1. M A R C H 2013Big data in theage of the telegraphCaitlin RosenthalDaniel McCallum’s 1854 organizational design for the New York andErie Railroad resembles a tree rather than a pyramid. It empowered frontlinemanagers by clarifying data flows.In 1854, Daniel McCallum took charge ‘Big data,’ then and nowof the operations of the New York andErie Railroad. With nearly 500 miles Just as information now floods into com-of track, it was one of the world’s longest panies by the tera-, peta-, and exabyte,systems, but not one of the most during the mid-19th century, governments,efficient. In fact, McCallum found that businesses, and universities producedfar from rendering operations more and grappled with what one historian hasefficient, the scale of the railroad expo- called an “avalanche of numbers.” 3nentially increased its complexity.1 To be sure, McCallum’s rail lines may not have generated even a megabyte of The problem was not a lack of infor- information. But this was indeed big data mation: the growing use of the telegraph for him and his senior deputies, who gave the company an unprecedented were managing a system of unprece- supply of nearly real-time data, including dented proportions. Although the reports of accidents and train delays.2 telegraph’s speed made more informa- Rather, the difficulty was putting that data tion available, organizing and acting to use, and it led McCallum to develop on it became increasingly difficult. One one of the era’s great low-tech manage- delayed train, for example, could dis- ment innovations: the organization rupt the progress of many others. And chart. This article presents that long-lost the stakes were high: with engines chart (see exhibit, “The first modern pulling cars in both directions along a organization chart”; and sidebar, single set of rails, schedule changes“Tracking a missing org chart”) and shows risked the deadly crashes that plagued how aligning data with operations 19th-century railroads. and strategy—the quintessential modern management challenge—is a problem that spans the ages. (continued on page 4)
2. 2 2 1Plan of Organization,New York and Erie Railroad,1855Courtesy of the Geographyand Map Division,Library of Congress.
3. 3ExhibitThe first modern organization chartDaniel McCallum created the first organization chart in response to the information problemhobbling one of the longest railroads in the world. In surprising contrast to today’s top-downorganization pyramids, in McCallum’s chart the hierarchy was reversed: authority over day-to-dayscheduling and operations went to the divisional superintendents down the line, who oversawthe five branch lines of the railroad. The reasoning: they possessed the best operating data, werecloser to the action, and thus were best placed to manage the line’s persistent inefficiencies.1Each superintendent was responsible for the physical geography of the tracks and stationsand for the men who moved along the rails: conductors, brakemen, and laborers. Coordinatingactivities between these two branches, the superintendents managed both the fixed depotsand the rolling stock that moved between them.2
4. 4As McCallum reflected, “A super- the tree. The first was a straight branchintendent of a road 50 miles [long] . . . representing the physical assets ofmay be almost constantly upon the line tracks and stations, the secondengaged in the direction of its details.” a winding branch consisting of the menBut on railroads like his, which stretched who moved along the rails, from thefor hundreds of miles, no individual conductors and brakemen on the trainsmanager could be responsible for all of to the laborers who maintained thethe necessary schedule changes. tracks. The divisional superintendents were responsible for coordinating these two branches—the depots and theReversing the information rolling stock, and the employees whohierarchy moved between them.In crafting the organizational plan, Even as McCallum decentralized decisionMcCallum sought to improve the way the making along the railroad, he alsorailroad used information. Through insisted that targeted metrics had to be21st-century eyes, the chart looks both reported back to its board of directors.antiquarian and surprisingly modern. That data flowed down from theFar from the static, hierarchical pyramids branches of the tree to its roots, wherethat we today associate with such McCallum and the board could usecharts, his was modeled after a tree. the information for oversight and long-McCallum drew the board of directors term decision making. Here, McCallum’sas the roots, himself and his chief goal was prioritization: assuringofficers as the tree’s trunk, and the rail- that the board, with its finite capacity,road’s divisions and departments as received relevant and actionablethe branches. data. As “interesting as this information is,” he reflected, it is only in its “practicalCritically, McCallum gained control by application . . . that its real valuegiving up control, delegating authority to consists.” McCallum therefore designedmanagers who could use information a system of hourly, daily, and monthlyin real time. He put what we would call reports that enabled him to calculatethe organization’s C-level at the ground practical metrics, such as cost perlevel, supporting the railroad, not direct- ton-mile and average load per car. Bying its operations. Following one comparing the profitability and efficiencyof McCallum’s key precepts—“a proper of different routes, the board coulddivision of responsibilities”—authority identify opportunities for improvement.over day-to-day scheduling went to thedivisional superintendents down the line. A message for today’s leaders?Most of the chart spans the domainsof these superintendents: the railroad’s Modern managers, of course, have morefive branch lines. Each superintendent sophisticated tools than McCallum did.was responsible for two subbranches of Top executives can now access detailed
5. 5Tracking a missing org chartFormer Harvard Business School professor at St. Lawrence University, in upstate New York.2Alfred Chandler (1918–2007), who helped estab- The discovery came in time for the chart to belish business history as a rigorous academic included in my doctoral dissertation, completingdiscipline, described the momentous impact of a quest Chandler began when researchingmanagerial innovations, such as the organi- his own dissertation on the pioneering financialzation chart. Chandler identified Daniel McCallum analyst Henry Varnum Poor.as the originator of the New York and Erie’spioneering plan and describes the chart in tan- Chandler continued writing about McCallum andtalizing detail in several of his books. I began the chart in many foundational business-searching for it during my doctoral studies at history studies, including his Pulitzer Prize winner,Harvard, writing to archives in New York and The Visible Hand: The Managerial RevolutionOhio and even combing through Chandler’s per- in American Business (Harvard Universitysonal papers. In the course of my search, Press, 1977), as well as Strategy and Structure:I learned that Chandler himself had never seen Chapters in the History of American Industrialthe chart and based his description on a Enterprise (MIT Press, 1962), which is widelydetailed advertisement in the American Railroad viewed as a seminal book in the developmentJournal.1 not just of business history but also of strategic thought.I was almost ready to give up on my search 1 handler explains that he had not seen the chart himself Cwhen the unexpected happened at an academic (at least as of 1988), in Alfred Chandler, “Origins ofconference: Peter Knight, a professor of the organization chart,” Harvard Business Review, 1988,American studies at the University of Manchester, Volume 66, Number 2, pp. 156−57.handed out a series of images on the history 2 n addition, Charles Wrege and Guidon Sorbo Jr. located Iof capitalism, and I immediately recognized that the chart and discuss it in “A bridge builder changesone was the missing organization chart! I was a railroad: The story of Daniel Craig McCallum,” Canalastonished to learn that Peter had found it History and Technology Proceedings, 2005, Volume 24, pp. 183–218.in the Library of Congress. With the help of itsreference librarians, I located another copydata, often in real time. Today’s powerful leaders report that they are overwhelmedbattlefield-management systems, for by copious data flows.4 A more fruitfulinstance, give generals the ability to direct approach might begin with McCallum’scombat missions at the microlevel, low-tech reflections on organizationalwhich was invisible to the board of the structures and priorities. Within today’sNew York and Erie. organizations, emerging social networks— often married to sources of big data—But while cheaper storage space and have a certain kinship with McCallum’sbandwidth will make such granular man- logic. These networks too provideagement options increasingly tempting, opportunities for greater information shar-will they be optimal? Executive attention ing and collaboration at relatively lowspans are already stressed, and many levels in the organization, and they too
6. 6 3 he phrase “avalanche of numbers” comes can improve operations, customer T from Ian Hacking, writing on the spread of service, and innovation. Curiously, digital probabilistic and statistical reasoning, in mappings of these social interactions The Taming of Chance, first edition, Cambridge, bear a resemblance to the nodes and UK: Cambridge University Press, 1990. 4 teve LaValle et al., “Big data, analytics and the branches of McCallum’s chart. S path from insights to value,” MIT Sloan Management Review, 2011, Volume 52, Number 2, Drowning in the details of operations, pp. 21–32. Daniel McCallum stepped back and redesigned the railroad’s organization. The author wishes to acknowledge Michael His insights on how to meld local Chui for his contribution to this article. authority with information gave his man- Caitlin Rosenthal, an alumnus of agers better operating tools—which McKinsey’s Houston office, is the are just as relevant in the age of Harvard–Newcomen postdoctoral fellow the Internet as they were in the age at the Harvard Business School. of the telegraph. Copyright © 2013 McKinsey Company. All1 his article’s details on the railway’s operations T rights reserved. We welcome your comments and organizational thought come from Homer on this article. Please send them to quarterly_ Ramsdell and D. C. McCallum, Reports of the email@example.com. President and Superintendent of the New York and Erie Railroad to the Stockholders, for the Year Ending September 30, 1855, New York, NY: Press of the New York and Erie Railroad Company, 1856.2 om Standage quotes contemporaries who called T the telegraph the “highway of thought” in The Victorian Internet: The Remarkable Story of the Telegraph and the Nineteenth Century’s On-line Pioneers, first edition, London, UK: Weidenfeld Nicolson, 1998. An excellent recent account of the telegraph’s impact is Richard John, Network Nation: Inventing American Telecommunications, first edition, Cambridge, MA: Harvard University Press, 2010.