In 1994, Gordon Bethune took the helm, with Brenneman becoming president and chief operating officer. They staved off bankruptcy by renegotiating with their creditors. And they launched an organizational turnaround that proved remarkably successful, catapulting Continental from worst to best among big U.S. carriers.

By 1995, Continental was moving up on the Department of Transportation’s (DOT’s) performance measures (see Figure 1). Its stock price was soaring. And the turnaround stuck. The latest rankings by Consumer Reports place Continental first among the seven big U.S. airlines. Zagat’s 2007 survey of frequent flyers found overall ratings for the big airlines were low and declining, with the “notable exception” of Continental. Continental was the only big airline, and one of only five overall, to be a Zagat Top Spot.

The mid-’90s were also a time for change in New York’s police department (NYPD). As W. Chan Kim and Renée Mauborgne describe in their 2003 HBR case study, “Turf wars over jurisdiction and funding were rife. Officers were underpaid relative to their counterparts in neighboring communities…. Crime had gotten so far out of hand that the press referred to the Big Apple as the Rotten Apple.” In response, then-mayor Rudolph Giuliani hired William Bratton to lead the NYPD, fresh from a string of successful turnarounds of other agencies, including NYC’s transit police.

Though crime rates in NYC had started to decline in the late 1980s, Bratton’s arrival accelerated the trend (see Figure 2). Time wrote in a 1996 cover story, “The drop became a giddy double-digit affair, plunging farther and faster than it has done anywhere else in the country, faster than any cultural or demographic trend could explain. For two years, crime has declined in all 76 precincts.” As Kim and Mauborgne note, the change wasn’t just a flash in the pan or a nationwide trend: “Statistics released in December 2002 revealed that New York’s overall crime rate [was] the lowest among the 25 largest cities in the United States.”

Finding the Keys

These turnarounds are classic: rapid U-turns from the brink of doom to stellar success. They may not last forever. But if a failing school could achieve similar results for several years, thousands of students would benefit permanently. How did they happen? This article explains what we know, from plentiful cross-sector research, about how to engineer turnarounds within existing organizations. It then identifies two critical policy issues that states and districts must address to accelerate the prevalence of real, successful turnarounds in education.

Education reformers faced with failing schools and districts tend toward one of two camps: The Incrementalists hold that meaningful improvement can only happen slowly, with soul-wrenching culture change leading to instructional change and eventual student success. The Clean Slate Club believes the only way to fix failing schools is to shut them down and start fresh, with entirely new rules, staff, and leadership.

Both camps have it wrong, but for different reasons. The slow and steady approach won’t work for chronically failing organizations. The fresh-start method is much more promising, based on the dramatic success of some newly formed schools serving tough populations. But most start-ups fail or bump along in the purgatory of mediocrity, even in sectors that, unlike education, enjoy abundant venture capital and a ready stable of capable entrepreneurs. Moreover, troubled organizations across sectors regularly transform themselves from bad to great without a clean slate. The consequence of education’s failure to recognize turnarounds as a means of school improvement is twofold: in education, turnarounds have been tried rarely and studied even less. While education researchers catch up, practitioners can use the turnaround lessons of other sectors.

Essential Actions

To identify what makes turnarounds successful, we reviewed dozens of studies across a wide range of organizations: nonprofits of differing sizes, some in highly regulated industries such as health care; government agencies with varying missions; and for-profits in numerous industries. Case studies of single turnarounds comprise most of this research, including studies of both large, stand-alone entities and small units within larger organizations, closer in size to schools. The turnaround precursors, patterns of action, and chronically challenging environments we found were surprisingly consistent across these varied venues, bolstering their potential relevance to both districts and schools. Turnarounds were attempted when organizations were failing by many measures, not just financial metrics.

While this article uses the well-documented Continental Airlines and NYPD cases as illustrations, what happened in these two organizations is similar to what we saw across the research. We coded the cases from this broad research to reveal two overall success factors.

First, turnaround leaders work in an environment that gives them what we call “the big yes.” Second, bad-to-great transformations require a point-guard leader who both drives key changes and deftly influences stakeholders to support and engage in dramatic transformation. To be sure, staff help effect a turnaround, but the leader is the unapologetic driver of change in successful turnarounds. Effective turnaround leaders follow a formula of common actions that spur dramatic improvement. The actions interact to move the organization rapidly toward impressive, mission-determined results that influence stakeholders to support additional change. Below, we explain the six most consistent actions in the bad-to-great formula and provide an example of what each action might entail in school and district turnarounds.

Focus on a Few Early Wins

Successful turnaround leaders choose a few high-priority goals with visible payoffs and use early success to gain momentum. While these “wins” are limited in scope, they are high-priority, not peripheral, elements of organization performance. Early wins are critical for motivating staff and disempowering naysayers.

At Continental, Bethune and Brenneman initially focused on what Brenneman calls “the customers in seat 9C, the business travelers who book the aisle seats in the front of the plane. They pay full fare, and they travel a lot.” To win these customers back, Continental launched a massive effort to refurbish airplanes inside and out, recarpet their terminals, and upgrade food service, all in six months rather than the four years originally estimated.

These changes might seem merely cosmetic. But in fact they addressed a major concern of the customers most important to the airline’s success. And the upgrades built positive momentum for further change. As Brenneman recalls, employees “could see senior management finally taking the actions they knew had been needed for years.” For a demoralized organization, this kind of mission-focused early win is vital to convincing the team that it can in fact be successful.

At NYPD, Bratton initially launched an effort to crack down on minor offenders. While their offenses weren’t the city’s biggest crime issues, the effort helped convince skeptical citizens and officers that the police could make a difference.

In schools, early wins must tackle similarly visible goals essential to the learning mission. An elementary school might aim to raise reading scores to within one grade level of year-end goals for 90 percent of 5th graders by the first semester’s end. This is challenging in schools where many children are multiple grade levels behind. But it is achievable, as many cases of high-poverty start-up schools have demonstrated, and a necessary step toward achieving grade-level pass rates at year’s end. All other changes can support this goal. Imagine the impact when teachers realize that the school need never again graduate a class of non-readers.

A district also must focus early wins on student learning to fit the turnaround formula, perhaps by adopting similar goals for one subset of struggling children or a few low-performing schools. To achieve the goals, the district must then tackle barriers blocking success for those students or schools. For example, a district might arrange to provide targeted schools with materials online to work around book shortages or improve dramatically their access to interim assessment data. Such online materials, assessment data, and other changes in district management systems are not themselves “early wins.” They must be used as tools to achieve rapid academic results and convince stakeholders that additional focused change will produce more success.

Break Organization Norms

In a failing organization, existing practices contribute to failure. Successful turnaround leaders break rules and norms. Deviating to achieve early wins shows that new action gets new results.

In response to Continental’s financial struggles, an entrenched norm of cost cutting pervaded the organization. As Brenneman explains, the company’s “myopic focus” on costs had led to perverse tactics: skimping on cabin air conditioning and flying more slowly to cut fuel use; removing high-revenue first-class seats to squeeze in more passengers; and eliminating corporate discounts even for the airline’s top customers. Brenneman calls the result a “doom loop. By focusing only on costs, the airline had created a product no one wanted to buy.” Declining revenues sparked more ill-advised cost cutting, such as morale-sapping wage reductions.

When Bethune and Brenneman took over, they pursued strategies that actually increased costs, like the plane and terminal upgrades. The airline started paying employees more, based on performance. For every month the airline finished in the DOT’s top five for on-time arrivals, each employee received $65. The on-time bonuses cost the company $3 million per month, but improving the on-time record boosted overall financials by an estimated $8 to $9 million per month.

Like many large organizations, Continental had accumulated hundreds of regulations. The result was a nine-inch-thick tome known as the “Thou Shalt Not” book. A central part of leadership’s plan was to free employees to do what was needed to solve problems and meet customers’ needs. To make the point, the executives took a copy of the book into the parking lot, soaked it with gasoline, and torched it in front of a crowd of employees.

Bratton, too, made a practice of norm busting. At NYPD, he soon learned that only 5 percent of the budget went to narcotics forces, even though a high percentage of crimes were drug-linked. The reason? An assumption that the department’s top priority was responding to 911 calls, rather than to the kind of long-term, preemptive work done by the narcotics unit. In addition, the narcotics squad worked Monday through Friday, while narcotics activities and related crime soared on weekends. One of Bratton’s early actions was a major reallocation of staff and resources into narcotics, including shifting officers’ time to weekends.

In an elementary school, the leader might bend time-use norms by having teachers provide rolling reading instruction as children arrive on buses in the morning. Rescheduling classroom volunteers into lunch-hour chaperoning could replace lost morning teacher-planning time. This schedule adjustment would add one to three weekly instructional hours per child in many schools.

For districts, delivering individualized reading assessment and instruction to every classroom via technology, for example, would require veering from textbook and technology budgets, as these line items are typically separate. Shifting dollars can ignite turf battles, because budgets are often equated with number of staff positions and job importance of district department leaders. The key is making the learning goal the organization’s clear priority.

Push Rapid-Fire Experimentation

Turnaround leaders press a fast cycle of trying new tactics, discarding failed tactics, and investing more in what works. They resist touting mere progress as ultimate success.

Bratton’s most famous innovation was the introduction of the Compstat system, short for computer statistics, which provided everyone from precinct staff to top brass with detailed statistics and maps showing how patterns of crime and law enforcement actions played out in different places and over time. The system made possible big, department-wide strategic decisions, like the reallocation of resources to narcotics work.

Perhaps more important was the system’s value for precinct commanders as a day-to-day management tool. The Time cover story on Bratton begins with an account of a semiweekly Compstat meeting, in which a precinct commander is grilled about a rise in robberies and his response. New problems demand new strategies, and the Compstat meetings were designed to keep that fast cycle of response-measure-adjust going.

In a school, the leader might redeploy a motivated, technology-capable staff person to provide Compstat-style reports of student-by-student, teacher-by-teacher, grade-by-grade results on mandatory quizzes. This effort would provide the fodder for making changes before semester’s end. Most important, each person and team would receive timely data about the progress of students for whom each is accountable.

In a district, new interim assessment data would provide feedback about what schools, grades, and student subgroups are meeting goals. Slow progress would be a trigger for district organizers to do some problem solving.

Get the Right Staff, Right the Remainder

Successful turnaround leaders typically do not replace all or even most staff at the start, but they often replace some key leaders who help organize and drive change. For remaining staff, change is mandatory, not optional.

At Continental, cleaning house at the top of the organization was a big part of the turnaround. Of 61 officers, Bethune and Brenneman showed 50 the door. Some housecleaning took place at lower levels as well, but an organization with 40,000 employees can’t possibly transform itself by swapping out all of its people.

Continental’s new “people strategy” focused on making dramatic change mandatory for employees already in their positions. When the maintenance department told Brenneman that plane and terminal upgrades, his key “early win,” was a four-year project, Brenneman insisted on his six-month schedule: he’d find someone else to do the work if the maintenance department wasn’t up to the job. As it turns out, the department was up to the job, once it was clear that change was mandatory.

Bratton also mostly replaced leaders, not the rank and file. His “number two” was a veteran officer who knew everyone at headquarters. One of his first jobs was to help Bratton identify members of top staff likely to oppose or seek to undermine his reforms, leading to what Kim and Mauborgne call “a dramatic changing of the guard.” Bratton did replace half of his precinct commanders, but not immediately. The turnover grew out of the Compstat process. As Time wrote in 1996 on Bratton, “Effective precinct commanders…merely get grilled to a medium rare at Compstat. Those who show up unprepared, without coherent strategies to reduce crime, are fried crisp, then stripped of their commands.” Swapping out people was core to Bratton’s approach, but it followed from his turnaround efforts rather than preceding them.

In a school, the total staff replacement advocated by the Clean Slate Club would not be necessary. While not every teacher would be willing and able to do what’s needed, most would rise to the occasion. The rest typically reveal themselves during the “early win” phase and must then be removed.

The most important early staff decision would be the selection of an organizer to drive the action plan. The person might or might not be selected from the current staff and might be given power exceeding the person’s current title and tenure. This individual would ensure, for example, that analysis of student progress and instructional problem solving happened regularly, timed with the quiz schedule.

For a district turnaround, the superintendent would need to tap a trusted leader who could cut through the usual district machinery. This leader’s team would need to include additional organizers who could focus on implementation issues in targeted schools or student populations, and each of these people would need to be accountable for learning success among their assigned students. The superintendent might also replace critical department leaders from the start, making room for team members who can drive change.

Drive Decisions with Open-Air Data

Successful turnaround leaders are focused, fearless data hounds. They choose their initial goals based on rigorous analysis. They report key staff results visibly and often. All staff who participate in decisionmaking are required to share periodic results in open-air sessions, shifting discussions from excuse making and blaming to problem solving.

Again, Bratton’s Compstat meetings are a powerful example. These regular gatherings brought together top brass with all 76 precinct commanders, the police force’s key line managers. At every meeting, one commander took the hot seat, facing questions about the precinct’s performance that emerged from the Compstat data. How was the precinct working to solve the problems the data revealed? Why was performance going down on some key metrics?

The result was what Kim and Mauborgne call “a culture of performance…. An incompetent commander could no longer cover up his failings by blaming…neighboring precincts, because his neighbors were in the room and could respond. By the same token, the meetings gave high achievers a chance to be recognized.” Some commanders used similar tactics within their own precincts, extending the new culture.

Bethune and Brenneman, too, used data to drive change. As they were poised to assume the leadership of Continental, the twosome met over dinner for a week, poring over data and writing down “everything that was wrong with Continental.” The result was a set of some 15 key metrics that the pair decided to track rigorously and publicly over time and compare with those of their competitors. Results on these metrics, good and bad, became the central focus of a massive communications campaign that leadership launched inside and outside the company.

“Using data to drive instruction” has become such a mantra in public education that it’s important to pause here and explain how data strategies in successful turnarounds differ from typical K–12 data systems. The keys are using the right data to drive change and requiring all relevant staff to put their data on display in an open-air forum and then face tough questions (and helpful problem solving). The process helps people improve their practice, but it also transforms the culture.

In a school, staff capable of leading instructional change for learning results would be identified by student progress data. Those not capable of leading or accomplishing instructional change would be identified as well. The progress data would provide the school leader with a guide to the staff changes that would further improve student learning, and the achievement of early goals would help build support for such changes.

In a district, progress reports would enable the leader to evaluate the school-level leaders and district team members responsible for implementing changes by tracking the results achieved for defined groups of students within or across schools. Each of the staff leaders affected would need to be included in regularly scheduled meetings to present their own performance data for discussion.

Lead a Turnaround Campaign

Successful turnaround leaders know that change of any kind is hard and that people resist it for many reasons unrelated to success. Leaders use a consistent combination of motivating and maneuvering tactics that include communicating a positive vision of success; helping staff personally feel the problems customers feel; working through key influencers; and silencing critics with speedy success of early wins, thereby casting vocal naysayers as champions of failure.

Continental’s leadership orchestrated a “forgiveness campaign” to apologize to its unhappy customers. Officers, from the CEO through the vice presidents, divided complaint letters and started placing calls. Each officer took a city served by Continental and contacted travel agents and corporate customers. Saying “sorry” was part of the script, but the other was outlining the airline’s bold plan to fix problems. “We heard our share of shouting,” recalls Brenneman, but he argues that the campaign helped reverse the “doom loop” by convincing many customers that change was happening. Of course, this communication onslaught only worked because leadership had results to show, flowing from its early wins.

According to Kim and Mauborgne, one of Bratton’s specialties was putting managers face to face with the operational problems as a way of convincing them, in ways that no amount of memos, speeches, and PowerPoint presentations could, of the change imperative. As head of the NYC transit police, Bratton had famously battled complacency by requiring all senior managers to ride the subway to work and meetings, including at night, and did so himself.

At NYPD, Bratton hired John Miller, an investigative journalist, to lead his communications efforts, both inside and outside of the force. And he needed all the help he could get. One key “early win,” processing small “quality of life” crimes, was nearly scuttled by court officials who feared these cases would clog the dockets. By allying with the mayor and running a smart media campaign, Bratton framed the issue as make or break for NYC’s future, causing judicial leaders’ concerns to appear selfish and petty. The strategy worked.

In both schools and districts, leaders and their teams would need to analyze the required involvement and likely reaction of all stakeholders: school and district staff, parents, students, unions, and community members. At the start, most stakeholder groups would feel that their power was being reduced as the turnaround leader focused sharply on early-win goals. Leaders would need to communicate clearly how success would affect children’s later learning and work prospects. They would need to find ways for staff to empathize with children experiencing slow or no change. And they would need to identify vocal supporters and work with them to rally others to advocate for change. Most important, the leaders would need to achieve naysayer support or silence by accomplishing early student-learning gains.

The Turnaround Environment

These six key actions recur in story after story of successful turnarounds. But don’t turnaround leaders also need a supportive environment? Yes and no. Some conditions prove to be not that valuable, or even detrimental. Some scholars, for example, conclude that too much money dooms turnaround efforts, by diluting leader attention rather than focusing it on early wins.

One environmental condition is critical. Turnaround leaders need a “big yes,” a clear nod from the top in support of dramatic change, even if it causes discomfort and political fallout. However, there is no evidence that the larger organization needs to be highly effective or in turnaround mode to grant the “big yes” to a unit leader. Indeed, breaking the norms and rules of the status quo to achieve support-winning early victories is what successful turnaround leaders do.

While leaders at both Continental and NYPD had a “big yes” from their ultimate bosses, they were not handed a clean slate. Instead, they faced the same tough environmental conditions plaguing failing schools and districts: tight budgets, deep-seated status quo routines, and tough opposition from organized employees. They turned around their organizations nonetheless.

Enabling School and District Turnarounds

To enable more widespread, successful turnarounds in education, state and district leaders need to focus on two critical policy changes. First, states (particularly governors) need to create much more political will to try turnarounds at the district level and to retry when some inevitably fail. They can only do this by developing much more capacity, in-house or through contractors, to take charge of failing schools when districts don’t act.

Second, states and districts could do much more to fuel the pipeline of K–12 turnaround leaders. One key step is to open the door to noneducation leaders with turnaround competencies, induce them to take the job, and invest to equip them with the education know-how they need to succeed.

A few states and districts, such as Chicago, the District of Columbia, and Louisiana, are attempting real turnarounds. Related efforts, such as New Leaders for New Schools and the University of Virginia’s School Turnaround Specialist Program, are underway to help more turnaround leaders succeed. Mass Insight Education has launched a national campaign to encourage state leaders to play a more active role.

All of these initiatives are promising. And the good news is they don’t have to start from scratch. From Continental Airlines to NYPD to countless others, turnarounds have happened with dramatic results. Turnarounds can happen in education, too.