Based on the frameworks offered in the Blenko et al. Harvard Business Review article and given your current or past history in an organization, how receptive would your organization's culture be to the tenets of a "decision audit"? How do you think the organization would score on the test questions shown on pages 58-59? Why?

SPOTLIGHT ON STRATEGIES FOR A CHANGING WORLD Many CEOs assume that organizational structure-the boxes and lines on a company's org chart-is a key determinant of financial performance. Like generals, they see their job as putting the right collection of troops in the right places. If the battle is about innovation, for example, then the CEO's duty is to create the best possible structure for channeling resources towards innovation. This belief helps explain why reorganizations are Instead of beginning with an analysis of strengths, so popular with chief executives. In fact, nearly half weaknesses, opportunities, and threats, structural of all CEOs launch a reorg during their first two years changes need to start with what we call a decision on the job. Some preside over repeated restructure audit. The goals of the audit are to understand the ings. The immediate motives vary. Some are about set of decisions that are critical to the success of your cutting costs; others are about promoting growth. company's strategy and to determine the organiza- Some are about shaking up a culture; others are tional level at which those decisions should be made about shifting strategic focus. Whatever the specifics, and executed to create the most value. If you can though, reorgs almost always involve making major align your organization's structure with its decisions, structural changes in pursuit of better performance. then the structure will work better, and your com- Despite the fanfare that usually greets them, pany's performance will improve. however, most reorganizations fall flat. A recent In this article we set out the basic principles for Bain & Company study of 57 reorgs between 2000 reorganizing around decisions. Let's begin by tak- and 2006 found that fewer than one-third produced ing a closer look at the link between decisions and any meaningful improvement in performance. Most performance. had no effect, and some actually destroyed value. Chrysler, for instance, reorganized its operations What Drives Your Performance? three times in the three years preceding its bank- Organizational structure is not the only determinant ruptcy and eventual combination with Fiat. Each of performance. In some cases, it is not even particu- ime, executives proclaimed that the company was larly important. That's why changing a company's on a new path to profitability. Each time, perfor- structure to meet a particular strategic goal can ac- mance didn't improve. tually exacerbate problems rather than help solve We believe that this failure is rooted in a profound them. For example, an organization struggling to misunderstanding about the link between structure innovate may try to gather more and more creative and performance. Contrary to popular belief, perfor- input-and end up getting too many people involved, mance is not determined solely by the nature, scale, thereby slowing the pace of decision making and sti- and disposition of resources, important though fling innovation. they may be. An army's success depends at least as Take the case of Yahoo. In December 2006, then- much on the quality of the decisions its officers and CEO Terry Semel announced a sweeping reorgani- soldiers make and execute on the ground as it does zation of the company, replacing Yahoo's product- on actual fighting power. A corporation's structure, aligned structure with one focused on users and similarly, will produce better performance if and advertiser customers. Seven product units were only if it improves the organization's ability to make merged into a group called Audience and another and execute key decisions better and faster than seven moved into a group called Advertisers and competitors. It may be that the strategic priority for Publishers. A unit dubbed Technology would pro- your company is to become more innovative. In that vide infrastructure for the two new operating groups. case, the reorganization challenge is to structure the The idea was to accelerate growth by exploiting company so that its leaders can make decisions that economies of scope across Yahoo's rich collection of produce more and better innovation over time. audience and advertiser products. Semel's team had For most companies, this requires a fundamen thought they'd carefully defined roles and responsi- tal rethinking of their approach to reorganization. bilities under the new structure, but decision making 56 Harvard Business Review June 2010THE DECISION-DRIVEN ORGANIZATION HBR.ORG Idea in Brief Reorganizations are popular In reality, a company's To reorganize around deci- of your organizational system with chief executives, who structure results in better sions, focus on six steps. First, to support decision making believe that making big performance only if it improves be clear about which decisions and execution. And sixth, structural changes will lead the organization's ability are most important. Second, equip your managers to make to better performance. But a to make and execute key figure out where in the orga- decisions quickly and well. Bain & Company study of 57 decisions better and faster nization those decisions need ions found that than competitors. If you to be made. Third, organize most reorgs had no effect-and can sync your organization's your structure around sources some actually destroyed value. structure with its decisions, of value. Fourth, figure out the then the structure will work level of authority your decision better and performance will makers need, and give it to improve. them. Fifth, adjust other parts and execution quickly became bogged down. Audi- higher for every country, industry, and company ence demanded tailored solutions that Technology size in our sample. Indeed, the companies in our could not provide at a reasonable cost. Advertisers sample that were most effective at decision making and Publishers needed its own set of unique prod- and execution generated average total shareholder uacts and so was constantly competing with Audience returns nearly six percentage points higher than for scarce developer time. In response, Yahoo execu- those of other firms. We also found that many com tives created new roles and management levels to panies have enormous scope to improve their perfor- coordinate the units. The organization ballooned to mance. Top-quintile companies score an average of 12 layers, product development slowed as decisions of 100 in decision effectiveness, while compa- stalled, and overhead costs increased. nies in the other four quintiles score, on average, 30 Yahoo's experience shows how a lack of atten- and below. This means that the typical organization tion to the decision-making process can thwart the has the potential to more than double its decision best-intentioned reorganization and undermine per- effectiveness formance. Ultimately, a company's value is no more What's more, the research revealed no strong sta- (and no less) than the sum of the decisions it makes tistical relationship between structure and perfor- and executes. Its assets, capabilities, and structure mance. Survey respondents' views about the struc are useless unless executives and managers through- ture of their company were not an accurate predictor but the organization make the essential decisions of either decision effectiveness or financial results. and get those decisions right more often than not. The conclusion we draw is simple: In a reorgani- Our research and experience confirm the tight zation, decisions rather than structure should be the link between performance and decisions. In 2008, primary focus. Let's see what that involves. we and our colleagues at Bain & Company surveyed executives worldwide from 760 companies, most Conducting a Decision Audit with revenues exceeding $1 billion, to understand Many reorganizations begin with a SWOT analysis: how effective those companies were at making What are our organization's strengths, weaknesses, and executing their critical decisions. We used the opportunities, and threats? What are our resources responses to assess decision quality (whether deci- and capabilities? What risks do we face? The idea is sions proved to be right more often than not), speed to determine if the company has everything it needs (whether decisions were made faster or slower than to support its strategy. All this sounds sensible, and competitors), yield (how well decisions were trans- in many ways it is. But the risk is that you'll end up lated into action), and effort (the time, trouble, and with an organization that's misaligned with your expense required for each key decision). Then we strategy, all because you have ignored decisions. The calculated a composite score for each company and proper place for this type of SWOT analysis is not, in compared that score with each firm's financial per- fact, as a prelude to organizational change but earlier, formance. (See the exhibit "Rate Your Decision Ef- when you are determining your company's strategy. fectiveness" to learn how your company compares A better way to begin a reorganization is with a with our sample.) decision audit. The first step in conducting one is to We found that decision effectiveness and finan- identify the key decisions you need to make and ex- cial results correlated at a 95% confidence level or ecute, given your strategy to create maximum value June 2010 Harvard Business Review 57SPOTLIGHT ON STRATEGIES FOR A CHANGING WORLD Quick Test Rate Your Decision Effectiveness Quality Speed Yield How do your organization's When looking back on We make critical decisions We execute critical critical decisions, we find decision abilities stack up against Much faster lecisions as intended that we chose the right than competitors 4 Most of the time the competition? While hardly a course of action Most of the time Somewhat faster Some of the time full-scale survey, this short test than competitors Some of the time 2 Infrequently Somewhat slower can give you a quick-and-dirty Never 2 Infrequently than competitors assessment of decision strengths 1 Neve Much slower than competitors and weaknesses. Q SCORE S SCORE Y SCORE for your shareholders. The set of key decisions for a platforms, for instance, which had been painfully growth strategy, for instance, will be different from difficult under the old structure. Set in this context, the set for a return-focused strategy. Of course, this the reorg made perfect sense and helped restore the exercise does not presuppose a change in strategy. It company to profitability in early 2010. may be that the reorganization is an attempt to im- As you conduct your own decision audit, you prove an existing strategy, in which case you'll end up need to consider two types of critical decisions: with a comparison between the decisions you ought . Big, one-off decisions that individually have a to be concentrating on and the ones you are actually significant impact. Petrochemical companies, for making. The bigger the difference-and the greater example, must make periodic multibillion-dollar the obstacle presented by your organizational structment decisions, such as if, when, or where ture-the more aggressive your reorg will need to be. to build a new ethylene cracker, which is critical to The ongoing turnaround at Ford illustrates the production. If a company builds at the wrong time, power of explicitly delineating a company's critical in the wrong place, or with the wrong technology, decisions. When Alan Mulally became CEO at the it will have to live with the consequences for many automaker in 2006, the company was in dire need decades. of change. Ford had been losing a point or more of . Small, routine decisions that cumulatively have market share every year since 2000 and was on the a significant impact. Amazon's continuing success verge of collapse. But rather than change the com- can be attributed partly to a host of savvy merchan- pany's structure first and then worry about decisions, dising decisions, including those related to special Mulally took the opposite approach. He and his team prices and shipping discounts, suggestions for com- outlined the decisions that were critical to a turn- plementary purchases, and targeted e-mail notices around. Only then did they begin to build the new about new offerings. None of these decisions carries organization around those decisions much value in any one instance. Cumulatively, how- Fixing the company's operations and restoring ever, they can mean the difference between success profitability centered on a schematic depicting Ford's and failure. critical decisions. It spelled out the key decisions Once you have identified which decisions are that needed to be made at each stage in Ford's value critical and categorized them, you can figure out chain, along with the infrastructure required to ex- where in the organization those decisions should be ecute them effectively. Every week, Mulally and his made. This requires an unbiased assessment of the team tracked their progress in making and execut- benefits of scale and coordination versus the ben- ing these decisions. They divested noncore brands efits of tailoring to local needs and staying close to such as Aston Martin, Jaguar, Land Rover, and Volvo; the customer. In which decisions is scale a critical reduced the number of production platforms; began factor? Which decisions are better made by business consolidating both suppliers and dealers; and so on. units or functions? Which need coordination across They also reorganized the company, moving from a many businesses? structure based on regional business units to a global Some decisions are fairly easy to place. Big matrix of functions and geographies. This new struc capital-allocation decisions, for instance, are typi- ture enabled Mulally's team to make its most impor cally best made by the corporate center so that se- tant decisions better and faster-creating global car nior leaders can design and execute a coordinated 58 Harvard Business Review June 2010THE DECISION-DRIVEN ORGANIZATION HBR.ORG Effort In making and executing To tally your overall score, multiply Using our database, you can More than 25 = You're doing great; critical decisions your scores for quality (Q), speed compare your decision score with keep it up. 4 We put in exactly the (S), and yield (Y) to get your QSY. the performance ranges for each 21-25 = Pretty good, but right amount of effort Next, divide your effort (E) score by benchmark quartile. If you score over could be great. We put in somewhat too 4, and multiply the result by the QSY. 25, for example, you are performing If you're putting in exactly the right at the level of the top-quartile 16-20 = Worse than 50% much/too little effort amount of effort, your QSY wouldn't companies in our survey. Relative to of companies. We put in way too much/ change. But if you scored effort as a 3, our sample, your organization would Time to act. nowhere near enough then your score would be 3/4, or 75%, appear to decide and deliver. 15 or less = Major decision effort of the QSY. reboot required. 1 We're off the charts E SCORE TOTAL SCORE [Q X 5 x Y X (E/4)] When reorganizing a company, decisions rather than structure should be the focus. portfolio strategy across the company. IT invest- office products at the time-within the senior lead- ments can usually be left to functions or business ership of the Product organization, which helped units. Decision placement is more of a challenge for accelerate the pace of new-product introductions in product, customer, and channel decisions, which this vital segment. The explicit focus on where deci- typically involve complex trade-offs. Pricing deci- sions should be made was critical to the successful sions, for example, need to be coordinated across turnaround at Xerox. customer segments and channels. Product decisions At the end of your audit, you may find that mak- must be considered from both an internal and an ex- ing and executing decisions better and faster than ternal perspective. In companies with many differ- competitors doesn't require reorganization and ent products or services, both the critical decisions that you can avoid the costly business of structural themselves and where they should be made may change altogether. If you find that you are already vary widely across the organization. focused on the right set of critical decisions and The Xerox turnaround launched in 2001 under you find that they are placed where they should be Anne Mulcahy offers a powerful contrast to the Ya- within the organization, then the source of perfor- hoo story. Unlike Semel at Yahoo, Mulcahy's lead- mance problems is unlikely to be structural. Prob- ership team took a decision-driven approach to the lems may stem instead from other organizational company's reorganization. Team members went issues. Perhaps there is an incentive problem with through every set of critical decisions Xerox needed the sales force. Maybe leaders place too much value to make and execute to fend off bankruptcy, and on building consensus, at the expense of decisive ac- they made explicit choices about where to locate tion. Whatever the issues, they can probably be fixed those decisions. Clarity and simplicity were the guid- without a wholesale structural redesign, which will ing principles. In the sales organization, for example, take people's minds off the competition and may ac- Xerox moved from a global customer structure, in tually add to your problems. (See the exhibit "Do You which sales and pricing decisions were made by Really Need to Reorganize?") global teams organized around industry verticals, to But for now, we'll assume that your audit does re- a simpler country structure, where those decisions veal that your structure needs to change. Let's look rested with local sales teams. The new structure at how to go about doing that. enabled Xerox to eliminate several layers of middle management, increase local accountability, and take Building the nearly $1 billion out of the company's cost structure Decision-Driven Structure in just two years. The simpler structure also con- All complex organizations must be broken down into centrated decisions related to the shift from analog manageable pieces to ensure that roles and respon to digital technology-critical to Xerox's success in sibilities for making and executing critical decisions June 2010 Harvard Business Review 59SPOTLIGHT ON STRATEGIES FOR A CHANGING WORLD Quick Test Do You Really Need to Reorganize? A comprehensive survey of your organization can help you understand what's undermining or supporting effective decisions-and whether structure or something else is likely to be the most important issue. But this quick test can serve as an early warning device. are clear. A good way to determine what the impor- tant decisions are in your company is to look at the STRUCTURE To get your total score sources of value in your business and then organize add up your individual Our structure helps-rather than hinders-the the macrostructure around them. scores . decisions most critical to our success. Take the case of British Gas, a division of the mul- L 3 4 tinational energy and utility company Centrica. In More strongly Disagree Agree more Strongly than 35 = You're doing 2006, faced with a serious performance crisis, the disagree more than than agree great; keep company's new leadership team started looking at it up gree disagree the sources of value in its business. Managers be- 31-35 = Good, but room for gan by examining differences in profitability by ser- ROLES improvement. vice, by geographic area, and by customer segment. Individuals understand their roles and 26-30 = Org is serious They discovered that profitability and growth varied accountability in our most critical decisions. barrier to much more by customer segment than by any other 1 2 3 decisions. variable. PROCESSES 10-25 = Major org One segment used large amounts of gas or elec- transformation Our processes are designed to produce effective, required. tricity and paid regularly through guaranteed direct timely decisions and action. debits. Decisions that helped the company retain 1 2 3 Compared with companies these customers, such as how to handle home moves in our database, a score INFORMATION and how best to offer additional services, were most above 35 puts you in the The people in critical decision roles have the top quartile-meaning important for this segment. A second customer seg- Information they need when and how they need it. that your organization is ment used less energy and paid regularly through a 1 2 3 4 pretty healthy. A score of system of prepayment cards. Here the key decisions 31 to 35 indicates room MEASURES & INCENTIVES for improvement, but related to controlling costs, particularly those associ- no immediate signs of ated with processing additional payments and with Our measures and incentives focus people on organizational breakdown. meter reading. A third segment wasn't as consistent making and executing effective decisions. A score of 30 or below in- in keeping up payments. For that group, the critical 1 2 3 dicates that you definitely have some organizational decisions related to managing receivables. PRIORITIES challenges to address. If Recognizing those different sources of value, People understand their priorities clearly enough you perform at 2 or less managing director Phil Bentley decided that the on any one issue, that to be able to make and execute the decisions they particular ailment likely best way to structure the company was by customer face. needs attention. segment. He established three separate businesses: 1 3 Premier Energy, Energy First, and Pay-As-You-Go DECISION STYLE Energy. This new structure allowed him to place We make decisions in a style that is effective (for accountability for decisions that directly affected example, that appropriately balances inclusiveness customers, such as service levels, positioning, and with momentum). product bundling, in the business units. Corporate 1 2 3 headquarters could focus on noncustomer-facing matters such as IT and finance. The alignment of PEOPLE We put our best people in the jobs where they can structure and decisions helped British Gas improve have the biggest decision impact. its performance significantly. It reduced customer 1 2 3 attrition from about 20% to less than 10%. Its bad debt fell, and the business began growing for the first BEHAVIORS time in years. Our leaders at all levels consistently demonstrate If decision-driven reorganization were just about effective decision behaviors. an audit and, if necessary, aligning the macrostruc- 1 2 3 ture with the sources of value, it would be relatively CULTURE easy to get your next reorg right. But the messy real- Our culture reinforces prompt, effective decisions ity of business is that big changes need to be followed and action throughout the organization. up with a whole host of smaller changes 1 2 3 In most reorgs, this stage of the process usually concentrates on assigning turf and setting out a hier- So Harvard Business Review June 2010THE DECISION-DRIVEN ORGANIZATION HBR.ORG archy with defined reporting lines. Once again, this BP had eliminated $3 billion in costs and was turning approach is misguided, because it does not use deci- in a profit that beat analysts' expectations by 50%. ions as the unit of analysis. Indeed, a focus on turf Any change in structure may necessitate changes often devolves into horse trading: Powerful manag- in decision roles, incentives, information flow, per- ers grab decision rights they shouldn't really own formance metrics, and processes. At UD Trucks-for- while weak ones surrender rights they really should merly Nissan Diesel-a new strategy designed to ex- own. People end up with responsibilities that are pand UD's penetration of national accounts required defined either too broadly or too narrowly, given the a reorganization of the company. Local branch of- decisions they need to make. Responsibilities that The Six Steps to fices were combined into regions, and national ac- are too broad result in insufficient supervision and Decision-Driven count teams were established. To get the most out limited accountability. Those that are too narrow Reorganization of the new structure, the team at UD Trucks clarified can create needless hierarchy-too many watchers decision roles between the national account teams and too few doers-and can encourage microman- Identify your and the company's Truck and Service units. Mea- agement. Without a focus on decisions, these power organization's key sures and incentives were reset to encourage col- struggles too often lead to creeping complexity in an decisions. laboration across units and focus the sales force on organization's infrastructure. key accounts. The reorganization at UD Trucks was The energy giant BP provides an excellent ex- Determine where critical, of course-the new strategy could not have ample of the consequences of an overly complex microstructure. Back in 1994, BP had just a handful in the organization been implemented without it. But goals, processes, information, measures, and incentives also needed of geographic and customer units, which were sup- those decisions to be aligned to make the new structure work. ported by a few central functions. After a period of should happen. Finally, you need to help managers develop significant merger and acquisition activity, however, the skills they need to make decisions quickly and the company added new geographic areas and cre- Organize the translate them into action consistently. Smart com- ated new functions. Spans of control narrowed, and _macrostructure panies mesh individuals' capabilities with the orga- the number of management layers increased. All of around sources ization's decision-making demands. They invest as these changes increased the number of "decision of value. needed to ensure that people have the skills required nodes"-the interfaces between regions, functions, to be better decision makers over time. and layers required to make and execute important Figure out what A good example of this took place at Hospira, the decisions-from about 500 in 1994 to roughly 10,000 level of authority $3.9 billion pharmaceutical company spun out from in 2007 decision makers Abbott Laboratories in 2004. Each decision node at the company introduced a In 2008, Hospira embarked on a major change different set of hurdles for new products, business need. program. The program included an effort to build development deals, and even options for reducing new decision capabilities across the company. The costs. The effects were predictable: Decision making Align other top 80 executives attended innovative training and execution slowed, and costs mounted elements of the workshops that showed them how to identify an or- In decision-driven reorganization, the challenge organizational ganization's critical decisions and outlined Hospira's is to determine exactly what authority decision mak- system, such as ew approach to effective, efficient decision making ers need, regardless of their organizational status, if incentives, infor- and execution. Executives learned to use tools that they are to make good decisions and execute them mation flow, and would help them get the who, what, when, where, effectively. BP saw that the people who were best processes, with and how of each decision right. The workshops also equipped to make decisions had to get too many those related to helped leaders learn and adopt the specific behaviors approvals from higher-ups or from regional heads, decision making. that would be required to make the changes stick which delayed execution. So its new chief executive Hospira's senior leaders began to track the com- Tony Hayward, launched a comprehensive simplifi- Help managers pany's changes in decision effectiveness at their bi- cation program designed to return decision rights to the appropriate people. The initiative eliminated lay- develop the skills monthly team meetings. They also began to gather survey data that could help them assess progress in ers of middle management, centralized some opera- and behavior building and sustaining decision capabilities. Since tions, and reduced overhead expenses by one-third. necessary to 2008, CEO Chris Begley and his team have seen sig It moved the number of decision nodes back toward make and execute nificant improvements in Hospira's financial and a target of 5,000. Both the company's decision effect decisions quickly stock price performance, which they attribute largely tiveness and its performance improved. By late 2009, and well. to improved decision making June 2010 Harvard Business Review 61SPOTLIGHT ON STRATEGIES FOR A CHANGING WORLD HBR.ORG Ultimately, a company's value is just the sum of the decisions it makes and executes. Adjusting to New Structures connected architecture, emerging countries, and A new strategy-or new execution of an existing connected business operations. These interlocking strategy-can require both macrochanges and micro- councils and boards have enabled Cisco to maintain changes to a company's structure. But any new struc its leadership position in the complex, fast-moving ture will create new boundaries that people may find telecommunications inications space and build on the strength hard to cope with and that may make effective deci- of its global functional structure. sion making more difficult. To get around this prob- Creating parallel decision-making authorities lem, it may be necessary to overlay your new struc may appear to be in conflict with our general princi- ture with some connections that help people reach ples of simplicity and clear accountability, but we be- beyond those boundaries. lieve this approach leads to a more streamlined pro- In 2001, faced with the bursting of the dot-com cess. Organization overlays such as Cisco's councils bubble, Cisco reorganized. It moved away from the and boards introduce valuable expertise that formal structure that had been in place for much of the structures cannot easily accommodate. They allow company's rise to prominence-a line-of-business fewer people to be involved in making and executing structure based on customer segments-and put critical decisions-in effect, reducing the number of in place a global functional organization. The logic decision nodes. was simple: Reorganizing around central functions would dramatically reduce the company's costs. IT'S EASY to see why so many CEOs are enticed by While the pros of this move outweighed the cons, the temptation of reorganization. Today's corporate the company's leaders acknowledged that organiz- structures can be inordinately complex. Some date ing around centralized functions could cause Cisco from a day when the demands of the business world to lose customer focus and intimacy. were quite different than they are today. Simplifica- To prevent that, the leadership team has created tion, alignment, modernization, a new vision of what over the past several years a series of corporate coun- the organization should look like-and all of it ac- cils and boards directly under the company's operat- complished with the stroke of a pen. But to focus ex- ing committee, Cisco's most senior decision-making clusively on structure is to confuse means with ends body. These cross-functional groups formulate and and to assume a connection that may not exist. The evaluate alternatives for each of the company's ma- reorgs that work best, such as those at Ford, Xerox, for strategic initiatives and then make recommenda- and Cisco, focus first on the organization's critical tions to senior management. We find that the pro- decisions. Then they build a zy build an organization that can cess accelerates decision making without sacrificing make and execute those decisions better and faster decision quality. The structural overlay of councils than the competition. The result is what executives and boards also seems to help functional leaders col- always seek from reorgs yet so seldom accomplish: laborate and make effective decisions about budgets improved performance. and resources. HBR Reprint R1006B Today, those teams are operating at full steam. Five "segment councils," for instance, support deci- Marcia W. Blenko (marcia.blenko@bain.com) sion making and execution for the company's enter- leads Bain & Company's Global Organization Practice and is a partner in the firm's Boston office. Michael C. prise, commercial, service provider, small business, Mankins (michael.mankins@bain.com) leads Bain's and consumer segments. Boards organized around Organization Practice in the Americas and is a partner in specific market opportunities, such as collaboration San Francisco. Paul Rogers (paul.rogers@bain.com) is a partner who leads Bain's London office; he formerly led and virtualization within the enterprise segment, Bain's Global Organization Practice. They are coauthors of develop tailored strategies and oversee execution. the forthcoming book Decide and Deliver: Five Steps to Finally, four cross-segment councils focus on the Breakthrough Performance in Your Organization (Harvard Business Press, 2010). Portions of this article are adapted arcane but critical topics of emerging solutions, from the book. 2 Harvard Business Review June 2010