SC

Bloomberg

Bloomberg gave bond traders usable calculations, then made data, news and conversations part of the same daily habit.

What shaped Bloomberg

  1. 1 · 1982 to 1989Start with the calculation a trader needsMichael Bloomberg had run information systems at Salomon Brothers. BL2 BL3Founder Domain Expertise
  2. 2 · 1990 to 2020Make the terminal useful throughout the working dayBloomberg added its own newsroom in 1990 after concern over continued access to Dow Jones news. BL2 BL3Collaboration & exchange network effects
  3. 3 · 1982 to 2025Spread the collection cost while continuing to earn the seatBloomberg sold continuing access to an integrated service and reused its data and news across subscribers. BL2 BL3

Arena: Market Conditions Before Bloomberg

Professional financial information · United States; later global · Bond traders and investment analysts · 1982

Specialist requirements

Bond desks collected prices from dealers and maintained calculations themselves. A quote alone did not show how a bond would behave under different interest-rate assumptions. Existing screens supplied prices, but traders still needed historical data and analytical work to compare securities. BL2 BL3

How each step happened

Step 1 of 3 · 1982 to 1989

Start with the calculation a trader needs

Michael Bloomberg had run information systems at Salomon Brothers. His team built a terminal that combined bond prices with calculations and historical data. Merrill Lynch accepted the working system, installed 20 terminals and invested in the company. Gerben Bakker’s history identifies an important product difference: Telerate could display prices but could not supply the same yield-curve analysis. BL2 BL3

A trader could examine a proposed trade while prices were changing, instead of assembling the inputs and calculations separately. The first sale therefore had a concrete analytical job behind it. Merrill’s backing also reduced the credibility problem facing a new supplier; its investment and its purchase were different contributions.

The early account is retrospective. The reported 20-terminal order should not be mixed with later accounts of 22 terminals delivered.

Rivals Telerate and in-house analysis are the relevant early alternatives. Merrill’s installation establishes adoption; Bakker’s historical comparison explains the functional difference. It does not isolate how much of the buying decision came from personal relationships.

Founder Domain Expertise

Step 2 of 3 · 1990 to 2020

Make the terminal useful throughout the working day

Bloomberg added its own newsroom in 1990 after concern over continued access to Dow Jones news. It expanded beyond bonds and supplied stories alongside the data and calculations they explained. WIRED’s 1999 reporting describes a customer using incoming messages throughout the day. Bakker records roughly 320,000 installed terminals by 2020. BL2 BL3

News and communication increased the number of tasks a subscriber could complete without leaving the terminal. This is the complementarity emphasized in Bakker’s history: a news operation supported the paid data business. As colleagues and counterparties used messaging, the destination of a conversation became another reason to retain access. Replacing a terminal could require changing both analytical routines and communication habits.

Messaging supports a qualified communication-network interpretation. The record does not measure how many subscriptions would survive on chat alone or the cost of migrating each customer.

Rivals Reuters also combined news and data and had its own communication network. The defensible claim is a useful installed workflow, not that Bloomberg invented every component or uniquely possessed a network.

Collaboration & exchange network effects

Step 3 of 3 · 1982 to 2025

Spread the collection cost while continuing to earn the seat

Bloomberg sold continuing access to an integrated service and reused its data and news across subscribers. Bakker links concentration in financial information to large sunk investments and discusses the failures of Quotron and other entrants. Merrill sold its remaining 20% interest for about $4.4 billion in 2008; that was a minority-stake transaction, not a sale of the whole company. BL2 BL3 BL5

A new provider must fund enough coverage and functionality before a trading desk will rely on it. An incumbent can spread that work across existing subscribers and add functions to a familiar environment. These cost and switching mechanisms reinforce each other, but a desk with narrower needs may still choose a specialist or reduce seats when its budget contracts.

The historical stake sale demonstrates substantial value. It does not establish a current whole-company exit or a current independently verified category-leader valuation under the outcome policy.

Rivals Reuters is evidence that another large network can sustain comparable investment. Specialist services compete for portions of the workflow; terminal count alone cannot establish their inability to replace particular tasks.

Key dates

  1. 1982Merrill Lynch installs the working system and becomes an investor. BL2 BL2
  2. 1990Bloomberg creates a news service under Matthew Winkler. BL2 BL2
  3. 1999WIRED reports a customer using the terminal for messages from overseas colleagues. BL2 BL2
  4. 2008-07-17Merrill sells its 20% Bloomberg interest for about $4.4 billion. BL5 BL5
  5. 2020Bakker records roughly 320,000 terminals worldwide by 2020. BL3 BL3

Sources

Oldest first.

  1. BL2 Michael Bloomberg’s Terminal Velocity. WIRED · 1999-02-01 Participant interview and reporting
  2. BL5 Merrill Lynch 2008 annual report. Merrill Lynch · 2009-02-24 Primary disclosure
  3. BL3 The terminal revolution: Reuters and Bloomberg as global providers of financial and economic news, 1960-2020. Gerben Bakker, LSE · 2025-09 Economic history research