

The collection score can be evaluated at a product type level; if one customer has more than one delinquent account, a customer level score can also be calculated, either through an additional dedicated scorecard or as the result of aggregation rules applied to the product scores.

CRIF’s decision engine allows collection cases to be segmented from different perspectives, capturing and summarizing all the relevant characteristics and ensuring the comprehensive profiling of all delinquent customers.

After the accounts and debtors are scored and segmented, the decision engine helps users automate the collection decision by calculating a repayment plan with settlement option.

In the following step, the contact strategy is executed, with recommendation of the optimal communication channel, always with a focus on the client/debtor as well as other conditions such as customer responsiveness, contact channel preference or other factors that can flexibly designed or modified by the business user.

CRIF’s decision engine supports customer-centric process management, integrating data coming from multiple data sources, calculating aggregated metrics at a counterparty level and making them available for decision automation as well as manual review.

The solution also includes a Business Intelligence & Reporting module, which enables the monitoring of collection performance.

The Strategy Optimizer identifies any necessary modifications to apply to the strategy in order to continuously improve performance, with the following benefits:

Provides powerful insights and visibility of the collection strategy effectiveness as well as performance of the collection process, including the individual performance of collection employees
Provides the option to integrate collection strategy data with external data (e.g., outcome of recovery actions)
Users can define KPIs, with the option to receive trigger alerts in the case of special events
Continues improvement of collection strategies thanks to test & learn functions such as Champion Challenger or Strategy Simulation (leveraging powerful what-if simulations)

Solving complex challenges with AI optimization

CRIF’s decision engine decision optimization module is an AI-driven solution, which helps solve, through mathematical optimization, complex challenges such as:

“Do we have right number of personnel dealing with collection cases?”
“If I had 10% more capacity, what would the financial impact be?”
“If we reduce our workforce by 5% or 10%, how will this affect our results and which adjusted strategy would maximize the collection goal under these new constraints?”
If you’re halfway through the month and you haven’t hit your targets — “How can we optimize our strategy to refocus?”

Intuitive graphics and easy implementation

The full collection strategy is available to expert and non-IT users at any time through the graphical user interface. CRIF’s decision engine designer is a self-service interface that enables the graphical authoring, design, testing, simulation, deployment, running and administration of decision strategies.

For example, scoring and strategies can be designed (and modified over time) by business users autonomously without IT intervention. Users can independently configure segmentation rules by leveraging a wide range of data pertaining to the delinquent customer profile, to its past due amount or to its current and past repayment performance through an intuitive graphical interface.

Full change management and audit tracking

The decision engine provides users with full audit tracking thanks to audit logs and versioning of all changes. The solution creates automatic documentation and is capable of comparing various process versions and highlighting differences.

A peer review allows the evaluation of the decision process configuration by one or more people (4 eyes check, 6 eyes check…etc.) as a precondition to deployment. Each reviewer must approve the work done to enable deployment of the decision process.

The figure below shows an indicative example of how the decision engine can support a standard collection process from end to end (not in relation to the application of the CRIF decision engine in the collection process).

NEEDS

Constant pressure to increase the recovery rate
Further optimization of process-related costs (OPEX and FTE efficiencies)
Tailored approach and handling of each case, reducing human interaction if possible
Data-driven decisioning with the option to apply advanced AI optimizations.
Constant policy review and execution monitoring

The CRIF decision engine can support collection and recovery at multiple stages of the process, including:

Early collection strategies
Collection scores, including self-cure models
Contact strategy optimization
Management of various collection channels including collection agencies
Allocation strategies for late-stage recovery

BENEFITS

For financial companies

- Lower NPL ratio
- Higher recovery rates
- Lower risk costs
- Lower operating costs
- Shorter time to recovery
- Lower provisions

For consumers / debtors

- Personalized approach
- Reduced collection fees
- Early collection reminders

HOW WE HELPED OUR CLIENTS UNLOCK THEIR POTENTIAL

Discover how some of the world's most important companies improved their performance and defined a brand-new digital path, both for them and for their clients.

for a Tier-1 Bank

UNIFIED COLLECTION SYSTEM INTEGRATED WITH EXTERNAL SYSTEMS

The client needed to create a unified collection system that could integrate with external systems

Unified Collection System
End-to-end management of origination and monitoring processes

for an important Telco Company in Middle East

ENHANCING ORIGINATION AND COLLECTION MANAGEMENT

The client needed to improve the risk evaluation while reducing rollover of customers into collection and optimize early stages collection strategies

Credit Scorecards
Strategies fo credit limit management and early stage collection

+30 %

Increase in on-time payments
22 %

Decrease in 30-day past due accounts and balance roll rates
+6 %

Increase in collection revenue (of 45-day collection portfolio)

for an important Telco Company

IMPLEMENT A UNIQUE COLLECTION SYSTEM

The client needed to implement a unique collection system while reducing time to decision and decreasing operational overtime

Decision Engine to speed up time to market
Decision Engine to support Collection process

GET IN TOUCH WITH THE SALES TEAM

What does the next digital journey of your business look like? Let's find out together.