Anyone who has tried to make even a little bit of change happen inside an established organization — much less engineer a company-wide transformation — has likely experienced their share of resistance.
In the new book How Change Really Works, co-authors Julia Dhar, Kristy Ellmer, and Philip Jameson write that “when organizations struggle with change, they usually do so not because leaders have a poor strategy…but because they don’t focus enough on how executives and employees are likely to behave, feel, and think throughout the process.” The book delves into the behavioral science behind why people resist or embrace changes at work. Dhar heads the Behavioral Science Lab at BCG; Ellmer is a Managing Director at the firm; and Jameson is a senior expert in organizational change.
In this excerpt, they caution that “take up” of change among your colleagues is never automatic, but must be earned. Then they present five steps for earning it.
• • •
Etsy, the New York-based e-commerce marketplace, connects artisans to the global economy. Every year, some eight million makers of handmade and vintage goods connect with around 96 million people looking for one-of-a-kind products listed among around 100 million items. Founded in 2005, Etsy harks back to a preindustrial age when products were not massproduced by factory workers but rather were crafted by skilled artisans. The sentiment is captured in Etsy’s mission: “Keep commerce human.”
And yet, while Etsy has developed an undeniably alluring “yesteryear” brand, its extraordinary success in a business dominated by behemoth retailers such as Amazon and eBay rests on its ability to stay at the forefront of the technology race. This has not been easy. Time and again, Etsy’s digital platform has been tested, as the company has strained to keep pace with the growing popularity of its marketplace.
Its first big test came in 2010, when Etsy’s gross merchandise sales reached $314.3 million, up from $180.6 million the previous year—a 74 percent jump. With such a jump in demand, Etsy needed urgently to upgrade its digital infrastructure.
At the best of times, every company’s digital infrastructure is susceptible to disruptive outages, when the whole online platform crashes and customers are temporarily unable to access the company’s services.
With every outage at Etsy, it was not only the company’s business that was negatively affected—so were the masses of solo entrepreneurs and small businesses that relied on Etsy’s marketplace. “Put frankly, outages suck,” acknowledged John Allspaw, Etsy’s then-Vice President of Technical Operations. “They suck for shoppers and sellers, and they suck for us who work here at Etsy.” There was, however, no easy fix to the problem. The very act of upgrading—introducing new, improved software through a process known as code deployment—increased the likelihood of further outages. How could the system be more stable and more resilient, while still allowing for necessary change and improvements?
Answering this question was Allspaw’s no. 1 task. Newly recruited from Flickr, the image and video hosting service, he needed to find a way to make Etsy’s digital platform more robust. One of his options was to create a culture where there was zero tolerance of failure—where engineers guilty of making a coding error would be singled out, made an example of, and either fired or required to undergo further training.
But Allspaw didn’t take this option. A student of behavioral science, he understood that such a culture of fear would be counterproductive. As he put it: “What do you do with those careless humans who caused everyone to have a bad day? Maybe they should be fired. Or maybe they need to be prevented from touching the dangerous bits again. Or maybe they need more training. This is the traditional view of ‘human error,’ which focuses on the characteristics of the individuals involved. It’s what [Dutch behavioral scientist] Sidney Dekker calls the ‘Bad Apple Theory.’ Get rid of the bad apples, and you’ll get rid of the human error. Seems simple, right? We don’t take this traditional view at Etsy.”
Under Allspaw’s leadership, Etsy’s technologists viewed the human errors—the failures—as a learning opportunity. Firing the person responsible for the “mistakes, errors, slips, lapses, etc.,” as he put it, was no guarantee that the digital platform would be any more reliable than it was before. Getting them to admit their errors, explain how it happened, and help develop solutions was a much better way.
But this brought up a question. How could Allspaw get his engineers to adopt this new collaborative behavior? How could he solve the behavioral puzzle we call “take up”? Take up is precisely what it sounds like: getting people to do things we want them to do. Just as you cannot force a customer to buy your product or service, you cannot force an employee to buy into the change that you want to see—in this case, getting them to admit their errors. (After all, why would they do this if they thought there was a chance that they might get fired?)
Yet if Allspaw’s engineers did not take up the new collaborative behavior, the long-term sustainability of Etsy’s digital platform could be in doubt. As we will show at the end of this chapter, he found a way to lower one of the barriers to take up that so often prevent employees from doing what their leaders want them to do.
Take Up: What It Is, and Why It’s Lower Than You Hope
Consider these familiar scenes from a struggling transformation:
- Months after the release of a new digital tool, few employees are using it. In the elevators, people say, “I don’t know what happened to it. It never really caught on. It’s still on my computer, I think.”
- Despite a comprehensive review with many steering committees, the redesigned business processes of a marketing team are gathering dust. When asked, people say that the new processes really represented a “future-state ambition,” and that they aren’t ready for immediate implementation. In the same breath, the marketing team complains about the inconsistent, duplicative, and slow ways in which they are forced to work.
- Following a painful reorganization, employees are not truly working within the new structure. They are still relying on their old managers for informal approval. They continue to ask their old direct reports to do work for them. Contractors have been brought in to backfill positions that were supposedly cut.
These regrettable situations have a common cause: low take up.
The movie Field of Dreams has the famous line, “If you build it, [they] will come.” Take up is a behavioral science concept that assumes Field of Dreams was wrong. In the world and workplace, people commonly fail to take up opportunities that they are offered or take up benefits to which they are entitled.
Transformations rely on widespread take up of new behaviors. When your revenue managers use an AI-powered pricing tool you developed for them exactly as you intended them to—that’s take up. When your customer service representatives use the new script their manager wrote for them instead of making up their own version—that’s take up. When your vice presidents maintain their new headcount after a cost program, rather than backfilling with contractors—that’s also take up.
In every failed transformation we have seen or analyzed, low take up was a key feature.
In every failed transformation we have seen or analyzed, low take up was a key feature. It frustrates executives endlessly. We often hear, “Why aren’t people using the new processes designed for them? The entire steering committee agreed that they are more secure and more consistent. We made substantial investments to develop them. I don’t understand why our employees are so resistant to change. There must be something wrong with our culture.” The next time a business case for new, improved tools lands on the table, someone says, “Why bother? They won’t use it anyway.”
We understand the frustration. We understand why executives feel tempted to attribute transformation failure to their employees’ unwillingness to change. But we take a different view. Executives know they are responsible for results, and not just the design of a transformation. This means they have no choice but to care deeply about take up in their organizations. Poor take up is just as much an executive’s problem as poor design is. As Jocko Willink and Leif Babin say in their book Extreme Ownership, “Leaders must own everything in their world. There is no one else to blame.” That is not to say that there will never be outright resistance to the change that you are planning. But a good default is that a failure to make change should first be treated as a take up problem until some other diagnosis is proven.
Generally, low take up is not a consequence of employees’ character flaws, low ambition, or poor understanding of the bigger picture. Rather, it represents your own failure to win them over. Remember: if you are executing a change, employees are the customers of that change. Just as you wouldn’t blame your customers for not buying your company’s products, you shouldn’t blame employees for not buying into your change.
The natural human inclination to misread situations and attribute a person’s actions to their innate characteristics has been studied by behavioral scientists. It was Stanford psychologist Lee Ross who first identified the cognitive bias he called fundamental attribution error—the “general tendency to overestimate the importance of personal or dispositional factors relative to environmental influences.” It was later given a more straightforward description by scientists and authors Chip and Dan Heath, who described it as “our inclination to attribute people’s behavior to the way they are rather than to the situation they are in.” When executives blame employees for poor take up, they are giving in to that inclination.
Most of the time, the best explanation for low take up is that employees are human. We commonly struggle to make behavioral changes that would benefit us. Medical researchers found that one-quarter of men who suffered a stroke or coronary heart disease event did not subsequently make improvements to their lifestyle: they continued to smoke, avoid physical exercise, and maintain an unhealthy diet. In the United States, 20 percent of low-income workers who are eligible for the Earned Income Tax Credit (EITC), which is effectively free-to-them money, do not claim it, even though many receive reminders from the Internal Revenue Service about their eligibility.
Rationally speaking, take up rates of a change that will save your life or put more dollars in your pocket should be 100 percent.
Rationally speaking, take up rates of a change that will save your life or put more dollars in your pocket should be 100 percent. But however well-intentioned we may be, there are all sorts of impediments that prevent us from doing things that are unambiguously good for us. So is it any wonder that employees do not automatically participate in things that may not be so unambiguously good for them, like the initiatives in your transformation?
Successful transformation leaders do for employees what successful product managers do for customers: they earn their take up, rather than expecting it. If you have ever bought a subscription service, you already know what this looks like. When Netflix invites you to “choose your plan,” it does so by reassuring you that you don’t have to make any commitments: you can “cancel at any time.” This pledge requires the company to work extra hard to earn your business—to get you to take up its products. Netflix can’t afford to rest on its laurels after you have chosen your plan for the first time. So it offers you “endless entertainment for one low price,” it provides easy navigation (“enjoy Netflix on all your devices”), and it entices you with ad-free viewing (even the “with adverts” option says there are “less than you might think”). Netflix earns your take up by anticipating the obstacles that might impede a well-intentioned person from completing their transaction and removing those obstacles.
You should do the same with employees. When leaders take this mindset, employees notice. As one employee happily reported to us, “We got training prior to rollout, and we had extra staff, so we did not become short-staffed. My manager did a great job being warm, approachable, and helpful with any concerns.
The Science of Take Up
Over the decades, behavioral scientists have tried different ways to describe when and why people start and continue new behaviors. In 2011, Susan Michie of University College London led a team that developed a framework known as COM-B, an acronym that summarizes the researchers’ central theory: capability, opportunity, and motivation interact to generate behavior. The concepts are defined as follows:
- Capability is defined as the individual’s psychological and physical capacity to engage in the activity concerned. It includes having the necessary knowledge and skills.
- Opportunity is defined as all the factors that lie outside the individual that make the behavior possible or prompt it.
- Motivation is defined as all those brain processes that energize and direct behavior, not just goals and conscious decision-making.
Or, to simplify further: capability relates to the question, “Can I do it?”; opportunity relates to the question, “Is my environment letting me do it?”; and motivation relates to the question, “Am I inclined to do it?”
Seven Conditions for Take Up
The COM-B framework provides a strong scientific grounding in the drivers of take up. But the three concepts are relatively broad. We’re often asked, “How can I put these ideas to work in my transformation tomorrow?”
To help leaders of change, we have broken down the framework into seven component parts—the seven conditions for take up. We find that employees don’t change their behavior until all seven are met. When one (or more) of the conditions for take up is not present, it becomes a barrier to take up. The table below lists the seven parts, expressed in their positive forms (“conditions for take up”) and their negative forms (“barriers to take up”).
Experienced leaders of change will quickly see how the seven barriers to take up typically show up in transformations. Here are three scenarios we have encountered:
- Knowledge gaps. Maria is a salesperson for a steel manufacturing company. Company leaders expect all salespeople to use a new process for reporting on leads, negotiations, and contracts. But Maria has received minimal information on the new process: a three-page document with a high-level summary, and a thirty-minute briefing call. She continues doing things the old way.
- Resource constraints. Andrew is a marketing director for an airline. Company leaders expect him to improve the effectiveness of marketing emails to customers by developing and implementing a new personalization tool. The tool is designed to optimize which customers should receive which marketing emails, and when. But Andrew hasn’t been allocated any digital or IT resources from central teams; hey tell him that the soonest they can offer developers is in six months. Andrew continues doing things the old way. By the time the resources do arrive, he feels jaded and implements slowly.
- Something to lose. Raj is a supply chain manager at a confectionary manufacturer. Company leaders have developed a new tool to optimize the distribution of finished goods. They want Raj to start using it in his part of the facility. But he has made his own decisions about distribution for twelve years, and he doesn’t want to hand over decision-making to a machine. Raj continues doing things the old way.
These and the other barriers represent formidable challenges to the take up of change in your organization. For a single employee to decide to change the way they work, they need to overcome all seven. For an entire organization to change, the majority of affected employees must follow suit. Framed in this way, is it really surprising that take up of change initiatives is so poor, so often?
Five Steps for Earning Take Up
To earn take up from employees, you need a deep knowledge of the potential barriers. But you need something else too: a great process through which to apply your understanding.
Our recommended process has five steps (see the figure below). The first two steps occur only once, and the remaining three occur continuously until the take up goal is met or abandoned. Let’s work through each step in the process.
1. Define whose behavior needs to change. Each of your initiatives is composed of people who will need to do something differently to make the change successful. Be clear on exactly who they are. Write a list of their names and positions. If you can’t be specific about who needs to behave differently, it is unlikely that your initiative will succeed.
2. Specify the change in behavior that you expect. For every person whose behavior needs to change, specify exactly what you expect them to do differently. This exercise often begins with understanding what they do now. If a new behavior is required (for example, start to use a new digital tool), specify exactly what that behavior is and when or how often it needs to occur. If an old behavior needs to stop (for example, ceasing the circulation of low-value data around a department), specify that too. The clearer you can be about what you expect to see, the more likely it is that your change will succeed.
3. Identify barriers to new behaviors. Identify barriers that may prevent employees from following the new behaviors you expect. Work through the seven take up barriers and ask yourself which of them might apply in your situation. Write them down and rate their potential severity. Ideally, conduct this task informed by employee feedback gathered through surveys, interviews, or both. Approach this task with empathy. Your goal is to accurately summarize the real barriers that employees will face in making this change. Success depends on your ability to spend time in their shoes.
4. Act to minimize barriers. For every barrier you’ve identified, identify concrete actions you can take to reduce it. When you’ve finished your list, estimate the likely effectiveness and potential cost (financial and nonfinancial) of each. Based on the information you have, determine which actions represent an acceptable return on investment and which represent an unacceptable return on investment. Start implementing the former and discard the latter. If your take up expectations have changed as a result of this process (for example, because the costs of earning a high level of take up are unacceptable), update the business case for your change and share it with those in your organization who would want to know.
5. Measure take up and collect feedback. Decide how you will measure whether take up of your change is really occurring and start measurements as soon as possible. One common method is surveys. For example, if you want employees to take up a new tool, you could send them a one-minute pulse survey every week, asking how easy the tool is to use and how well it fits into their workflow. Another common method is to use existing data from your systems. For example, in the situation described above, you could ask a data analyst to count your tool’s user logins every week. In addition to measuring take up, start a process for soliciting, receiving, and reviewing employee feedback. You want to understand how they are experiencing your new tool, process, or structure. Without feedback, you won’t be able to interpret and respond to low take up results early in your change. Establish this feedback loop from day 1.
After you’ve undertaken these five steps, your work isn’t done. Take up planning is not something that occurs only once; it must be conducted iteratively. You should continuously repeat steps 3, 4 and 5 until you either achieve your take up target or lower it enough to render further action unnecessary. The duration of this cycle can vary from situation to situation. Some urgent situations—like the organization-wide deployment of critical software upgrades—will warrant a daily review. Other changes will call for a weekly or monthly cycle. In general, we caution against a take up review cycle that is longer than a month—it’s unlikely that you’ll be able to respond fast enough to evolving take up barriers experienced by employees. Scientists have found that it takes the median person—the middle in a range of people—sixty-six days for a new behavior to become automatic; your goal should be to fit at least two take up cycles into that period, so that you can positively influence employees as they consider and try out the new behavior.
The real question is not whether current behaviors are ‘good’ or ‘bad’; the question is whether they are helping the business to execute the strategy it needs to be successful under the current conditions.
This chapter has described a method for creating behavioral changes among large groups. As you execute, remember this: behaviors that need to be changed are not always bad. Often, behaviors that executives want to change may be desirable ones—but in a different business environment. For example, a company competing for share in a rapidly growing, winner-take-all market may benefit from bold, entrepreneurial behaviors from employees. But when market growth slows and the relative positions of market players settle, that same entrepreneurialism can sometimes lead to poor processes, poor communication, and unacceptably risky decisionmaking. The real question is not whether current behaviors are “good” or “bad”; the question is whether they are helping the business to execute the strategy it needs to be successful under the current conditions.
Putting It All Together: How Etsy Created Take Up
As John Allspaw, Etsy’s Vice President of Technical Operations, considered how he could make the firm’s digital infrastructure more robust, he came to the conclusion that his software engineers should, as we say, take up a crucial behavior: total frankness about their thoughts, actions, and observations in the wake of an outage. If they did this, he believed, Etsy would be able to maximize its learning from each incident and minimize future recurrences.
He knew, however, that the engineers would be naturally disinclined toward total frankness, for fear of being blamed for the outage. In our language, they had something to lose, which is one of the two motivation-based barriers to take up (the other being little to gain). As he reasoned: “An engineer who thinks they’re going to be reprimanded are disincentivized to give the details necessary to get an understanding of the mechanism, pathology, and operation of the failure.” This would be a problem for the company: the failure to understand the root cause of the outage “all but guarantees that it will repeat—if not with the original engineer, [then with] another one in the future.”
To lower the barrier to total frankness, Allspaw introduced a specific practice to drive take up of the behavior he wanted to see from the engineers: the blameless post-mortem. As he explained: “The idea of a blameless postmortem is to change the focus from the ‘who’ to the ‘how.’” He went on: “Blameless postmortems try to change the focus to ‘how did it happen’? It assumes that people do what makes sense to them at the time. If you start from the assumption that nobody comes to work to do a bad job, then you’re left with the question: Why did somebody do what they did?”
The practice is part of a broader “just culture” fostered by Etsy’s leaders to institutionalize accountability as well as the support for “anyone who has ever taken down a website” and feels “a huge amount of guilt and embarrassment,” as Allspaw described Typically, the process starts when an engineer who has made a mistake self-initiates what Etsy calls an internal PSA, or public service announcement, which tells colleagues what has happened. “It’s similar to placing a ‘wet paint’ or ‘caution—slippery floor’ sign,” Allspaw once explained. This email triggers a rigorous postmortem, which follows a formal process that Allspaw later documented in his debriefing facilitator guide. And in a sign that the practice is intended to be collaborative rather than accusative, Allspaw handed out an annual award of a three-armed knitted sweater to the engineer who made the most surprising error.
The blameless postmortems worked. “A funny thing happens when engineers make mistakes and feel safe when giving details about it,” Allspaw wrote, in a much-read blog post. “They are not only willing to be held accountable; they are also enthusiastic in helping the rest of the company avoid the same error in the future. They are, after all, the most expert in their own error. They ought to be heavily involved in coming up with remediation items.” And to those leaders who suggested that the engineers were being let off the hook, he offered this riposte: “Technically, engineers are not at all ‘off the hook’ with a blameless postmortem process. They are very much ‘on the hook’ for helping Etsy become safer and more resilient in the end.”23 With a more robust digital platform, Etsy went from strength to strength during the 2010s, and by 2024, the company was able to report $12.6 billion in annual gross merchandise sales, reflecting nearly a decade of sustained global expansion. Of course, not all of this success can be attributed to blameless postmortems. But they certainly contributed to the sense that everyone was pulling in the same direction. As Etsy showed, you can’t expect take up—you have to earn it.
Excerpted from How Change Really Works: Seven Science-Based Principles for Transforming Your Organization by Julia Dhar, Kristy Ellmer, and Philip Jameson. Published by Harvard Business Review Press. Copyright © 2026. All rights reserved.
Julia Dhar is a global authority on behavioral science in leadership and organizational change. A Managing Director and Partner at Boston Consulting Group (BCG) and founder of its Behavioral Science Lab, she advises Fortune 100 companies and governments on how to make better decisions under pressure.
Kristy R. Ellmer is a global authority on transformation. A Managing Director and Partner at BCG, she is the global lead, emeritus, for Culture and Change Management and leads the firm’s North America Transformation practice and global Behavioral Science team. She guides organizations through their largest and most complex changes, delivering multibillion-dollar value.
Philip Jameson is a senior expert in organizational change at BCG, where he serves companies undertaking transformation. He is a classical musician and was Chief of Staff for the Sydney Symphony Orchestra. Philip draws on his passions for economics and group decision-making to help leaders navigate complexity.
(Featured image by Declan Sun on Unsplash.)
You must be logged in to post a comment.