Gunjo · Business Intelligence for the AI Era
← Sticker Wall JOURNEY · DETAIL

Michael Bloomberg / Bloomberg L.P. — Leveraging terminal data moats to build financial media and news influence

Founded: Michael Bloomberg · Bloomberg L.P.

JOURNEY

Key Fields

FIELD STAMPS
IndustryFintech
RegionUS
ScaleGiant
ChannelB2B

Origin

In 1981, Michael Bloomberg was laid off from Salomon Brothers and received a $10 million severance package. He noticed that Wall Street traders lacked a real-time system that integrated bond data, news, and analytical tools, so he decided to use the funds to start a company focused on financial data services. Having previously managed equity trading systems at Salomon, Bloomberg understood that the trading desk's need for instant, accurate data was far from being met.

Milestones

1981
Inception Failure
Michael Bloomberg was laid off from Salomon Brothers and received a $10 million severance package. He initially aimed to build a multi-functional terminal for financial institutions, but the first version only provided bond pricing and yield data, lacking news and analytical features. Bloomberg later recounted that the product was unsellable, as financial institutions felt it was no better than Reuters terminals, resulting in almost no actual orders in the first 6 months.
1982
Early Product Turning Point
Bloomberg approached Merrill Lynch, which agreed to invest $30 million for a 30% stake, becoming Bloomberg's first and largest institutional client. Merrill required the terminal to integrate real-time bond prices, calculators, and news, forcing Bloomberg to upgrade the product from a single data query tool to a comprehensive trading workstation. Merrill's initial order of 22 terminals marked Bloomberg's transition from concept to commercial reality.
1985
Expansion Growth
Bloomberg Terminal installations surpassed 5,000 units, with annual revenue exceeding $100 million for the first time. Bloomberg consistently reinvested terminal revenue into news gathering; after Bloomberg News was founded in 1984, it grew rapidly to 300 reporters. The high-margin subscription revenue from terminals provided stable cash flow for media expansion, establishing a model of software subsidizing news.
1990
Media Pivot
Bloomberg acquired a small New York radio station and renamed it Bloomberg Radio, entering the broadcast media space. That same year, Bloomberg News began providing content to newspapers, and the terminal-plus-news closed loop gradually took shape. Although the news business itself was not profitable, it significantly increased terminal stickiness for traders and fund managers; competitor Reuters could not match Bloomberg's terminal user scale in the media segment.
2001
Politics Pivot
Michael Bloomberg was elected Mayor of New York City and stepped down as CEO of Bloomberg, moving the company into a professional management phase. Peter Grauer took over, leading global market expansion and terminal penetration. While Bloomberg emphasized independence in politics and Bloomberg News maintained its style of investing heavily in exclusives and data accumulation, the media division's long-term losses were subsidized by terminal cash flow.
2008
M&A Growth
Bloomberg acquired BusinessWeek magazine, renaming it Bloomberg Businessweek, to strengthen its in-depth business reporting capabilities. The number of Bloomberg News reporters globally exceeded 2,300, with over 5,000 stories published daily. Annual terminal revenue exceeded $7 billion, and global terminal installations surpassed 300,000, making it the only giant in the financial information field capable of competing comprehensively with Reuters.
2023
AI Transformation Turning Point
Bloomberg released BloombergGPT, which built a financial training dataset of 363 billion tokens and a public dataset of 345 billion tokens based on its 40 years of accumulated financial data, training a 50-billion parameter financial large model. Bloomberg explicitly stated that due to concerns over data leakage, it would not directly connect private terminal data to external AI models, but would instead integrate AI capabilities into terminal workflows.

Turning Points

  • Starting a business with severance pay after being laid off by Salomon in 1981, turning a setback into the foundation of a financial information empire.
  • Merrill Lynch's $30 million investment and initial order of 22 terminals in 1982, which defined the business direction of a terminal-plus-comprehensive-workstation.
  • The decision to continuously invest terminal subscription cash flow into news gathering, using news stickiness to fuel terminal sales, creating a dual-engine drive of data and media.
  • The 2023 release of BloombergGPT, marking the company's transition from a data pipeline provider to an AI-enhanced financial workstation.
  • Bloomberg's three terms as Mayor of New York City led to the de-founderization of corporate governance, yet the core values and news independence have remained locked in place.

Failures & Pitfalls

  • Early products provided only single bond data without news or analysis, resulting in almost no sales for the first 6 months.
  • Bloomberg News has long operated at a loss, relying on high-margin terminal revenue for subsidies, unable to achieve independent profitability.
  • When facing the impact of AI generative products, BloombergGPT cannot open up proprietary terminal data due to data security concerns, limiting the scale of training data.
  • Reuters, investors under Masayoshi Son, and various crypto-data projects have attempted to reduce reliance on Bloomberg Terminals using blockchain and open data.

关键成功要素

  • Terminal subscriptions are the core cash cow, with annual subscription fees per unit exceeding $20,000 and over 300,000 global installations.
  • The partnership and 30% equity investment from Merrill Lynch allowed Bloomberg to lock in top-tier institutional clients from the start.
  • Using high-margin terminal cash flow to continuously subsidize news gathering, acquiring Businessweek and broadcast assets to expand influence.
  • Data moats and compliance frameworks are key to why the terminal is difficult to replace by AI; the depth of private data far exceeds that of public large models.
  • Founder Michael Bloomberg's political capital and personal influence have provided long-term endorsement for the brand.

Lessons

  • Being laid off is not necessarily a bad thing; having industry insight and startup capital allows one to focus on solving overlooked needs.
  • Deep-pocketed institutional clients can often drive product iteration; a product definition brought by a client like Merrill is more important than 100 ordinary clients.
  • Businesses with strong cash flow can invest in peripheral content that may not be profitable but increases customer stickiness, forming a moat.
  • The financial information industry has extremely high data security and compliance requirements; any seemingly convenient open solution can become a fatal vulnerability.
  • A founder's personal brand can serve as a sales tool, but if they are absent for a long time, the company must have a stable product value proposition to maintain growth.

Core Data

  • 2023 global Bloomberg Terminal installations：Over 300,000 units
  • Annual subscription fee per terminal：Approximately $20,000
  • 2023 annual Bloomberg Terminal revenue：Over $10 billion
  • BloombergGPT financial training dataset tokens：363 billion tokens
  • BloombergGPT parameter scale：50 billion parameters
  • 1982 Merrill Lynch investment：$30 million for a 30% stake
  • 2008 global Bloomberg reporter count：Over 2,300
  • Bloomberg News daily story volume：Over 5,000

Competitors / Peers

Bloomberg's direct competitors in the financial information terminal field are Refinitiv (formerly the Financial & Risk division of Thomson Reuters, later acquired by the London Stock Exchange) and S&P Global Market Intelligence. Both provide real-time quotes, news, and analytical terminals, but cannot fully replicate Bloomberg's terminal-plus-media model in terms of news independence and data granularity. Emerging challengers include crypto-data products that use blockchain and open data protocols to reduce data intermediary costs, as well as startups using open-source AI models to generate research reports, though they have not yet shaken Bloomberg's institutional client base.

Sources

Related

Same industry / scale
FintechUS

Assurant: The Hidden Mobile Protection Giant That Started with Funeral Insurance

FintechUS

Berkshire Hathaway: Building a Century-Long Compounding Machine with Insurance Float

FintechUS

Block: A Fintech Giant Built on Mobile Card Readers, Expanding into Cash App and Bitcoin

FintechUS

Ramp: A fast-growing fintech SaaS company disrupting corporate credit cards and expense management

⚠️ This entry draws on self-reported accounts and public reporting, not independently verified, and is not investment, startup, or legal advice; revenue figures are case-specific and not representative of typical results.