EXECUTIVE SUMMARY: HOW FICO MAKES MONEY

Imagine you want to buy a house. You walk into a bank, fill out a mortgage application, and the loan officer pulls up a three-digit number that essentially determines whether you get the loan and at what interest rate. That number is almost certainly a FICO Score. Fair Isaac Corporation invented consumer credit scoring in 1956 and has spent seven decades embedding its product so deeply into the American financial system that it has become invisible infrastructure — like the electrical grid or the plumbing behind the walls. You rarely think about it, but virtually nothing in consumer lending works without it.

FICO makes money two ways, and the simplicity is part of what makes the business extraordinary. First, every time a lender — a mortgage company, an auto dealer, a credit card issuer, an insurance underwriter — needs to evaluate a consumer's creditworthiness, the lender requests a credit report from one of the three major credit bureaus (Equifax, Experian, or TransUnion). The bureau applies FICO's proprietary algorithm to the consumer's data and produces a score. The lender pays the bureau for the report, and the bureau pays FICO a royalty for each score generated. FICO does not store the data, does not deliver the report, and does not interact with the lender — it simply licenses an algorithm and collects a fee every time it runs. The cost to FICO of producing one additional score is effectively zero: the algorithm already exists, the bureaus run it on their infrastructure, and FICO's capital expenditure in fiscal 2025 was just $8.9 million on nearly $2 billion in revenue. This is a toll booth with no maintenance costs.

Second, FICO sells decision management software — a cloud platform that helps banks and other financial institutions automate complex decisions like whether to approve a loan, whether a transaction is fraudulent, how to price insurance, or how to manage a delinquent account. This software business generates recurring subscription revenue and is growing at 33% annually for the platform component, though it remains smaller and lower-margin than scoring. The genius of the combined business is that the regulatory monopoly in scoring documented in our earlier chapters provides a natural customer acquisition channel for the software: every bank that uses FICO Scores is a potential buyer of FICO's decision management platform, and no competitor can offer the native integration between the industry-standard credit score and the decisioning software that acts on it.

In fiscal 2025, FICO generated $1.99 billion in revenue, $925 million in operating income (47% margins), and $735 million in free cash flow — from a company with total assets of just $1.87 billion and capital expenditures of $9 million. The business converts revenue into cash with an efficiency that most companies cannot approach.

1. HOW DOES THIS COMPANY ACTUALLY MAKE MONEY?

Walking Through a Transaction: The Mortgage Score

A family in suburban Ohio wants to buy their first home. They apply for a mortgage at their local credit union. The loan officer initiates the process by requesting a tri-merge credit report — a combined report from all three national bureaus. Each bureau retrieves the family's credit data, applies FICO's proprietary scoring algorithm, and produces a FICO Score. The credit union receives three scores (one from each bureau) along with the underlying credit reports.

For this single mortgage application, FICO has just earned three score royalties — one from each bureau — without lifting a finger. FICO did not acquire the data, did not store it, did not deliver the report, and did not interact with the credit union or the family. Its algorithm ran on bureau infrastructure, and a royalty flowed back. If this family also gets auto-quoted for insurance and applies for a credit card the same month, FICO earns additional royalties on those score pulls too. Multiply this across hundreds of millions of credit decisions annually across the entire U.S. economy, and you understand why FICO's Scores segment generated $305 million in a single quarter — $1.22 billion annualized — growing 29% year-over-year.

The Mortgage Direct Licensing Program, discussed extensively by CEO Will Lansing on the January 2026 earnings call, is restructuring this value chain. Instead of royalties flowing through bureau-bundled pricing, FICO is establishing direct contractual relationships with mortgage lenders through approved resellers. Five reseller participants signed in Q1 2026, with production integration testing near completion. This gives FICO direct control over pricing, visibility into lender-level volume, and the ability to sell adjacent products — the FICO Score Mortgage Simulator, performance monitoring models — directly to the end customer. It is a structural shift that captures more value per transaction.

Walking Through a Transaction: Platform Software

A major European bank wants to reduce fraud losses on its credit card portfolio. It evaluates FICO Platform's Enterprise Fraud Solution, which uses real-time machine learning to score transactions and flag suspicious activity. After a 9-month evaluation and proof-of-concept, the bank signs a multi-year SaaS contract with annual committed revenue plus usage-based fees tied to transaction volumes. FICO recognizes this as recurring ARR. Over the next two years, the bank deploys FICO Platform for three additional use cases — account origination, customer management, and collections optimization — each adding incremental ARR. This is the land-and-expand motion that produced 122% platform net dollar retention in Q1 2026.

Revenue Breakdown by Business Segment

Scores Segment Deep Dive:

The Scores segment is FICO's economic engine — the regulatory monopoly mapped in Chapters 1 and 2 manifesting as financial performance. It has two sub-segments:

B2B Scores (~90% of Scores revenue, ~$1.1B annualized): Sold through credit bureaus (and increasingly through the DLP) to lenders, insurers, and other commercial users. B2B revenue grew 36% year-over-year in Q1 2026. The sub-segment breaks down further by origination type: mortgage originations accounted for 51% of B2B revenue and 42% of total Scores revenue in Q1 2026, with mortgage Scores revenue up 60% year-over-year. Auto originations revenue grew 21%. Credit card, personal loan, and other originations grew 10%. Pricing is per-score royalty — the specific dollar amount per score is not publicly disclosed but has been increasing significantly, particularly for mortgage originations through the DLP program. Customers are the three national credit bureaus (as distributors) and, increasingly, mortgage lenders directly (through DLP resellers). This sub-segment's gross margin is estimated at well above 90% given near-zero marginal cost of score generation.

B2C Scores (~10% of Scores revenue, ~$120M annualized): Sold directly to consumers through myFICO.com subscriptions and indirectly through credit card issuers and monitoring services. B2C revenue grew 5% year-over-year in Q1 2026. This is a mature, slower-growth segment facing competition from free alternatives (Credit Karma, card issuer programs). Pricing is subscription-based for myFICO.com (approximately $30-40/month for premium tiers) and royalty-based for indirect channel partners.

Software Segment Deep Dive:

The Software segment is undergoing an internal transition — migrating customers from legacy on-premises products to the cloud-native FICO Platform — that makes headline growth (2% year-over-year) misleading. The two sub-businesses within this segment have dramatically different trajectories:

FICO Platform (~40% of Software ARR, $303M ARR, growing 33%): Cloud-native decision intelligence platform for real-time decisioning at scale. Deployed across fraud detection, origination, customer management, and collections use cases. Recognized by Gartner as the leader in Decision Intelligence Platforms. 150+ customers, more than half on multiple use cases. Pricing is SaaS subscription with committed annual revenue plus usage-based components. Platform NRR of 122% demonstrates expanding customer relationships. ACV bookings hit a record $38 million in Q1 2026, up 36% on a trailing 12-month basis. This is the growth engine management is building to diversify beyond scoring.

Non-Platform Legacy (~60% of Software ARR, $463M ARR, declining 8%): Legacy on-premises products including Falcon (fraud), TRIAD (account management), and Originations Manager. NRR of 91% reflects managed decline as customers migrate to Platform or churn legacy products. Includes the CCS (Customer Communication Services) business where ARR growth was flat. Revenue declined 13% year-over-year in Q1 2026 due to lower point-in-time license revenues and end-of-life product retirements.

2. WHO ARE THE CUSTOMERS AND WHY DO THEY CHOOSE FICO?

FICO's customers fall into three distinct categories with very different relationship dynamics.

Credit Bureaus (Scores distribution partners): Equifax, Experian, and TransUnion are FICO's primary revenue channel for B2B Scores. They integrate FICO's algorithms into their credit report delivery systems and pay royalties on each score generated. The relationship is simultaneously cooperative (bureaus need FICO Scores because lenders demand them) and adversarial (bureaus co-own VantageScore, the only competitive alternative). No single bureau accounts for more than approximately 33% of Scores revenue, but collectively the three bureaus represent the overwhelming majority of B2B Scores distribution. The DLP is partially disintermediating this relationship by establishing direct lender connections.

Financial Institutions (end users of Scores and Software): Major banks, credit card issuers, mortgage lenders, auto finance companies, and insurers. These institutions use FICO Scores for credit origination, account management, and portfolio monitoring. A subset also licenses FICO Platform for decision management. The relationship is driven by the institutional switching costs analyzed in Chapter 2: these customers choose FICO because the FICO Score is the regulatory standard, the systemic standard, and the career-safe standard. They are not delighted customers choosing FICO over delightful alternatives — they are institutionally bound customers operating within a system built around FICO's product. This distinction matters: the 122% platform NRR suggests software customers are genuinely expanding by choice, while scoring customers are locked in by structure.

Consumers (B2C): Individual consumers access FICO Scores through myFICO.com subscriptions and card issuer free-score programs. This is the weakest customer relationship — free alternatives from Credit Karma and card issuers provide adequate substitutes for most consumer needs.

If FICO Disappeared Tomorrow: The mortgage industry would face an immediate operational crisis. Automated underwriting systems at Fannie Mae and Freddie Mac require FICO Scores. Every mortgage lender's risk pricing model is calibrated to FICO Score ranges. Secondary market investors price mortgage-backed securities based on FICO Score distributions. Bank examiners reference FICO Scores in assessing credit risk management. There is no quick substitute — rebuilding the system around an alternative scoring standard would take years of coordinated institutional effort. For FICO Platform customers, the disruption would be severe but not existential — alternative decision management platforms exist, though migration would be costly and risky.

3. WHAT'S THE COMPETITIVE MOAT IN SIMPLE TERMS?

The moat analysis in Chapter 2 documented FICO's advantages in detail. In the simplest possible terms: FICO's algorithm produces a number that the entire American financial system has agreed to use as the standard measure of whether you're likely to repay a loan. That agreement is not just commercial convention — it is embedded in federal regulations, government-sponsored enterprise requirements, automated underwriting systems, risk-based capital calculations, and insurance pricing guidelines. Changing the standard requires coordinating dozens of regulatory bodies, thousands of financial institutions, and millions of interconnected systems simultaneously. No individual participant has the incentive or authority to drive that coordination.

If Jeff Bezos decided to compete with FICO tomorrow with unlimited capital, he could hire the world's best data scientists and build a statistically superior credit scoring model within a year. He would then discover that statistical superiority is worthless without regulatory acceptance by Fannie Mae and Freddie Mac (timeline: undefined — FICO's own upgrade to Score 10T has been in process for years), adoption by thousands of lenders willing to recalibrate their risk models (timeline: years), integration into automated underwriting systems (timeline: years), and institutional trust built through demonstrated performance across multiple economic cycles (timeline: decades). The algorithm is the easy part. The institutional infrastructure is the moat.

4. SCALE ECONOMICS: DOES GROWTH MAKE THIS BUSINESS BETTER OR JUST BIGGER?

Returns to Scale Assessment: INCREASING — dramatically.

The evidence is unambiguous. From 2016 to 2025:
- Revenue CAGR: 9.4% ($881M → $1,991M)
- Operating Profit CAGR: 20.7% ($170M → $925M)
- Net Income CAGR: 21.8% ($109M → $652M)
- FCF CAGR: 17.0% ($188M → $770M)

Operating profit has grown at more than twice the rate of revenue for nine consecutive years. This is textbook increasing returns to scale: each additional dollar of revenue drops to the bottom line at a dramatically higher rate than the average dollar because the cost base is largely fixed. The scoring algorithm costs the same whether it processes one million scores or one billion. The FICO Platform's cloud infrastructure has marginal costs that decline with scale.

The margin expansion trajectory makes this concrete:
- 2016: 19.2% operating margin → $0.19 operating profit per dollar of revenue
- 2025: 47.0% operating margin → $0.47 operating profit per dollar of revenue
- Q1 2026: 54.0% non-GAAP operating margin → $0.54 per dollar

At double the current revenue (~$4B), operating margins could theoretically approach 55-60%, because the incremental cost of delivering additional scores is near zero and platform software has high incremental margins once infrastructure costs are absorbed. The theoretical ceiling is determined by R&D investment levels (which management chooses) rather than by economic necessity.

4.5 CAPACITY UTILIZATION & EMBEDDED OPERATING LEVERAGE

FICO's business is fundamentally different from Vinall's Carvana example because it is an intellectual property licensing and software business with effectively unlimited capacity. The scoring algorithm can process any volume of queries on bureau infrastructure without FICO investing in additional capacity. The FICO Platform is cloud-hosted with elastic scaling. Capital expenditure of $8.9 million in fiscal 2025 — 0.4% of revenue — confirms that there is no meaningful physical capacity constraint.

The embedded operating leverage in FICO is not capacity-vs-utilization leverage (where infrastructure is built ahead of demand) but rather pricing-vs-cost leverage (where pricing power allows revenue to grow while the cost base grows modestly). Operating expenses grew approximately 4% quarter-over-quarter in Q1 2026 while revenue grew 16% year-over-year — the gap between revenue growth and cost growth flows directly to operating profit.

Capacity Utilization Ratio: Effectively Unlimited — MASSIVE LEVERAGE inherent in the algorithm-licensing model.

The correct way to understand FICO's embedded leverage is not "how many more scores can the infrastructure handle" (answer: infinite) but "how much more pricing power can be exercised before triggering regulatory backlash" (answer: uncertain, but significant headroom remains). Each percentage point of price increase on billions of scores generates tens of millions of incremental revenue at near-100% incremental margin.

5. WHERE DOES THE CASH GO?

FICO's cash flow dynamics reveal a business with extraordinary cash generation efficiency and a singular capital allocation philosophy: generate cash, borrow additional capital, and buy back shares with both.

Operating the Business:

Total operating expenses were $278 million in Q1 2026, with personnel costs as the largest component. Stock-based compensation of $157 million in fiscal 2025 (approximately $40 million per quarter) represents a significant non-cash expense — roughly 21% of reported FCF. Depreciation has declined from $36 million (2017) to $15 million (2025), reflecting the asset-light nature of the business and the absence of significant physical infrastructure. R&D investment is embedded within operating expenses but not separately disclosed at the segment level.

The business requires virtually zero maintenance capital expenditure. Fiscal 2025 CapEx of $8.9 million is less than what many companies spend on office furniture. This means that virtually all operating cash flow converts to free cash flow: $779 million OCF → $735 million FCF, a 94% conversion rate.

After the Bills Are Paid:

FICO's capital allocation strategy is aggressive and mono-dimensional: nearly all excess cash (plus significant borrowed capital) goes to share buybacks. The ten-year track record:

The pattern is striking: in most years, FICO spends significantly more on buybacks than it generates in free cash flow, funding the difference with debt. Cumulative buybacks from 2016-2025 total approximately $5.8 billion against cumulative FCF of approximately $3.9 billion — meaning $1.9 billion of buybacks were debt-funded. Total debt has grown from roughly $1 billion in 2021 to $3.46 billion in 2025, while stockholders' equity has gone deeply negative (-$1.75 billion) because the buybacks have eliminated the equity base entirely.

This is a leveraged recapitalization disguised as a capital return program. Management is betting — correctly so far — that the scoring monopoly's cash flows are so stable and growing so consistently that the business can sustain substantial leverage while returning cash to shareholders through buybacks. The weighted average interest rate of 5.22% on $3.2 billion of debt (Q1 2026) implies approximately $167 million in annual interest expense — meaningful but comfortably covered by $770 million in FCF (4.6x interest coverage).

The share count reduction has been dramatic: from 31 million shares (2016) to 24 million (2025), a 23% reduction that has amplified per-share metrics beyond underlying business growth. EPS grew at a 25.4% CAGR versus net income's 21.8% CAGR, with the 3.6 percentage-point difference attributable to share count reduction.

5.5 HOLDING COMPANY / CONGLOMERATE DISCOUNT ANALYSIS

Not applicable — FICO is a single operating business with two reportable segments (Scores and Software) that are managed as an integrated entity. There are no publicly traded subsidiary stakes, no holding company discount, and no sum-of-parts analysis required.

6. BUSINESS MODEL EVOLUTION & TRANSITIONS

Historical Transition: From Analytics Consulting to Algorithm Licensing to Software Platform

FICO's business has undergone two major transformations. In its early decades (1956-2000s), the company was primarily an analytics consulting firm that built custom scoring models for financial institutions. Revenue was project-based, cyclical, and labor-intensive. The first major transition came as the FICO Score became standardized and embedded in regulatory frameworks — the business shifted from selling custom consulting engagements to licensing a standardized algorithm distributed through credit bureaus. This shift dramatically improved economics: recurring royalty revenue replaced one-time project fees, marginal costs dropped to near zero, and the regulatory embedding created the monopoly position that exists today.

The second transition, currently underway, involves building the Software segment into a significant growth engine through the FICO Platform cloud migration. Legacy on-premises software (Falcon, TRIAD, Originations Manager) is being migrated to or replaced by the cloud-native FICO Platform. This transition mirrors the broader enterprise software industry's shift from perpetual licenses to SaaS subscriptions: near-term revenue headwinds (non-platform declining 13% year-over-year) offset by long-term improvements in recurring revenue visibility, customer lifetime value, and expansion economics. Platform ARR of $303 million growing at 33% suggests this transition is well past the trough and entering the acceleration phase.

A third transition is emerging within Scores: the shift from bureau-intermediated distribution to direct licensing through the DLP. This is not a business model change per se — the per-score royalty model remains — but it is a distribution and pricing architecture change that gives FICO greater control over its economic relationship with end customers. CFO Steve Weber's Q1 2026 commentary attributed 60% mortgage Scores revenue growth primarily to "higher mortgage origination Scores unit price," confirming that the DLP is a pricing power amplification mechanism.

CEO and Leadership:

Will Lansing has served as CEO since 2012, overseeing the entire modern era of FICO's margin expansion and strategic repositioning. Under his leadership, operating margins have expanded from approximately 22% to 47%, revenue has more than doubled, and the share count has been reduced by over 30%. Lansing's strategic philosophy is clearly focused on maximizing the economic value extracted from the scoring monopoly while building the software platform as a long-term diversification play. His tone on earnings calls is confident and direct — when the Wells Fargo analyst raised lender concerns about DLP liability, Lansing dismissed them as "misplaced, misguided" — suggesting a management team that is not apologetic about its pricing power and willing to exercise it aggressively.

CFO Steve Weber, in his role since 2021, has overseen the acceleration of the buyback program and the debt-funded capital return strategy. His Q1 2026 comment that management is "pretty confident we're going to be able to beat our guidance" while maintaining conservative published guidance reflects the disciplined under-promise-and-over-deliver pattern that has characterized FICO's investor communication.

6.5 VALUE LAYER DECOMPOSITION

Revenue Split Summary:
- Revenue from AI-RESILIENT layers (proprietary data + regulatory + transaction): ~70-75% of total
- Revenue from AI-VULNERABLE layers (data access + workflow logic + interface): ~25-30% of total

This composition is exceptionally favorable. The vast majority of FICO's revenue derives from value layers that AI cannot replicate or displace because the value is in regulatory acceptance and institutional entrenchment, not in the interface or workflow logic.

6.6 REVENUE MODEL AI RESILIENCE

Per-Score Royalty Model: FICO's scoring revenue is not per-seat or per-user — it is per-transaction, tied to the volume of credit decisions made across the economy. AI agents making automated lending decisions would increase, not decrease, the number of score queries. If AI enables faster, more automated credit decisioning, FICO benefits from higher scoring volumes. The revenue model is intrinsically aligned with automation and AI adoption in financial services.

Platform SaaS Model: Software revenue is subscription-based with usage components. While per-seat licensing risk exists for some on-premises legacy products, the platform's pricing is increasingly tied to transaction volumes and use-case deployment rather than human user counts. AI agents processing more decisions on FICO Platform would generate more usage-based revenue, not less.

Revenue Model Durability Verdict: RESILIENT. The per-score royalty model is consumption-based and benefits from any increase in automated credit decisioning. The platform's transition to usage-based pricing further insulates against AI-driven seat compression. No material revenue stream faces structural threat from AI agent substitution.

7. WHAT COULD GO WRONG?

Regulatory Intervention on Scoring Pricing: The most consequential risk. If FHFA, Congress, or CFPB imposes price caps on FICO Scores or mandates acceptance of multiple scoring alternatives with sufficient force, the pricing power that has driven 28 percentage points of margin expansion could reverse. FICO's scoring revenue is approximately 55% of total company revenue and likely contributes 70%+ of operating profit — pricing compression here would have outsized impact.

Leverage Risk in a Downturn: With $3.46 billion in debt, $55 million in cash, and negative stockholders' equity of -$1.75 billion, FICO is among the most aggressively leveraged companies in enterprise software. If mortgage origination volumes collapse in a severe recession (reducing scoring volumes) while interest rates spike (increasing debt service costs), the company's FCF could compress enough to create refinancing risk. The current 4.6x interest coverage is adequate but provides less cushion than a business of this quality should carry.

Software Platform Execution: The Software segment must successfully transition 150+ customers from legacy products to FICO Platform while simultaneously winning new business. If platform growth decelerates or non-platform churn accelerates beyond expectations, the software segment could stagnate — leaving the company over-dependent on scoring revenue and vulnerable to the regulatory risks described above.

Munger's Inversion — How This Business Dies:
1. Political scenario: A populist Congress investigates FICO's pricing power, holds hearings on the cost of credit scoring to consumers, and legislates mandated acceptance of multiple scoring alternatives at regulated prices. Within 5 years, FICO's scoring margins compress from 90%+ to 50%, eliminating $400M+ in annual operating profit.
2. Secular scenario: The GSEs are reformed or privatized, new conforming mortgage standards no longer mandate specific scoring products, and lenders gradually diversify to cheaper alternatives over a decade. Slow erosion rather than sudden collapse.

BUSINESS MODEL VERDICT

In One Sentence: FICO earns a royalty every time any American's creditworthiness is evaluated — an event that happens billions of times annually — while selling the decision management software that acts on those evaluations.

Overall: This is a "wonderful business" by any Buffett/Munger definition — an unassailable competitive position, near-zero capital requirements, extraordinary cash generation, and demonstrated pricing power that has widened the moat for over a decade. The lone reservation is capital allocation: management's decision to lever the balance sheet to -$1.75 billion in equity to fund buybacks converts a business with fortress-quality economics into one with meaningful financial risk. Whether this is brilliant optimization or imprudent leverage depends entirely on whether the scoring monopoly's cash flows are as indestructible as management believes.

Understanding how the business makes money — the toll-booth scoring model, the expanding software platform, the near-zero capital requirements — the next question is whether the financial statements confirm the story across time. Do the margins, returns on capital, and cash flow trajectories reflect the pricing power, scale advantages, and customer lock-in we have described? And does the balance sheet's aggressive leverage enhance or endanger the long-term compounding of intrinsic value? That is precisely what the numbers will reveal.