Examining Amazon’s Return to Office Mandate

Anthony Coppedge
10 min readJan 3, 2024

Working in an office or working remotely isn’t a binary this-or-that solution. But with the so-called ‘return-to-office-mandates’ going around on social media from larger companies in particular, it is worth asking better questions about the value of these locale options so that the decisions made have clear data and insights to back up why an abundance of senior leaders are making these closed-and-shut cases instead of “pick what works best for your goals and business outcomes.”

To be clear: this is not a post about the pros and cons of either approach. What this post is about is thinking about how to ask better questions with both quantitative and qualitative data to inform the decisions.

In my opinion, the current situation at Amazon is the most clear example of the this-or-that approach of working in a company office versus working from anywhere. In an internal memo to Amazonians, current CEO Andy Jassy made the statement that went counter to the core culture of Amazon:

“It is past the time to disagree and commit…if you can’t disagree and commit, it’s probably not going to work out for you at Amazon,” said Jassy in late August of 2023.

While it’s fairly easy to disagree with the tone and even the message, it’s entirely within the scope of a CEO to make such decisions. The question is rather about the impact of his decision, and how he framed it, in light of what is a serious value of Amazon: “Make decisions with data.”

“Data without action is useless; action without data is roulette.” — Unknown

A core value and sentiment of Amazon has been the practice and belief that Amazonians should “make decisions with data.” So when audio was leaked of an employee asking Jassy about the return-to-office mandate, Jassy admits that he has no “perfect data,” but cites conversations with “60 to 80 other CEOs,” all of whom prefer in-office work to remote work. An Amazon SVP, Mike Hopkins, doubled down saying of in-office work, “I don’t have data to back it up, but I know it’s better.”

Question: If ‘make decisions with data’ is a core value or principle, wouldn’t such a weighty decision be most respected and well-received when it’s made with data?

The question to ask here is not if they could make this decision and double down on their office preference without data, but if they should have done so. After all, if the company has been successful by ‘making decisions with data,’ it comes across as tone-deaf to flippantly disregard the lack of data for their RTO decision when data for decision-making has been a key part of Amazon’s culture and success.

Again, I’m not here to pick on Amazon; I’m merely using a well-known and oft-quoted situation that sums up the crux of this issue around in-office and remote work.

Question: “Are executives making decisions about the importance and value of in-office requirements mainly out of preference over performative data?”

The reasons cited by Jassy back in early 2023 were outlined in another internal memo, now made public:

“It’s easier to learn, model, practice, and strengthen our culture when we’re in the office together most of the time and surrounded by our colleagues.” — Andy Jassy, CEO of Amazon

Question: “Is an organization’s culture primarily defined by the osmosis of physical proximity or the value of relational proximity?”

Having been in dozens of offices over the years across many different organizations, it’s easy to observe how cliques form and how being in a shared office space — heck, even in a shared swath of cubicles — and how people in close physical proximity are emotionally and relationally far apart. Relational proximity — the way people choose to interact, communicate, and share — is a hallmark of the human condition and may be bolstered by physical proximity, but is not defined by it.

Jassy goes on to bolster his assertion with some observations:

  • “When you’re in person, people tend to be more engaged, observant, and attuned to what’s happening in the meetings and the cultural clues being communicated.”
  • “It’s also easier for leaders to teach when they have more people in a room at one time, can better assess whether the team is digesting the information as intended; and if not, how they need to adjust their communication.”
  • “For those unsure about why something happened or why somebody reacted a certain way, it’s easier to ask ad-hoc questions on the way to lunch, in the elevator, or in the hallway; whereas when you’re at home, you’re less likely to do so.”

Generally speaking, I’d agree with the first two observations. However, how many organizations have 100% dedicated teams with assigned individuals all co-located together?

In today’s knowledge work, teams are often comprised of people distributed across time zones, so Jassy’s assertion is primarily based on co-located groups and, when possible, teams of people. So it is unlikely that those employees not co-located with their team(s) will get the same value out of in office as those who work in the same office together.

Jassy’s third point is based on the introversion or extroversion of the individual and their relational dynamics with those in the same office as opposed to asynchronous responses asked via Slack, for instance.

“Collaborating and inventing is easier and more effective when we’re in person.” — Andy Jassy, CEO of Amazon

Question: “Since collaboration doesn’t happen without coordination, and coordination only happens with consistent communication, how does an organization need to alter communication systems, processes, and feedback loops in order to make invention and innovation possible?”

Jassy shared his personal experiences where physical proximity led to breakthrough moments:

  • “Interjecting happens less frequently in virtual calls because it blocks out all speakers when it transpires.”
  • “Teams working on new ideas often find that a whiteboard enhances the group understanding and iterating.”
  • “Some of the best inventions have had their breakthrough moments from people staying behind in a meeting and working through ideas on a whiteboard, or walking back to an office together on the way back from the meeting, or just popping by a teammate’s office later that day with another thought.”
  • “Invention is often sloppy. It wanders and meanders and marinates. Serendipitous interactions help it, and there are more of those in person than virtually.”

The common denominator for each of Jassy’s observations is based on interpersonal communication. This is related directly back to the point above about relational proximity, though the Amazon CEO is tying these communication breakthrough moments to physically shared space and time.

As a facilitator for both large and intimate brainstorming, ideation, design thinking, and process mapping meetings, I can say with certainty that the best ideas almost always come from strange or so-called ‘bad ideas.’ This is not new and has been codified in books for decades, my personal favorite among them being Six Thinking Hats by Edward DeBono. And I’d also say that, when possible, in-person meetings and workshops that allow for the space and time to think, question, ask, marinate, reconsider, challenge, and clarify are notably more productive and effective than almost any virtual session when all attendees are fully present and not distracted by laptops, cell phones, or other people demanding their time and attention.

Yet I must ask the question:

“How often do these meetings/workshops/sessions happen and for what duration?”

Because if they’re not frequent, then co-location for this purpose, at least, isn’t a good reason for required office attendance. Even after these sessions are complete, there’s still the deep work needed to actually do the work and coordinate and communicate outside of the team. That kind of dedicated, head-down, uninterrupted work needs time and space, both of which are not easily realized in a large open office.

“Learning from one another is easier in-person.” — Andy Jassy, CEO of Amazon

Question: “What is the cost for easier in-person learning? And how is that cost compared to the productivity cost for dedicated, uninterrupted deep work?”

Jassy offers his additional opinions to support his point of view around the ease of learning in person together:

  • “Being able to walk a few feet to somebody’s space and ask them how to do something or how they’ve handled a particular situation is much easier than Chiming or Slacking them.”
  • “We have a lot of functions and roles where learning from peers is very useful and critical. And, our newer employees, especially those who have joined us in the past few years, stand to be most disadvantaged by not having the learning and mentorship opportunities from peers that many of us who joined much earlier had.”

As someone who has been onboarded several times and onboarded others on numerous occasions, there is absolute value in face-to-face, in-person shadowing. And yet, if the process is not well documented, repeatable, and scalable, then it’s a high-cost context switch for both the person leading the onboarding and the person being onboarded. Said simply: no amount of personal face-time can make up for a poorly designed onboarding system and process, so the challenge is less about physical proximity and more about practical reality.

The dark side of co-location is also in the frequency and duration of interruptions due to physical proximity. In his book Deep Work, Professor Cal Newton documents the high cost of interruptions and the related context-switching costs on productivity, including:

Time Wasted:

  • Switching Time: Each interruption requires time to switch your focus from one task to another. Research indicates that it can take an average of 23 minutes and 15 seconds to return to a task after being interrupted.
  • Cumulative Time Loss: Frequent interruptions can lead to a substantial cumulative loss of productive time throughout the day.

Reduced Productivity:

  • Decreased Efficiency: Constant context switching may result in decreased overall efficiency as it disrupts the flow of work and makes it challenging to achieve deep focus.Quality of Work: Interruptions can impact the quality of work as attention and concentration are divided among multiple tasks.

Increased Errors:

  • Mistakes and Oversights: Context switching may lead to errors and oversights in tasks, as the brain struggles to re-engage with the details of the work.

Stress and Frustration:

  • Cognitive Load: Frequent interruptions increase cognitive load, leading to higher levels of stress and frustration.Job Satisfaction: Constantly being interrupted can negatively affect job satisfaction and overall well-being.

Long-term Impact on Health:

  • Burnout: Chronic interruptions can contribute to burnout, a state of emotional, physical, and mental exhaustion.
  • Health Issues: Prolonged stress from interruptions may contribute to health problems such as headaches, fatigue, and sleep disturbances.

Project Delays:

  • Timeline Disruptions: Context switching can lead to delays in project timelines and deadlines, as tasks take longer to complete than anticipated.

Communication Breakdowns:

  • Miscommunication: Frequent interruptions may result in miscommunication or incomplete transfer of information between team members.

Impaired Decision Making:

  • Decision Fatigue: Constantly switching between tasks can lead to decision fatigue, impairing the ability to make sound judgments and decisions.

Impact on Creativity:

  • Creativity and Innovation: Interruptions can stifle creativity and innovation as it disrupts the creative thought process.

Training and Onboarding Costs:

  • Learning Curve: Constant interruptions may require additional time for individuals to get back into the rhythm of their work, resulting in a prolonged learning curve.

“Teams tend to be better connected to one another when they see each other in person more frequently.” — Andy Jassy, CEO of Amazon

Question: “People, in general, work and play better together when in physical proximity, but how will the work improve if employees don’t have the uninterrupted, distraction-free time to do deep, heads-down work?”

Again, Jassy offers his rationale:

  • “There is something about being face-to-face with somebody, looking them in the eye, and seeing they’re fully immersed in whatever you’re discussing that bonds people together.”
  • “Teams tend to find ways to work through hard and complex trade-offs faster when they get together and map it out in a room.”

Once more, the context of “team” in Jassy’s world is based on teams being 100% co-located together so that these interactions he’s referring to can happen repeatedly. I don’t disagree with his assertions at all when talking about teams of people who all work in the same space at the same time. My question is how realistic is this for the teams doing the work when the talent may very well be distributed around the globe? Perhaps Jassy’s executive privilege and perspective have the unintended bias of having his executive leaders and direct reports all within the hallways of his local office; great if you have it, but what about the vast majority of teams not co-located? What then?

In the domains of knowledge work, the challenge to the habits of the 20th century and the advantages of in-office work locales is perhaps most simply and eloquently summed up in an article by an executive recruiting company, Alumni Global:

“In building a company culture, and improving engagement and productivity, many organizations would benefit from letting go of the notion of the importance of their physical office. Instead, emphasize and consider the essence of their employee’s experience and their human needs. Employees want to feel connected to their colleagues and managers, to feel their work has meaning and impact on the company, and to be appreciated for the work they do. This can be achieved in a hybrid and remote workplace just as well as within the walls of an office.”AlumniGlobal.com

The value of co-located teams and in-office benefits for serendipitous conversations is not in question; those are nice to have potential benefits. The value being questioned is in the cost of employee retention, employee burnout, employee morale, and indeed even the employee culture.

Nothing is free and everything has a cost, so the argument is not about in-office, hybrid, or remote work. It’s about using data to understand, represent, share, and make choices based on the impact of those decisions on the employees, teams, and business in light of how much it benefits the customer. The question is instead how should executives be using both quantitative and qualitative data to inform their decisions?



Anthony Coppedge

I lead the vision for how business agility is infused in Digital Sales at IBM. I relish the chance to sabotage mediocrity.