The Financial Terminal That Started on a Bicycle

In the late 1970s, a Wall Street analyst could ask FactSet for a company report and receive four pages by bicycle messenger. The packet held fundamentals, ratios and the other figures required to understand a business. Howard Wille, then a director of research, designed the overview. His colleague Charles “Chuck” Snyder licensed data and assembled it on a shared mainframe. They called the compact result a “fact set.” It was a plain name for a product with a durable insight: financial professionals did not need more loose information. They needed the right information arranged for a decision.

Wille and Snyder founded FactSet in 1978. Nearly half a century later, the company is no longer moving paper around Manhattan. It is moving licensed data, proprietary content, client data, calculations and instructions through an enterprise platform. FactSet reported 247,766 users and 9,130 clients as of May 31, 2026. Its fiscal 2025 revenue was $2.32 billion, the latest entry in a run of more than 45 consecutive years of annual revenue growth.

01 / The jobMaking the market usable

A modern investment firm has the opposite of an information shortage. Prices arrive in real time. Filings, transcripts and broker research accumulate. Portfolio positions live in internal systems. Risk models, benchmarks, alternative datasets and compliance rules come from different places, with different identifiers and permissions. The hard problem is connection. An analyst may find an idea, test it, place it in a portfolio, explain the resulting risk, send an order and report the outcome. Every handoff can invite a new screen, export or reconciliation.

FactSet sells a way to reduce those seams. Its Workstation connects more than 800 proprietary, third-party and client data sources in a common environment. Research teams can screen companies and monitor news. Portfolio managers can model trades. Risk teams can inspect exposures and run scenarios. Performance teams can attribute returns. Wealth advisers can prepare meetings and proposals. Developers can take data out through feeds and APIs or embed FactSet widgets in their own applications.

The enduring product is not the screen. It is the distance removed between a question and a defensible answer.

That distinction explains why FactSet is more than a financial terminal, though the terminal remains an important doorway. The company increasingly describes itself as a financial digital platform and enterprise solutions provider. The language is less vivid than a glowing desktop, but more accurate. A client may encounter FactSet in Excel, a portfolio report, a cloud warehouse, an adviser portal, an order-management system or a custom application without spending the day in one branded window.

02 / The evolutionFrom Visicalc to agents

FactSet's history reads like a tour of office computing. In 1981, Snyder found a way to download company data into Visicalc, the early personal-computer spreadsheet. Screening followed. In 1989, the Private Database Service let customers combine their own information with FactSet's - decades before “bring your own data” became standard enterprise language. FactSet for Windows arrived in 1990. The company went public in 1996 and added 24-hour live telephone support in 1999.

The support milestone matters. Financial information is perishable, and a model that breaks before a client meeting is not an abstract software issue. FactSet made service part of the product. Its client-value retention remains visible in annual subscription value, or ASV, where retention exceeded 95 percent in the third quarter of fiscal 2026. Once data and calculations become woven through research templates, portfolio reports, compliance checks and client communications, replacing the vendor becomes a large operational project.

In 2009, the company combined several platforms into one FactSet product. More recently it has expanded in both directions. Below the visible workflow sit identifiers and reference data. FactSet's $1.925 billion purchase of CUSIP Global Services in 2022 placed it closer to that market plumbing. Above the data sit decisions and execution. The $246.5 million acquisition of LiquidityBook in 2025 added cloud-based order management, pre-trade compliance and an investment book of record, extending FactSet from portfolio construction toward the actual order.

03 / The customerA professional tool, on purpose

FactSet's customers include asset managers, asset owners, hedge funds, banks, insurers, corporations, private-equity firms and wealth managers. They are not one audience. An equity analyst wants estimates and transcripts. A quantitative researcher wants clean history and programmatic delivery. A chief investment officer wants a whole-portfolio view. An adviser wants a useful reason to call a client before the client calls first.

The buy side remains the center of gravity: it represented roughly 82 percent of organic ASV in fiscal 2025. Yet wealth has become an important expansion area. FactSet offers adviser dashboards, prospect monitoring, model portfolios, compliant proposal generation and client reporting. Its pitch is consolidation. A wealth firm can connect institutional analytics to its CRM and portal instead of asking advisers to swivel among disconnected tools.

The commercial model is correspondingly durable. Clients subscribe to combinations of content, applications, seats and enterprise delivery, with additional revenue from feeds, APIs, managed services, professional services and issuer-related products. FactSet reported organic ASV of $2.486 billion in May 2026. In practical terms, it sells access to the platform and then grows by adding users, datasets and adjacent workflows.

04 / The contestNot the keyboard company

The unavoidable comparison is Bloomberg, whose terminal, news operation and messaging network make it a fixture of trading floors. LSEG's data and analytics business and S&P Capital IQ Pro are the other broad rivals; Morningstar, MSCI, SimCorp, BlackRock Aladdin, ICE and specialists compete in narrower lanes. There is overlap, but these are not identical products. A banker's preferred research desktop may not be the best operating layer for an asset manager's portfolios.

FactSet's answer is configurability and openness. It integrates client-owned information rather than insisting that every useful fact originate inside its walls. It offers an enterprise SDK and more than 1,100 datasets for quantitative workflows. Its portfolio analytics cover performance, attribution, exposures, risk and scenarios across asset classes. Its Workstation connects research with Microsoft Office, internal tools and downstream systems. High-touch support accompanies the software.

This is not a cheap or casual product, and it is not intended to be. FactSet earns its place when the cost of a mismatched security, stale report, unexplained return or broken handoff is higher than the subscription. Its differentiation is less a dazzling single feature than a collection of reduced annoyances, governed calculations and connections that work every morning.

05 / The next interfaceAI with receipts

Generative AI creates a peculiar challenge for financial data companies. A model can summarize a transcript or draft portfolio commentary in seconds, but plausible language is not enough for a regulated institution. The source must be licensed. The calculation must be repeatable. Permissions must hold. An analyst needs to inspect why the machine reached its answer.

FactSet is betting that its existing data and workflow position can make AI more useful. FactSet Intelligence includes conversational search, transcript analysis, portfolio commentary and Pitch Creator for investment-banking presentations. Its Model Context Protocol tools let compatible AI systems interact with financial data and analytics. In the third quarter of fiscal 2026, the company said more than 90 percent of its top 50 clients were using four or more FactSet AI products.

Partnerships broaden that bet. Google Cloud is bringing infrastructure, enterprise search and Gemini models to a new generation of financial agents. TIFIN.AI is working with FactSet on adviser workflows. J.P. Morgan's Fusion platform supplies normalized fund-accounting information for whole-portfolio analytics. Valutico links private-market data to more frequent valuation work. Each arrangement follows the same old pattern: put trusted information nearer to the decision, then remove manual steps around it.

The bicycle was never the business. Delivery was.

There is risk in the transition. AI can lower the effort required to search and summarize, inviting new competitors and pressing established vendors to prove the value of their data. Technology spending can weigh on margins. Clients may prefer models that operate across vendors rather than within one platform. FactSet's defense is that reliable financial work requires more than a language model: it requires identifiers, entitlements, historical data, analytical engines, client context and a path into production.

That makes FactSet an instructive company for builders outside finance. Start with a recurring professional decision. Gather fragmented inputs. Let customers mix in what only they know. Meet them inside familiar tools. Add human service where mistakes are expensive. Then expand to the next decision without breaking the first. The surface can move from paper to spreadsheet to agent. The job remains recognizable.

Forty-eight years after its founding, FactSet still packages facts. The package is now a sprawling system of data, software and service, and the messenger is a network connection. Somewhere beneath the AI agents and risk models is the same four-page discipline: not every number, just the connected set that helps someone act.