ABOUT THIS DOCUMENT
Forward-looking statements
Product roadmaps and similar marketing materials should be considered forward looking and subject to future change at Fair Isaac’s discretion. Future functionality, features or enhancements as shown are Fair Isaac’s current projections of the product direction but are not specific commitments or obligations.
This is Revision 3 of this document and replaces all previous versions.
Fair Isaac, Capstone, FICO, PlacementsPlus and ScoreNet are registered trademarks of Fair Isaac Corporation in the United States and may be trademarks or registered trademarks of Fair Isaac Corporation in other countries. Blaze Advisor, Debt Manager, Falcon, Recovery Management System, RMS, Small Business Scoring Service, TRIAD and Vectus are trademarks of Fair Isaac Corporation in the United States and may be trademarks or registered trademarks of Fair Isaac Corporation in other countries. Other product and company names herein may be trademarks of their respective owners.
CONTENTS
Message from the CEO . . . ii
Defining EDM . . . 1
The Power of Connected Decisions . . . 6
Scenario 1: Existing Customer Wants a New Credit Line . . . 7
Scenario 2: Valuable Customers Enter Collections . . . 8
Scenario 3: Customer Using New Card Gets Flagged for Fraud . . 10
The EDM Architecture . . . 14
Roadmaps . . . 18
EDM Application Roadmaps: Financial Services. . . 18
Originations . . . 20
Customer Management . . . 21
Collections and Recovery . . . 22
Fraud . . . 23
EDM Technology Roadmaps . . . 25
Analytics Roadmap. . . 26
Migration . . . 27
Message from
the CEO
Fair Isaac is now and always has been focused on decisions. From its earliest days more than 50 years ago to the present, the people of Fair Isaac have been working with clients to make their businesses more profitable by improving the way they make decisions. By developing innovative analytics and decision technology, Fair Isaac has changed the way the retail banking industry works and made thousands of clients—in multiple industries and geographies—more successful.
Today we have a phrase to describe the use of data, analytics, business rules and adaptive control to manage and improve operational decisions:
Enterprise Decision Management, or EDM. While the phrase may be new, make no mistake: EDM is not dependent on technology that is still on the drawing board. It is a market-tested approach that works.
We believe there is great potential for continued value growth from applying EDM in retail banking. Linking decisions across the lifecycle, more precise customer targeting and more complete customer information all create the possibility for growth.
We also believe that the time is right for many other industries to see the same benefits. The technology on which EDM relies is more mature and more scalable than ever. Companies have collected, organized and begun to understand their customer data, though most still fail to derive full business value from it. The analytic techniques needed to understand this data are established and continually improving. Service Oriented Architectures (SOA) and Business Process Management make the integration of powerful decision services both straightforward and necessary. The time is right for EDM to become a mainstream business practice.
Fair Isaac is committed to the development of a new EDM architecture that delivers greater flexibility in how you design and execute decisions. This architecture can give your systems higher analytic performance to maximize the value of your data, while enabling you to make better marketing, fraud and risk management decisions throughout your operations and across the customer lifecycle.
This document shows how we intend to evolve and extend our current software, analytic and product architecture to help you become truly decision-centric. We welcome the opportunity to provide more detail on our products and EDM technology to you, to discuss your particular business challenges, and to explore how your organization can make smarter decisions.
Challenge—
Creating Value in Each Transaction
Every business transaction, especially thoseinvolving your customers, creates or destroys value. While companies spend a great deal of time and money focusing on low-volume, high-value transactions, they typically spend much less time working out how to make high-volume transactions that have lower value per transaction. Yet the total value of these “low-value” transactions can be very high thanks to their volume. Ensuring you create value in these high-volume, rapid-turnaround, operational transactions is a critical challenge for the 21st century.
Figure 1:
Some of the trends making it harder to add value to operational transactions
There are many reasons why adding value in high-volume operational transactions is difficult and getting more difficult all the time.
»Pace of change. The pace of change continues to
accelerate. Consumers are more fickle in the products they seek and more willing to move to your competitors. Regulators continue to impose more rules, adding new compliance requirements faster than ever. You have more competitors that are easier for your customers to find, and non-traditional entrants using the internet are just “three clicks” away.
»Need for speed. Your customers and business
partners are demanding that you do business in real time. More of your transactions start and must be completed on-demand; coordinating and balancing the demands of these transactions is harder. Customers and partners expect straight-through processing, forcing you to eliminate exception processing.
»Uncertainty and complexity. The competitive reality
is that you must handle an increasing number of offers and products. You must define your customer segments more finely and deliver consistently across many channels. You have more products, aimed at smaller segments, constrained by more regulations, which you must deliver through more channels.
»Risk. Fraud, privacy and security are of great
concern to customers. You must balance reputation and business risk, privacy and data security, fraud and growth more carefully than ever before.
»Loyalty and attrition. Established industries, such
as credit cards, are seeing growing attrition while all industries see that customer loyalty is both more valuable and harder to come by.
Days Timeliness of Decisions Point of ContactReal-Time/ Local/
Simple Level of Objectives Corporate Objectives/Trade-Offs Static/
Simple Regulatory Constraints Dynamic/Complex Every 3-5
Years Changes to Decision Strategy AdjustmentsFrequent Well-Defined Decision-Making Boundaries Geography/DeptsCross Channels/ Low/
Manual Operational Volume AutomatedHigh/
Workforce
In addition, you must continue to focus on your profitability. Having eliminated inefficiency, you must now focus on improving your top line, the effectiveness of your operations.
Key requirements for a solution to these challenges include:
Greater flexibility
Your operational transactions are handled by a wide range of systems on a wide range of technologies, both internal and external. A solution must integrate flexibly with these systems and support constant change and improvement. As you find new
opportunities, you must be able to react immediately.
Higher analytic performance
As more of what customers do leaves a digital trail, you have more data available both internally and externally. You must deliver higher performance by turning this data into analytic insight more rapidly and effectively.
Better fraud and risk management
Fraud and risk cut across all your transactions. You must be able to identify and mitigate all forms of risk and fraud in alignment with your overall business strategy.
Customer lifecycle management
The best way to treat customers is to consider the complete customer lifecycle. Success means moving from operational silos focused on products or discrete decision areas—such as marketing or fraud—to coordinated, customer-centric decisions that advance customer relationships, additional sales and increased profitability.
Most importantly, you cannot afford to wait—you need to act now. You must ensure that the decisions you take in your operational transactions create value and reliably build your success.
Solution—
Enterprise Decision Management
Enterprise Decision Management is an approach that automates, improves and connects decisions to enhance business performance.
Enterprise Decision Management delivers a
measurable business impact by increasing customer profitability and loyalty, reducing fraud and credit risk, and lowering the costs of making decisions. It automates decisions for speed and consistency, constantly improves the targeting, relevance and results of decisions, and connects decisions across customer views, functions, and channels for maximum value. It is a closed-loop approach that systematically improves decisions so you can inject learning directly into processes, and use insights from point solutions to make better decisions across your enterprise.
Enterprise Decision Management, or EDM, gives you the control you need over decisions that are growing more complex all the time.
Enterprise Decision Management is a proven approach within critical business areas such as marketing, origination, customer management, collections and recovery and fraud. Fair Isaac clients have led the way in showing the kinds of benefits to be gained by focusing on automating and improving customer decisions in these areas, delivering results like those summarized in Figure 2.
Figure 2:
Some example benefits from adopting EDM in specific areas
Marketing Originations ManagementCustomer and RecoveryCollections Fraud
30%
15%
–25%15%30%
50%
Profit
per account Customersbooked at same risk level
Decrease
in bad debt in collectorIncrease productivity
Decrease in fraud
While every decision has its own measure of success—click-through rates, debt ratios, false positives and so on—there is a “universal” metric for measuring the quality of decisions, known as the Decision Yield (Figure 3). This metric considers five dimensions of decision effectiveness—Precision, Consistency, Agility, Speed and Cost.
Figure 3:
The Decision Yield
To illustrate how these dimensions of a decision can contribute, consider:
Precision
The degree of targeting in a decision or the number of risk tiers used makes a material difference to the profitability of an origination decision, for instance.
Consistency
Consistency of a decision across channels can improve customer satisfaction, while consistency across lines of business can prevent cannibalization of one line by another.
Agility
The ability to adapt dynamically to changing conditions quickly and cheaply can reduce IT costs for system maintenance and, more significantly, reduce losses incurred by failing to respond to a competitor’s offer to your customers.
Speed
The speed with which you can make a decision can determine whether customers look into competitive products, and can improve satisfaction by reducing wait times.
Cost
Spending less money by using less staff time to process normal applications or conducting
Building success with EDM
While businesses adopt different technologies and approaches to EDM in different sequences, many see the benefits arising from three different stages.
Automate
Many companies focus first on the straightforward automation of high-volume operational decisions. Using business rules management software,
companies take control of their operational decision-making. They reduce the time needed to make changes and introduce offers and products, reduce the cost of their decisions and move to taking those decisions in real-time.
Improve
From automation, many companies focus on improving the quality of their decisions and the results from them. Using predictive analytics,
decision analytics (optimization) and adaptive control for closed-loop testing, companies can segment their customer base and target decisions using more precise assessments of risk and opportunity. This leads to improvements such as increases in customer profitability, reduced fraud losses, and higher cross-sell penetration.
Connect
The most advanced stage of EDM is connecting decisions across channels, systems and stages in the customer lifecycle. Using reusable and open
components, businesses can connect their systems; similarly, advanced analytics such as decision models can explicitly balance the trade-offs between different decisions. More sophisticated control over decisions delivers stronger alignment of execution with business strategy and ensures channel and product-line integration.
AUTOMATE High-Volume Operational Decisions
IMPROVE Quality and Results from Decisions
CONNECT
Across Channels, Systems and Customer Lifecycle TECHNOL OGY IMP A C
T • Incremental increases in customer profitability
• Coordinated decisions • Greater alignment of execution with business strategy • Business rules management software • Predictive analytics
• Decision models (optimization) • Adaptive control (closed-loop testing)
• Reusable and open components • Integrated analytics
• Reduce time needed to make changes and introduce offers/products • Reduce cost of decisions • Real-time decisions
• More precise assessments of risk and opportunity
• More profit per customer, lower credit/fraud losses • Systematic improvements in results
Figure 4:
Each Fair Isaac client focuses on these stages differently, depending on the kinds of decisions they need to automate and their experience in decision automation. What’s important to note is that most businesses are a long way from maximizing the full potential of their decisions across all three stages.
The Power of Connected Decisions
One of the critical business drivers for Enterprise Decision Management is the need to make connected decisions—decisions where data and analytics from a decision at one point in the lifecycle affect other decisions later in the lifecycle. Even companies that have enjoyed success through implementation of Fair Isaac’s applications and technology to automate and improve decisions stand to gain by building stronger connections between their decisions, and adopting a more customer-centric approach. This is where the real competitive edge lies for companies in industries that are at the more sophisticated end of the EDM spectrum.To show why this is important, we will consider three scenarios in financial services.
Scenario 1:
Existing Customer Wants a New Credit Line
ChallengeMaking the right offer, at the point of origination, requires a balance of the lender’s objectives and customer’s needs. These factors include: risk, utilization, share of wallet, product competition and exposure. This is best accomplished when looking at the customer holdings across all the lender’s products and outside accounts.
Situation
»Offers for financial products often account for risk; for example, pre-approved credit cards. However, when customers respond to these offers, the originations process may be missing an opportunity.
»By making adjustments to the original offer, the company can better match the customer’s and the lender’s needs. These adjustments take the form of an alternative product, an additional product (cross-sell), or an enhanced product of the same type (up-sell).
»Knowing precisely how to make these adjustments
requires having information that is not typically available during the originations process.
Disconnected Decision
»The offers are based primarily on the revenue potential and risk of the customer. While originations can present alternative offers at the time of
application, they do not have all the information about the customer’s holdings and history.
»It is also difficult for the originator to determine the product that best fits the customer and the lender’s profit or reserve requirements.
Connected Decision
»The originations system has access to important information about the customer across all of their accounts that reside in the customer management system. This enables the lender to incorporate propensity to use, debt capacity and profitability in any adjusted offer.
»Cross-sell approval is significantly accelerated, as the eligibility for products is pre-determined in the customer management system, and made available to the marketing and originations systems through a common data format between the systems.
»The ability for applications to share data and outputs readily means that cross-sell strategies can be quickly adjusted to reflect changing take-up/offer ratios, utilization/take-up ratios and performance/exposure ratios.
»Proper understanding of utilization/take-up ratios also has a significant positive impact on contingent liability. Why offer high limits if they will be taken up but not used, when a product with a high propensity for utilization can be offered instead?
»Overall exposure risk is also better controlled, through use of a “whole customer” view in establishing the offers to be made.
»It also streamlines the process, allowing the originator to spend more time selling.
Business Benefits
»Improved offer management that reflects risk and reward across all the customer holdings
»Faster approvals
»Increased cross-sells, up-sells
»Reduced risk exposure
»Less conflict between competing product offers
»Improved effectiveness of marketing campaigns
Scenario 2:
Valuable Customers Enter Collections
ChallengeMany segments of a portfolio require specialized treatments when entering collections. Customers in these categories include those that typically self-cure, or those that still require contact for payment but hold or purchase products that generate superior profits. Treating these kinds of customers in isolation inside the collections system can not only increase collections costs unnecessarily, it can also increase attrition among those that generate above-average fees, interest income and cross-sell revenues.
Situation
»Customers are passed to the collections systems when delinquent. The policies governing the treatment of these customers while in collections are shared across the customer management and collections operational areas.
»The customer management department
determines an approach to treatment that considers value both during and after the collections episode.
»Goals used to measure success of the collection department may, however, consider value related to the collections episode only—typically recovery rates.
Disconnected Decision
»The collections system isn’t aware of a customer’s true value to the bank. So, the collections
department may load all customer accounts into the automated dialer for a call, or perhaps send everyone a letter requesting payment.
»The self-cure customers will, of course, still pay. However, the cost of a call or letter will have been incurred with no added value to the outcome of the collections process.
»At worst, due to the unnecessary contact, the customer may pay down their balance and cease to use their account, denying the lender revenue that would otherwise have been gained.
Connected Decision
»A connected customer management and
collections system ensures that collection treatment is appropriate given the overall value and risk each customer represents.
»Feedback from the collections system to the customer management system provides visibility into all actions that were taken and how those actions impacted subsequent customer
performance and long-term profitability. With this feedback loop in place, collections strategy and operational results can be continuously adjusted to balance cost of collections, amount collected and customer retention.
»A connected customer management and
collections system also provides feedback that shows the impact of individual and cumulative actions on payment performance, whether from the customer management system, or the skills of the individual collectors.
»Collections strategies can then be continuously adjusted to take advantage of the most effective actions, applying the expertise of the company’s best collectors to all cases.
Business Benefits
»Improve alignment of action with strategic intent, minimizing operational negation
»Quantify costs relative to returns, allowing faster creation of the most cost-effective strategies
»Improve understanding of relationship-damaging actions and adjust collections approach, to reduce attrition without reducing collections rates
»Better identify effective operational actions and extend their reach to more collections cases
Scenario 3:
Customer Using New Card Gets Flagged for Fraud
ChallengeActivation and usage incentive strategies for
dormant or never-active account customers can have their value negated, if the behavior being
encouraged is subsequently blocked as a result of the fraud detection decision not being aware of the activation or usage incentive strategy.
Situation
»Some customers who have opened an account as
part of a new-account recruitment program have not used the card after a number of months.
» The customer management group runs a new
account activation program designed to encourage the customer to use their card, via a mailing that offers a preferential interest rate for the first 6 months of use.
»A customer responding to the campaign attempts
to make a purchase for a high-value item, hoping to take advantage of the preferential interest rate.
Disconnected Decision
»The fraud detection system sees this transaction— for a large amount, on an account with no previous activity—and flags the transaction as potentially fraudulent, due to an “out-of-profile” event.
»The customer suffers the embarrassment of being
»The bank loses not only the revenue that would have been gained from the card product, but also all potential profit from the broader relationship.
Connected Decision
»The fraud detection system references the customer management rules used in establishing the reactivation program, identifies these
customers as ‘selectees’ and sets a new rule allowing the transaction to take place.
»A case is created after the transaction to call the customer, thanking them for their business and asking them if they require any further help or information about their card.
»Verification that the transaction was not fraudulent is achieved, the customer is appreciative of attentive service, and the future profitability of the relationship is protected.
Business Benefits
»Improve alignment of decision strategies, minimizing negation of strategy value
»Reduce false positive rates to minimize attrition
»Increase revenue
Key Elements of an EDM Architecture
Before describing Fair Isaac’s EDM direction, it is worth highlighting some of the general characteristics required for an effective EDM architecture.Decision Services
»Decision services allow you to take control of operational decisions independently from the process and systems that deliver your products and manage your business.
Key driver: The need to manage the value created by each operational transaction.
Business Rules Management Software
»Business rules management software allows
you to automate policies, rules and regulations to create an agile enterprise that makes consistent decisions.
Key driver: The need to follow increasingly complex business policies, operating rules and regulations.
Predictive Analytics
»Predictive analytics predict behavior to guide customer treatment to make precise, targeted decisions.
Key driver: The increasing complexity of decisions and the need to manage uncertainty and risk.
Adaptive Control
»Adaptive control lets you test decision strategies and learn what works and what does not, allowing you to systematically advance and improve results. Key driver: Very dynamic, highly competitive markets.
Decision Models and Optimization
»Decision models allow you to optimize results by explicitly balancing trade-offs such as risk and reward.
Key driver: The need to make decisions with many trade-offs and constraints.
Data
»Data both to build analytic models and to make operational decisions in real time must be available and shared throughout the architecture. This includes internal data, data from external data providers and transactional data.
Key driver: The need to make better use of real-time analytics based on both historical and transactional data is important.
Fair Isaac and
Enterprise Decision Management
Fair Isaac is committed to delivering the premier solutions for Enterprise Decision Management, as well as best-in-class technology that can be used to improve any client’s decision environment. To this end, we are building our offerings on an EDM architecture that will ensure interoperability across systems, and that will enable connected decisions across the customer lifecycle.Decisions can only be made optimally if you can effectively combine business rules (regulations, policies, best practices, experience) and insight from data (descriptive, predictive and decision analytics) into high-performance decision services. Decision services are different from other services or applications in that they change all the time—no amount of investment in automating a decision will ensure that it remains optimal for long.
The decisions that you need to automate with EDM can be very visible, such as the need to check a transaction for fraud, or much less so, such as the need to customize the options on an Interactive Voice Response system to deliver great customer service. The Fair Isaac EDM architecture ensures that you can add value in every one of these decisions.
The diagram in Figure 5 illustrates some of the major concepts of Fair Isaac’s EDM architecture. These include:
Decision service. This is a service that applies rules
and models to make decisions. The decision service can be invoked anywhere a decision is made. That
may be when someone applies for a credit card or new service, when a customer requests additional service, when a transaction occurs or when an account goes into collections.
EDM repository. This is a central
repository that, like the decision service, can be accessed for any decision across the lifecycle. It contains the organization’s full complement of predictive models, rules and strategies.
Fair Isaac Applications
Fair Isaac is well-known for its series of decision applications that automate and improve decisions at critical points in the customer lifecycle—marketing, origination and management for example—as well as for systems that manage fraud at every stage Business Intelligence Decision Service Analytic Development Rules Management EDM Repository ERP CRM Core Banking Billing Supply Chain E-Business OPERATIONAL SYSTEMS DECISION APPLICATIONSCustomer Behavior and Strategy Performance
Process Data, Rules and Analytics to Determine Best Decision
Decision Request Decision Data Warehouse Marketing Originations Customer Management Collection and Recovery Fraud Development and Management Transaction Execution EDM APPLICATIONS Marketing Precision Marketing Originations Capstone® Customer Management TRIADTM Collections and Recovery Debt ManagerTM Fraud FalconTM Figure 5:
EDM in Other Industries
While Fair Isaac’s applications for financial services represent a core area of focus and expertise, Fair Isaac is committed to leveraging this experience across other areas. The underlying technologies lend
themselves to the development of new and innovative decision services in many industries.
Any area with the following characteristics is a good candidate for Enterprise Decision Management:
»High volume of decisions
»Degree of uncertainty
»Real-time decisions
»Measurable business impact of decisions
Common examples are risk-based decisions, such as operating risk or underwriting risk, and customer treatment, including marketing, multi-channel synchronization, customer service exception-processing and account opening.
For information on Fair Isaac solutions for your industry, contact your Fair Isaac representative or visit www.fairisaac.com.
Historically our focus has been on these decision applications and the individual enabling
technologies. Today that focus has expanded to include a common EDM architecture that supports systems—linked and independent, packaged and custom—making customer decisions across the lifecycle.
The following pages present three parallel EDM development efforts:
»The development of the common underlying EDM
architecture
»The product roadmaps for some of our major
products, including new functionality as well as their integration with the EDM architecture
»The product roadmaps for the core technologies, which are available as separate offerings as well as components within our EDM applications
Future documents will expand on the information presented in the following sections, and present more information both on the architecture itself and on our solutions for financial services, insurance, telecommunications and other industries.
The EDM Architecture
Fair Isaac’s EDM architecture brings greater flexibility, higher analytic performance and better decisions across the lifecycle. Building on Fair Isaac’s broad and deep experience in developing EDM
applications, the architecture is service-oriented, designed for easy standards-based integration with your core systems, and will maximize the ability of companies to leverage technology, expertise and analytics across their decisions. The architecture supports and delivers ever more powerful analytics that operate both within particular kinds of decisions and across them.
Some of the new capabilities that the EDM architecture gives you are:
»Common analytic development, testing and deployment technologies and processes across the entire lifecycle. By having consistent analytic
development and deployment processes and technologies, your operational complexity is reduced substantially. Training costs are reduced, error rates go down, and personnel can more easily move from one area to another.
»Common data model. The common data model
makes the development of new predictive models and decision models, especially those that connect across the customer lifecycle, much easier. It provides current and consistent data, in a format that is readily available for the model building tools. The common data model also enables consistent reporting, monitoring and notification capabilities that span the customer lifecycle.
Portal
Calling Application
System Mgmt . . .
DESIGN TIME EXECUTION TIME CLIENT ENVIRONMENT
Common Browser-Based GUI
Strategy Design Services Layer Model Management Rule Management Simulation & Optimization Configuration & Deployment
Decision Orchestration Mgmt & MonitoringBusiness Process
Case Management
Business Services
Security
Performance Learning
Rules Models AdaptiveEngines Service Invocation Ent e rprise Ser vic e Bus
»Decision/outcome data. One of the significant
advantages you gain is the ability to see how decision strategies you applied early in the lifecycle impact later stages in the lifecycle. For example, you can see the impact of a specific marketing campaign on your collections segmentation.
»Integration with core business systems.
Application components built on this architecture will via standards-based integration technologies. The integration to your core business backbone systems, such as core banking or policy
management and claims systems, becomes easier.
»Support for advanced analytics. The architecture is
designed to enable the deployment of adaptive and other advanced analytics.
»Unified case management facility. Decision
services often identify the need for a case, such as a potential instance of identity theft or in collections management. As a result, Fair Isaac’s applications have long had sophisticated case management systems, such as those in Falcon® Fraud Manager, Debt ManagerTMsolution and Vectus®
software. The EDM architecture will deliver a consistent, sophisticated case management capability, which will also help in connecting decisions. For instance, a collections case may be identified as fraud, and a common case management component will enable this to be managed more effectively.
These new capabilities will give you more control and prepare you to make better, more consistent and more informed decisions in today’s increasingly competitive global environment.
Fitting into the Modern IT Architecture
In designing the EDM architecture, we are building on top of an industry-leading middleware technology stack, IBM WebSphere, to help ensure the robustness, performance, and scalability of our applications. We have also taken pains to ensure that it will be easy to implement in our clients’ IT environment.»Use of IBM WebSphere Middleware technologies:
• Business Process Management: WebSphere Process Server, Business Monitor,
and Business Modeler
• Application Environment: WebSphere Application Server and Portal Server
• Integration: WebSphere Enterprise Service Bus, Message Broker, and Registry & Repository • Data Management: WebSphere InfoServer
»Support for standards:
• Support for the leading database management systems
• Browser-based user interfaces that support portals
• Support for common modeling, rules and business process-related standards • Support for common deployment and
integration standards, such as J2EE and Web Services
We have been careful to avoid the creation of a monolithic platform. The architecture consists of many component services for use in a service-oriented architecture. This enables third-party components to be used where advantageous.
Decision Services
Figure 8 represents the core of this architecture, the decision services layer. Designed to deliver
sophisticated, high-performance decision services usable in a wide variety of environments, the architecture has a number of critical characteristics.
»Sophisticated decision services
Sophisticated decision services require that rules and predictive models can be brought together and orchestrated to deliver the right decisions, actions, in operational transactions. The
architecture brings Fair Isaac’s historic strength in both rules and models together into a single environment for decision orchestration.
»Powerful point analytics
Different decisions require different forms of analytics. The architecture brings the right analytics to bear on each decision by integrating all internal and external data and delivering them to a rich set of analytic models.
»Integrated analytics
As the state of the art moves to integrated
customer decisions, the analytics supporting those decisions must likewise be integrated. By
integrating across decisions and sharing common analytic components, the architecture ensures that these analytics can be developed and deployed.
»Interoperability
While decision services are necessary for effective operations, they are not sufficient. They must be
Adaptive Analytics Engines
Strategy Orchestration
Champion / Challenger Learning
Decision Orchestration Multi-Step Flows Rules Models Adaptive Analytics Engines Context Performance Event Analytic Application Analytic Ent e rprise S e rv ic e Bus EDM Repository Decision S e rvic es C o nnec tors Decision Invocation Decision Decision S e rvic es La y e r
able to interoperate easily with a wide variety of client systems including but not limited to portals, CRM systems, core banking systems, claims management systems, marketing campaign systems, etc. By leveraging IBM WebSphere—integration middleware that is well-established in Financial Services—the EDM architecture will readily
interoperate within existing client IT environments.
»Adaptive control
Smart decisions aren’t static—they adapt. The architecture integrates adaptive analytics and decision strategy control to ensure that decisions can evolve effectively and remain optimal.
»Monitoring and management
Companies increasingly run their business “by the numbers” with reporting, dashboards and Key Performance Indicators all playing a role. Decision services, and the decisions they make, must be integrated with this measurement. The architecture ensures that companies understand what decisions are being taken in real time.
»Decision service integration
Decision services are deployed within a variety of application environments that have a wide range of performance and throughput requirements. With that in mind, the EDM architecture is designed with multiple methods for integrating with decision services—supporting standard web service interfaces as well as native language (e.g. Java, Cobol) integration where performance is critical.
Summary
This EDM architecture is a combination of existing Fair Isaac products (described in the EDM Technologies Roadmaps section), standard capabilities and components delivered as part of EDM applications (described in the EDM Applications Roadmaps section), and processes and standards used throughout Fair Isaac. The various decision service components can be accessed by applications within the architecture, and by other applications, through standard integration mechanisms such as web services.
This EDM architecture builds decision services— services that make critical, high-volume business decisions—that can easily be assembled into composite applications. A composite EDM
application, made up of a number of independently developed services, delivers the value of the decision service in an operating environment. A composite application of this sort combines a decision service with elements such as case management and reporting and “plugs” into a core banking system or business process management system to become part of a whole process.
The next section describes how Fair Isaac is using this EDM architecture to develop composite applications to deliver its core solutions for financial services.
EDM Application Roadmaps:
Financial Services
Fair Isaac offers applications for multiple industries, and we focus on financial services in this section because this is the industry for which we have developed the most “point solutions.” Similar development efforts are underway for our full product portfolio, but the first examples of
applications moving to the EDM architecture will be in financial solutions.
Fair Isaac’s development effort for its EDM
applications for financial services has two objectives. First, we will continue to improve our applications by adding features and functionality that ensure they
remain or become best-in-class solutions. Second, we will bring these solutions fully onto our EDM architecture, in order to bring clients the full value of connected decisions, as well as reducing client costs such as training and maintenance.
Incremental Improvement Priorities
»Continued upgrades: We will not put needed
improvements on hold while we undertake the work to bring our applications onto our EDM architecture.
»Gradual inclusion of standard components: As
standard components are developed through the emerging EDM architecture, we will integrate them into new releases to reduce development and deployment costs and improve consistency.
»Quick-value “connectors”: We will develop a series
of “product connectors” that support explicit integration of existing products. These connectors will allow customers to connect decisions in the near term, while our architecture development efforts proceed.
EDM Architecture Priorities
»Product consolidation. Over the next three years,
we will consolidate our product portfolio in areas where we currently support more than one primary application—such as originations (Capstone® Decision Manager, LiquidCredit®
service) and collections and recovery (Debt ManagerTMsolution,
Recovery Management SystemTM
solution).
»Component adoption. We will use the standard
components being developed as part of the emerging EDM architecture to standardize the development, installation and operation of EDM applications, and thus reduce costs and risk.
»Continued lifecycle integration. We will continue to
identify areas where there is significant potential value from connecting decisions, or from integration with our current and older products, and focus our development efforts on these critical intersections.
Originations
The initial focus of the originations roadmap is on upgrades and incremental improvement to specific originations products:
»Capstone®Decision Manager, Fair Isaac’s high-end
application processing solution, released version 3.3.4 in June 2007. This release is the culmination of over 50 maintenance items, addressing many different user needs. Another release to enhance security features and provide customer-driven enhancements is planned for the second half of 2008.
»Capstone Decision Accelerator Fair Isaac’s “built for
purpose” origination decision service, built upon the Blaze Advisor®
business rules management system, released version 1.2.5 in August 2007. This release builds on the robustness of the prior release and provides several new requested features, including extended architecture support. Release 2.0, available in summer 2008, will offer access to the current release of the Blaze Advisor system.
»LiquidCredit®Decision Manager, Fair Isaac’s
recommitment to small business application processing services, released in November 2007. In particular it addresses clients’ need for speed (timely access to Small Business Scoring ServiceTM models) and choice (flexibility in determining what
»Capstone Intelligent Data Manager, Fair Isaac’s
data acquisition product, release 1.1.1 is coming in summer 2008. This release will add access to the major Canadian consumer credit reporting agencies and will provide an interface to administer data source subscriber information and manage default settings for processing parameters.
»LiquidCredit Decision Engine, Fair Isaac’s hosted
analytic and data acquisition service, has release 6.6 coming in spring 2008. This release will include new and improved Application Risk Models version 3.0, as well as small business scoring models that utilize Paynet data.
In 2008 our main focus is on our current product set’s evolution towards our vision of four service-oriented components that work seamlessly with other EDM applications.
»Enterprise Origination Platform 1.0, scheduled
for release in late 2009, includes four modules: application processing, decisioning, data acquisition and analytics. These modules will build on work already completed, particularly the Capstone Decision Accelerator (decisioning) and Capstone Intelligent Data Manager (data acquisition) products.
While the focus will shift to the new enterprise originations products, the current plan is to continue to make select enhancements to existing products such as Capstone Decision Manager and LiquidCredit
Customer Management
The roadmap for customer management is the roadmap for Fair Isaac’s market-leading TRIADTM adaptive control system.
»TRIAD 8.2 was made available in September 2006
and was the first step in adopting the emerging EDM architecture. TRIAD 8.2 uses Blaze Advisor as the rules and analytic deployment engine, as will all products in the EDM application suite in the future. Blaze Advisor brought greater flexibility to the control of models and decision keys in the TRIAD system, allowing manipulation of nested arrays of data, for example, and greater usability.
»TRIAD 8.3, released year-end 2007, is the latest
version of the TRIAD 8.x series which first brought account management and customer management together within the same platform. Among the key features of TRIAD 8.3 are: improved system processing, greater flexibility for creating and maintaining strategies and scenarios within the Credit Facilities decision area, and a sophisticated extract tool for the TRIAD Analytic Datamart (TAD) that facilitates using TRIAD data for strategy and model optimization purposes. Additionally, TRIAD 8.3 focuses on making migration easier for our processor partners and the 40+ end user clients on TRIAD 7.0. This is the first in a series of migration tools for customer management which will assist in the migration of control tables (including strategies) and linkage files. Further, tools have been developed to help TRIAD users migrate Blaze Decision System projects to Blaze Advisor. It is expected that these tools will reduce this part of the typical TRIAD upgrade by 50% and 75%.
»TRIAD 8.4, released in 2008, will include improved
strategy definition and visualization, making it easier to find problems and opportunities early in the design process. It will also include an update from Business Objects 6.5 to Business Objects Xi
»TRIAD 9.0, scheduled for 2009, will sport a new,
more intuitive user interface based on the common UI framework. The structure of constructing a decision, proven over 20 years and 300+ implementations, will be retained but will now include greater flexibility in decision design with an upgraded TreeViewTM
system. The new TRIAD 9.0 user interface will be able to be implemented directly on top of the TRIAD 8.4 control tables and host system components, making it simpler for TRIAD 8.4 clients to install. Additionally, TRIAD 9.0 will extend the real-time decisioning capabilities to other decision areas outside of Authorizations, and will provide improved reporting capabilities using OLAP.
Collections and Recovery
For 2007, the collections and recovery product family continued to see an investment in the individual products.
»Debt ManagerTMsolution 6.3, released in November
2007, added online document printing and storage capabilities and platform support for SQL 2005 and Internet Explorer 7.0. The release also enhanced support for flexible installment plans, with capabilities to cope with longer-term and more diverse plans.
»RMS NGTMsolution 2.0.1, released in March 2007,
added multi-streaming to enhance performance for a number of critical batch functions, field masking for Tax ID and date of birth, and platform support for SQL Server 2005.
»PlacementsPlus®service is currently available on
release 4.18. Recent improvements have been made to the service to include a Debt Sale module that allows for great visibility and control over the process of identifying, bidding and transfer of assets to a debt buyer. Additionally, with release 4.18, credit grantors now have the capability to manage outsourcing activities of multiple accounts for a single customer as one collection activity. This customer-level functionality will allow for increased visibility to the overall debt obligations of a
customer and greater overall returns from
Moving into 2008, the collections and recovery platforms will converge on an enterprise collections and recovery product family built on the emerging EDM architecture.
»Enterprise Collections and Recovery release 1.0 in
December, 2008 will adopt the EDM architecture’s common user interface and reporting framework to improve user configurability and decision control. The new platform will be service-oriented and will use the common EDM case management
capabilities. This common case management engine will allow a single instance of case management to accommodate treatment of different cases across the enterprise. Fair Isaac will provide an upgrade path for all existing Debt ManagerTM
solution, RMS NG solution and Atlas clients that contains all of the functionality available in those applications today, along with additional features and capabilities to be the absolute best-in-class collections and recovery tools.
»Release 2.0 of the platform, in 2009, will incorporate PlacementsPlus®service as an end-user deployable module, giving more choice in access to the exceptional agency management and reporting capabilities provided today only as an ASP service.
Fraud
There are currently two primary software development paths for Fair Isaac fraud solutions across multiple industries—one path (previously referred to as Falcon®
Fraud Manager 5.x) is based on Java, and another path is based on enhancements and innovations to the more widely deployed fraud solutions architecture which, in the financial industry, was delivered as Falcon 4.x for Unix and Falcon 4.x for Mainframe. This latter path also supported other Fair Isaac industry fraud verticals (such as Telecom Fraud and Merchant Fraud) and other Fair Isaac financial fraud solutions (such as Advances Fraud). The improvements to the more standardized software which have been made over the last few years will be applied to the current Falcon 4.x Unix and 4.x Mainframe deployments with a merged Falcon 6 solution.
As we are moving to a more standardized delivery approach with EDM, the fraud solutions roadmap is taking the “best of the best” of both paths and combining them into the merged Falcon 6. Before delivering the merged Falcon 6 solution, each existing Falcon path (Java, Unix and Mainframe) will have at least one final release (5.X) that meets the more near-term needs for their particular path.
»Falcon 5.1.3 in 2007 will allow better response time
from support, and allow engineering teams to monitor more effectively and respond more quickly to performance/system issues in the Java version of the software.
The 4.x Unix and 4.x Mainframe versions of Falcon system have not had significant enhancements for some time. Before moving to a merged Falcon 6, we are investing in several areas in the 5.x versions that will quickly provide value over the current 4.x versions. These enhancements will also be re-used in the 6.x release.
»Falcon 5.2C (Unix) and 5.2M (Mainframe) in 2007
included features to achieve PCI compliance. Two key areas for PCI are to encrypt “data at rest” and to provide “masking” for privacy of sensitive fields. Enhancements in these releases allow the solution to take advantage of several analytic innovations, such as global profiling, multi-level profiling, adaptive analytics and shared memory profiling (for high performance in more complex analytic scenarios). Falcon 5.2C also added a browser-based case manager for the analyst and configuration workstations, as well as an expanded API to include additional fields for use by the analytics and rules. The remaining 5.2M focus is on code optimization to reduce MIPs.
»Falcon®ID is a comprehensive solution for
detecting and resolving cases of suspected identity fraud. Falcon ID 1.2.3 will provide support for Solaris and Windows, and will also provide for encryption of the data store.
»Falcon®
Fraud Manager for Merchants is a
comprehensive solution for merchants, which detects fraud in a Card-Not-Present (CNP) environment such as e-commerce, mail order or telephone order. The primary enhancement in release 4.0 are providing PCI compliance and web services.
»Falcon Transfer, our Faster Payments fraud solution,
analyzes the risk of funds transfer in real time, reviewing both the sender’s and receiver’s account profile. When it detects signs of fraud, it stops the payment before settlement and sends it to the case management system for further review by a fraud analyst.
»Falcon Enterprise 1.0, to be delivered in late 2008,
incorporates a combination of the best capabilities of the current Falcon 5.x solutions, such as a sophisticated server component of the case
manager, and the best of the advances made to the more standardized fraud solutions, such as more configurable APIs, multiple data source inputs, the ability to deploy several analytic innovations, and more sophisticated profiling. The vast majority of the components and capabilities for Falcon Enterprise 1.0 have been fully tested in a prior Falcon version or another Fair Isaac fraud solution.
»Falcon Enterprise 2.0 will allow for the delivery of
fraud solutions across multiple channels, creating an enterprise fraud solution capable of handling several channels in the payment card industry, in retail banking, and in other industry verticals. It will also extend the case management capabilities to include research and investigative views, and tools to manage more complex forms of fraud and facilitate uncovering previously undiscovered linkages and patterns. Falcon Enterprise 2.0 also adopts key elements of the emerging EDM architecture, including Blaze Advisor for rules authoring, and the standard enterprise case management functions.
EDM Technology Roadmaps
While we develop the applications for particular decision areas, Fair Isaac will continue to improve upon the core technologies we use to deliver Enterprise Decision Management. We use these technologies to develop our applications, and to deliver custom solutions for our clients; in addition, some of them—including the Blaze Advisor®business rules management system and the Model Builder predictive modeling software—are available as standalone software tools.
Available now:
»Blaze Advisor 6.5 has support for lifecycle
management and rule verification and validation to make it easier to create and deploy sophisticated rules-based decisioning. With its support for PMML it is now easier than ever to integrate rules with analytic models, especially those developed using Model Builder.
»Model Builder 3.6 adds greater ease of use for
new and veteran users, particularly in importing, analyzing, and sampling new datasets, and in rapidly constructing new scorecard models to learn the predictive relationships among the data.
In 2007-2008:
»ScoreNet network’s “active/active” infrastructure
was added in summer 2007. This dual-data center environment will support true real-time, 24x7, high availability delivery of data and analytics.
»Blaze Advisor 6.6 will add more deployment
management, improving the flexibility and control over decision services in complex environments.
»A new decision simulation capability will allow business users to better understand the impact of proposed changes.
In 2009:
»Blaze Advisor and Model Builder will share a
single, extensible development environment.
»A wide range of ScoreNet data sources, including common schemas and characteristics, will be easily deployable in Blaze Advisor and Model Builder, enabling rapid development of data-rich decision services. Not only will this make adoption easier for those already familiar with standard development tools, it will make the integration of rule and model development tighter.
»Web 2.0 technologies and AJAX will also be brought to bear to maximize the effectiveness of business users as they work on the rules in their applications. Extensibility, thanks to the adoption of widespread standards, will allow clients, partners and Fair Isaac to bring new development innovations to market faster, further aiding clients and Fair Isaac development teams.
Analytics Roadmap
Fair Isaac’s applications are powered by advanced analytics—the science behind smarter decisions. We will continue to update and enhance the analytics that are embedded in our applications, in order to provide clients with the sharpest predictions possible. We will also add new analytics to applications. Additionally, we continue to design new types of custom analytics and strategy design services.
A short selection of planned releases includes:
»Fraud: Short-term improvements in consortium
models include new releases of the UK credit and debit models, new releases of US credit and debit (both signature and PIN) models and new releases of credit models in Japan, Malaysia, Europe and Switzerland. Further development of a number of these models, and of credit consortium models in APAC, Australia, Brazil, Mexico and South Africa, will follow. We will also develop adaptive models for bust-out fraud, which will be part of Falcon 6.
»Originations: We are currently developing
new consumer and small business application risk models.
»Collections: We are developing pre-developed
collections models for integration within Debt Manager solution.
»Standalone Analytics: We are developing a
methodology to design customer cross-sell strategies that improve responses and future revenue—by designing segments-driven short- and longer-term actions and effects. We are also developing cross-lifecycle customer segmentation and strategy services.
Migration
To take maximum advantage of the emerging EDM architecture and new releases, clients must be able to adopt new versions and new products and still be able to maintain an integrated, stable operating platform. At the same time, it is important to preserve previous investments in Fair Isaac technology by making any migration a no-loss exercise that focuses on delivering maximum business value to the client.
Fair Isaac Professional Services works with clients as an open business partner committed to client satisfaction at every step of the EDM journey. Our cumulative experience in technology delivery and our tight integration with Fair Isaac’s product development and product management
organizations means that clients receive the highest level of service and support, from the initial
assessment to project completion. Fair Isaac Professional Services’ EDM methodology is being applied to all products, regions and services to provide a consistent, customer-centric delivery experience. Early adopters can even take advantage of extra services and access to Fair Isaac resources.
In just the past three years, Fair Isaac Professional Services staff members have worked on over 9,000 individual projects in 77 countries worldwide. As our product offerings continue to expand, we are committed to developing and enabling a network of partners that will grow with us, providing increased global coverage, local availability and
complementary service offerings, all under the governance of Fair Isaac. Together with our clients’ own organizations, this eco-system of technology and service providers means that projects can be designed and delivered faster, with appropriate skill sets and cost savings.
As our clients move forward in adoption of EDM technology, we will make maximum use of automated migration tools, combined with flexible services to migrate customized components. Fair Isaac Professional Services is prepared to engage with clients immediately in an EDM Assessment that will evaluate readiness for Enterprise Decision Management from both business and technology perspectives. By developing a business case-founded roadmap, all parties have a clear view of the possible initiatives, deliverables and expected return on investment to be gained. This planning process will lead to projects executed by the most appropriate resources, using an outcome-based approach focused on client satisfaction.
Looking Forward
There are many ways that Fair Isaac’s emerging EDM architecture and the applications being developed on it can support your business.
Distributed Decisions
With ScoreNet network, Fair Isaac provides connectivity to hundreds of external data sources and service providers. The ScoreNet network allows many new and sophisticated “decision points” to be automated and enhanced with our EDM technologies, and then delivered seamlessly over the network.
Decisions required in a particular location can be made elsewhere and delivered using ScoreNet network. Examples include the ability to dynamically access multiple external data sources and score the
combined information before returning the results, model-driven actions triggered by an event using the network for delivery of both event and results, and optimization of account placement with outside agencies. The combination of the emerging EDM architecture and ScoreNet network allows for decisions to be made and rapidly distributed using a state-of-the-art network.
Customer Centricity
Companies that are out of alignment with their customers or losing customers across lifestages are becoming more focused on those customers. Enterprise Decision Management, and the emerging EDM architecture, is a perfect platform for customer centricity. Customer centricity has six guiding principles, shown below, and an EDM approach supports all six.
PRINCIPLE
EDM
Choose Which Customers
to Focus On
Clear, Insightful
Customer Segmentation
Winning Value Proposition
for Each Segment
Deep Alignment
Across the Enterprise
Analytics can identify not only the current high-value customers, but also the future high-value customers,
as well as those prospects that are likely to respond to an offer or other action.
The integration of analytics, including those for segmenting customers, and powerful analytic techniques and
tech-nologies from Fair Isaac’s R&D labs ensure that the most effective customer segmentation is applied to all decisions.
The use of decision models and optimization means that the most compelling value propositions are identified
quickly, tested scientifically and deployed easily.
Customer treatment decisions can only be aligned across the enterprise if an infrastructure for sharing, managing
and deploying decisions across the enterprise exists. EDM delivers such an enterprise decisioning backbone.
Customer centricity is a cross-functional exercise focused on developing insight and applying actions to ensure maximum value by neither under-serving or over-serving specific customers. Companies adopting customer centricity have seen industry-leading results. EDM and the EDM architecture are critical enablers for a company’s customer centricity program.
Enterprise Fraud Management
Fraud is evolving from an opportunistic, repeatable, mostly local problem to one that is orchestrated, dynamic and cross-border. New fraud problems caused by organized gangs attempting massive compromises, combined with phishing and malware attacks, must be repelled while still handling
everyday local fraud and delivering customer service. Fraud losses must be minimized yet customers cannot be disrupted or they will move elsewhere.
EDM and the use of the EDM architecture as the backbone of Fair Isaac’s enterprise fraud solutions represent the state of the art in dealing with today’s fraud challenge. Through business rules that capture known best-practices for fraud prevention, and predictive analytics that detect fraudulent activity mid-transaction, EDM provides the ideal capabilities for today’s fraud environment.
Extending Legacy and Core Platforms
Unlike so many approaches, especially those that contain the dreaded “E” word—Enterprise—EDM is a highly additive strategy. Enterprise Decision Management takes advantage of the infrastructure and data you have today, integrates with the service-oriented and process-centric architecture you are developing and adds value at every stage.The EDM architecture, and the EDM technologies, can easily extend existing legacy applications, giving them a new lease on life by making core decisioning components more agile and more compliant. Instead of replacing complete legacy applications, a high-risk and costly proposition, EDM allows you to renovate the critical components for a high return on
investment.
For companies running core banking systems or claims management systems, EDM offers a powerful tool for improving the straight-through processing rates and bringing best-in-class origination, customer management, fraud and other decisions to bear.
Fair Isaac’s approach of focusing on the development and integration of decision services for high-volume, high-value operational decisions allows your
ongoing adoption of service-oriented architecture and business process management to proceed without surrendering control and optimization of the decisions that run your business.
If, like most companies, your architecture mixes these different approaches across lines of business or geographies, the connectivity of decision-making offered by the EDM architecture and Fair Isaac’s focus on connected decisions ensures consistency of critical business decisions, despite the variety of technologies used to capture, display or process information about your products and customers.
Universal Decisioning
As companies adopt EDM, many of them are considering the possibility of a universal decisioning platform. Such a platform, the logical outcome of EDM adoption, results in smart decisions across the enterprise and maximizes the value of coordinating interrelated decisions while reducing operational negation. A universal decisioning backbone allows companies to rapidly add and adapt decisions within their operational infrastructure, maximizing the value they gain from those decisions.
Conclusion
Fair Isaac’s EDM architecture and the connected applications being developed on it can bring a new measure of value to the operational decisions that drive your business. By automating, improving and connecting your decisions using our applications and technology, you can improve results at every stage of your customer lifecycle, based on a coherent and integrated set of data, rules and models.
This document is intended to start the dialogue with our clients about the EDM architecture. Whatever Fair Isaac solutions you are using today, and whatever your business challenges, we believe that we can help you achieve a higher level of performance. We have been helping companies improve their decisions for more than 50 years— EDM is what we were born to do.
We invite you to discuss our EDM approach, and how it can benefit your business. Please contact your Fair Isaac representative or visit www.fairisaac.com.