The initiative was under way. But Andrea, the head of the company’s change office, was feeling frustrated. For a while the news had been good: workstreams proceeding as expected, key activity milestones tracking green, decisions happening on time. But now one critical decision—about discontinuing certain activities—was stuck. Sometimes it fell off the senior team’s agenda. Other times the team debated it without reaching a conclusion. Meanwhile, two groups of employees were performing tasks that were no longer essential. And nothing Andrea did seemed to help.

Under normal circumstances, making and implementing decisions can feel like juggling several balls at once. During a change effort, as someone once put it, decision making can feel like juggling chainsaws—while blindfolded and balancing on a large ball. Too many decisions fall by the wayside. Too many turn out wrong.

Why is change so difficult? We see three main reasons:

One is that every transformation involves so many key decisions. Just to launch an initiative, corporate leaders must identify the facts that indicate a need for change choose the best course of action, validate the business case, determine the highest risks, select the right leaders, and so on. The list of critical decisions only expands once the initiative gets underway. In the meantime, of course, someone has to make and execute the decisions required to run the business.

Another reason lies in the mood swings that can affect an organization undergoing change. The initial phase of a transformation usually provokes fear and skepticism. Once the possibilities begin to materialize, and people feel optimistic, even exhilarated. Later, as new obstacles appear, they become anxious and pessimistic again. Various cognitive biases intensify these moods and impede people’s ability to make good decisions. In the first phase, for instance, the bias known as anchoring, or relying on familiar reference points, locks people into conventional thought patterns and clouds their judgment about alternatives.

A third reason: The stakes are high. Every major change is highly visible to employees and peers, to shareholders and board members, and also to customers. The company has invested a lot of resources. A decision maker’s credibility is at stake. The cost of failure is steep. In many cases, decisions made during a transformation effort can define an executive’s career.

Companies that master these decision-making difficulties typically follow a few common practices:

They establish a “decision drumbeat”

Most companies organize change efforts around deliverables. We find it’s more fruitful to organize the work around key decisions, creating a kind of “decision drumbeat” that governs the process and ensures adherence to a timetable.

When one major brewer acquired another brewer, for instance, team members set an aggressive timeline for integration and stretch targets for synergies and results. To deliver, they mapped out the major decisions required to consolidate manufacturing facilities, distribution channels and advertising contracts. The team specified all the smaller decisions that had to happen before the major decisions could be made, and the team laid those out on a detailed calendar. So right from the start this decision drumbeat kept the integration marching forward— everyone knew what decisions needed to be made next, what information they would need and the time they would need it by. A decision drumbeat isn’t relevant only for acquisitions. It’s useful for any change process that involves linked decision making, such as strategy development or the launch of a new product.

They track leading indicators

It also helps to track leading indicators of the change process, not just indicators of what has happened so far. One such critical indicator is what we call a “risk predictor.” Many risks, such as poor sponsorship and change overload, can threaten to disrupt change efforts. These risks tend to occur in predictable patterns over the life cycle of a change effort, but only a handful will determine success or failure at each stage. A risk predictor can help a company understand the unique risk profile of an initiative. It can help identify the four or five risks that pose the biggest threat, the sequence in which they will arise, and the tools that will be most effective for containing and managing each one. When a decision gets stuck, it’s often because decision makers have been blindsided by some unexpected development. Tracking leading indicators can arm them with the information they need to make better choices if and when the risks materialize.

They find the right decision-making style

During times of change, leaders often find they need to modify their accustomed style of decision making in order to handle the volume and difficulty of decisions they must make. Companies that rely on consensus decision making, for example, often discover that consensus doesn’t work during a transformation effort: People are too emotional and reaching a consensus is takes too much time. Yet a directive style—decisions made by one person with little or no input from anybody else—may undercut change by reducing buy-in. Most of our clients find that a participative style with clear decision roles works best, but even then decision makers may seek out less input than usual. What’s most important is that leaders spell out the decision style they intend to use, so that everyone knows what to expect.

Ultimately, the success of any change effort depends on the behavior of people—those on the front line and elsewhere—who must execute all those decisions. What they do reflects how well the company communicates its decisions. Studies have shown that in times of stress, people can take in only a small fraction of the information they would usually be able to process. Communication has to be brief, positive and delivered face-to-face by trusted messengers, preferably by an individual’s direct supervisor. It also has to be timely, so that it can influence people’s perceptions before they become solidified into beliefs. Without this kind of communication, too many decisions will languish unimplemented, and the change effort is likely to stall.

Change leaders, like our fictional Andrea, have a tough job because the results they hope to achieve depend on many different decisions made by many different people. Good strategies for making and executing those decisions won’t guarantee the success of a transformation effort. But poor decision making and execution are sure to undermine even the most carefully planned change.

Patrick Litre is a partner in Bain & Company’s Atlanta office. Paul Rogers is the managing partner of Bain’s London office and leads Bain’s Global Organization practice.