What Is Decision Optimization and How Can It Improve Results Across the Customer Lifecyle?

Key Takeaways

• Decision optimization uses prescriptive analytics and mathematical modeling to find the best possible solution to a complex problem, such as maximizing profit or minimizing risk, while respecting business constraints.
• Decision optimization combines predictive modeling, action effect modeling, and business expertise to quantify the outcome of every possible decision and simulate how changing a strategy will affect KPIs.
• Decision optimization can be applied at every stage of the customer lifecycle from acquisition and onboarding through customer management, retention and collections.
• Real-world results speak for themselves. Home Credit International achieved a 26% increase in portfolio profit and a 29% increase in new sales using FICO's AI-powered optimization, and HSBC UK saw a 15% uplift in monthly card spend.
• Decision optimization is becoming a core technology in financial services because it replaces gut-feel strategy changes with data-driven scenario simulation so organizations can test outcomes before committing to a new approach.

Decision optimization is the process of using prescriptive analytics and mathematical modeling to find the best possible solution to a complex problem given at least one objective, a set of constraints, and available data. It is widely used across industries to automate and improve logistically complex decisions like resource allocation, scheduling, routing, and planning.

Decision optimization is also becoming a core technology in financial services, because of its ability to deliver substantial increases in profit for specific strategies, driving the business forward in a systematic fashion.

Watch the short video below to hear Marc Drobe of FICO explains how decision optimization works.

As Mark explains, decision optimization can be described as a mathematical optimization process to develop better decision strategies. What we mean by better is that those strategies are more profitable, more manageable, and most importantly meet business goals and constraints.

What Are the Objectives of Decision Optimization?

The core objective of decision optimization is to find the best possible outcome to a scenario. That can mean maximizing a value (like profit, efficiency, or customer satisfaction) or minimizing a cost (like risk, waste, or time), while respecting all relevant constraints (resource limits, business rules, operational requirements, etc.).

Beyond finding an optimal solution, decision optimization also helps organizations understand why one solution is better than another (for example, why accepting a lower profit margin on one product might lead to better overall returns across a portfolio). It enables faster and more consistent decision-making at scale and allows organizations to test scenarios to see how outcomes would change.

Using decision optimization you can:

• Define scenarios using objectives and constraints
• Solve for the optimal strategy within those given constraints
• Analyze expected effects on KPIs and drill into trade-offs
• Choose the scenario you want to implement

Understanding the Role of Decision Impact Modeling in Decision Optimization

At the core of the optimization approach is decision impact modelling, which combines:

• Predictive modelling: to estimate the probability of a customer behaving in a certain way, such as the likelihood of responding to an offer, or cancelling a service
• Profit modelling: to quantify the financial outcome of each possible decision, including the revenues it could generate and the costs it would incur
• Action-effect modelling: to understand how a specific action will shift behavior compared to taking no action, for example how much a discount increases the probability of a customer accepting an offer
• Business expertise: to incorporate domain knowledge, strategic priorities, and judgment that data alone may not capture, so the decision model reflects real-world context and goals

How to Use Decision Optimization Across the Customer Lifecycle

Decision optimization is often used to derive the specific actions that should be taken to achieve a particular objective, such as raising profit by XX%, under user-defined constraints. 

For example, a loan portfolio manager setting pricing might ask:

• What is the impact on my portfolio if I change my strategy?
• What are the possible points I can run my portfolio at and evaluate the tradeoffs between two competing objectives across all points to make the best decision? 
• What are the possible impacts if I want to increase both NPI and volume, even though they are competing objectives?
• What is the impact on my KPIs if I move away from business as usual and maximize net interest income or origination volume?

Decision optimization can be used to improve strategies and outcomes across the customer lifecycle:

Customer acquisition

• Identifying which prospects to target, with what offer, and through which channel
• Maximizing the likelihood of conversion while staying within budgets

Customer Onboarding

• Determining the right product configuration or service tier to assign to a new customer based on their predicted behavior and risk profile

Customer Management

• Deciding which customers to proactively engage, with what product or service adjustment, and at what point in their journey
• Optimizing for long-term value by balancing cross-sell and upsell opportunities against the risk of over-contacting or eroding trust

Customer Retention

• Predicting which customers are at risk of leaving and determining the most cost-effective intervention, such as a targeted offer or outreach, to retain them

Expansion

• Identifying which existing customers are most likely to benefit from, and accept, additional products or higher credit limits
• Determining the right offer and timing

Advocacy

• Identifying customers who have had consistently positive experiences and are most likely to refer others or act as brand advocates
• Determining the right incentive or recognition to encourage advocacy

Decision Optimization in Action: Examples and Use Cases

A good example of decision optimization is action is Home Credit International’s loan pricing optimization project. In this post, Petr Kapoun, Erste Group Risk Management Advisor and former Chief Risk Officer at Home Credit Russia, explains how he has used FICO AI-powered optimization to achieve remarkable results: 26% increase in portfolio profit and a 29% increase in new sales. According to him, one of the key benefits of decision optimization is to use analytics, rather than hunches, to simulate outcomes of changing strategies.

 How FICO Can Help You Optimize Decisions

• Discover the optimization capabilities in FICO Platform
• Explore FICO Xpress Optimization, the world-class mathematical optimization solver used for large-scale problems in a variety of industries such as banking, ecommerce, supply chain, logistics, manufacturing, and more.
• Request a free trial license for Xpress to see its decision optimization capabilities for yourself
• Download this case study to see how HSBC UK achieved a 15% increase in monthly spend thanks to Xpress
• Read this case study to learn how Xpress helped Pepsico reduced daily production solve times by 30%

This is an update of a post from 2020.

Frequently Asked Questions
 

What is the difference between decision optimization and predictive analytics?

Predictive analytics estimates what is likely to happen, for example, the probability that a customer will default or churn. Decision optimization takes those predictions and determines the best action to take given those predictions, your business objectives, and your constraints. The two work together: better predictions lead to better optimization.

How long does it take to implement decision optimization? 

Implementation timelines vary depending on the complexity of the problem, the availability of historical data, and the maturity of your existing analytics infrastructure. That said, platforms like FICO® Xpress are designed to accelerate optimization speed so organizations benefit from fast deployment and faster returns.

Do you need a data science team to use decision optimization? 

While advanced optimization projects benefit from operations researchers and data scientists, modern decision optimization tools are increasingly accessible to business analysts and risk managers. FICO® Xpress, for instance, allows business users to define objectives, set constraints, and simulate scenarios without needing to write complex mathematical models from scratch.

How is decision optimization different from rules-based decisioning? 

Rules-based decisioning applies fixed, conditional logic to decisions, for example, "approve if credit score is above X." Decision optimization, by contrast, dynamically determines the best action across a large number of variables and possible outcomes simultaneously, allowing organizations to balance competing objectives and adapt strategies as conditions change. Rules still play a role, but optimization determines the best configuration of rules to meet your business objectives.