Author Talks: Return to the office, or work from anywhere?

In this edition of Author Talks, McKinsey Global Publishing’s Ramya DRozario chats with Prithwiraj Choudhury, associate professor at Harvard Business School and part of the Forbes’ 2023 Future of Work 50 list, about his new book, The World Is Your Office: How Work from Anywhere Boosts Talent, Productivity, and Innovation (Harvard Business Review, Spring 2025). Analyzing the limitations of return-to-office (RTO) mandates, Choudhury explores why work-from-anywhere models may offer more sustainable benefits for both employees and organizations, and why they could ultimately shape the future of work. Describing it as a “triple win,” Choudhury shares examples of companies that have adopted this working model successfully and offers solutions for those facing implementation challenges. An edited version of the conversation follows. You can watch the full video at the end of this page.

What makes ‘work from anywhere’ the future of work?

Work from anywhere is not the same as work from home. It’s a remote-work and distributed-work arrangement that lets individuals choose where they want to live. That is the most important distinction. They could choose the city, the town, the state, and, in some cases, even the country where they want to live.

Why is work from anywhere the future of work? I see work from anywhere as a triple win. It’s a win for individuals because they can choose where to live and choose a location based on their priorities. If I am raising a young family, I can move to a low-cost living location. I could be closer to my aging parents or to a location where I love the weather, the sports, or the cultural amenities.

For organizations, it expands their labor market access. Instead of looking for talent locally in a city, now the whole country—or the whole world—becomes the accessible labor market. There’s also a third win, which is for smaller towns and places that have experienced brain drain.

It’s a talent strategy that’s a win–win–win for individuals, companies, and society.

How do communities and governments benefit from a work-from-anywhere model?

I’ll give you two data points. Prior to the pandemic, only a couple of countries, Estonia and Barbados, used special visa programs for remote workers. Now, about 60 countries, including the United Arab Emirates, Portugal, and Brazil, do that, too. Most recently, New Zealand announced a digital-nomad visa.

Also, within countries, there are municipalities, smaller towns, and regions that are trying to attract remote workers. Just in the US, there are about 50 cities—in Michigan, Wisconsin, Ohio, Indiana, Arkansas, and Oklahoma—that I’ve studied quite deeply and that are offering financial and nonfinancial incentives to remote workers to move there. This is an interesting phenomenon because for decades, many of these places have lost young people to larger cities—the coastal cities. And now, it’s an opportunity for these places to reverse the brain drain.

Now, not all these programs will succeed. I write about the conditions needed to make such a program succeed. But it’s an exciting phenomenon that countries, states, and cities are now competing for talent, just the way companies compete for talent.

What are some examples of companies that have successfully implemented this policy?

Airbnb, Atlassian, and Nvidia, as well as many other companies, all have work-from-anywhere policies. The most interesting thing for me as a researcher has been to see how start-ups have embraced this model. Hundreds, if not thousands, of start-ups have embraced the work-from-anywhere arrangement. They’ve gone completely remote; they don’t have any offices.

Yet work from anywhere does not mean that you never have in-person interactions. That’s a very strong misconception. You can have an all-remote company, but there’s no all-remote worker. You must meet your peers and teammates periodically. For example, start-ups that are practicing work from anywhere have a retreat, an off-site-based model, where the teams or the whole company gather in person. They meet at a nice location with some frequency—could be bimonthly or quarterly—and they spend a whole week together. In the book, I present my research with GitLab. I’ve completed research with Zapier, which is also an all-remote start-up with nearly 800 workers.

There’s a relatively large company called Deel, which is all remote. SEEK and Doist are others. An example abroad is Cactus Communications in India. There are many start-ups around the world that have embraced work from anywhere.

What are the main challenges of a work-from-anywhere model?

Like any new management practice, work from anywhere has both benefits and challenges. As for the benefits, the business case is simple: It’s about attracting and retaining diverse talent from a wider geographic pool.

But there are lots of challenges, because work from anywhere is not only remote, it’s distributed—and hyper-distributed. For example, Zapier has about 800 employees, who come from more than 20 countries.

In a prepandemic world, a start-up would probably have been located in Silicon Valley. Zapier would have been competing for talent against Google and Facebook. But now, because many start-ups have embraced work from anywhere and the work arrangement is global, they are able to go to 20 countries around the world.

But work from anywhere creates challenges. I write about three main challenges: isolation, knowledge sharing, and social interactions, such as mentoring and onboarding.

I also present a lot of best practices that I have learned over the years by studying the GitLabs and Zapiers of the world. So, yes, there are new challenges, but there are also solutions.

What are a few solutions to these challenges?

A must-do new practice in work-from-anywhere companies is codifying knowledge in real time. For example, in a traditional, in-person company, you can have a lot of knowledge that’s tacit. It has not been written down, because new employees can go and tap the shoulder of a senior colleague and ask questions.

In a work-from-anywhere environment, the organization is hyper-distributed. There’s no shoulder to tap. Instead, you need to tap a code book: a common, shared place where all knowledge is codified in real time. It’s very hard to do that, because individuals love to do the work but often don’t love to write about it. We don’t like documentation, but it is an absolute must for work-from-anywhere companies. You need to have the knowledge codified in real time, because you don’t want workers accessing stale information.

Generative AI offers a great opportunity to codify knowledge. Some of the best practices that I’ve observed relate to using generative AI. A team could have its meeting transcribed by AI and then have that text codified to a common depository.

That’s just one example. The other one I spoke about earlier was, instead of meeting in an office every week, how can companies reimagine both the frequency and venue of in-person meetings?

The general principle here is that instead of mandating an RTO policy top down, teams should have the freedom and flexibility to choose the venue and frequency of in-person meetings.

I keep saying that the right unit of analysis is not the entire company or the individual. It’s the team. Every team should decide where and how frequently to meet based on its tasks, priorities, and the places where people on the team live.

What are some of the trending technologies that support a work-from-anywhere model?

I don’t want to evangelize any one technology. But you need synchronous and asynchronous communication technologies, and there are many now, including Slack and others.

Doist has some of its own technologies, such as Todoist and Twist. The real opportunity is to use generative AI not only for knowledge codification. I write about how, in this era of big data, there’s an opportunity to create what I learned from Doist: to build an information oasis out of an entire ocean of information. We can use predictive-AI models to glean the most relevant knowledge that each one of us needs every day. Between communication technologies and AI, there’s a great opportunity to leverage technology.

The other technological innovation driving work from anywhere is what’s known as “digital twin” technology. For many years, the argument has been, “Is work from anywhere only for white-collar workers? Is it a work arrangement only for people who work on a computer or at a desk?” Now, with digital-twin technology, work from anywhere can unfold in blue-collar, “semidesk” settings and deskless settings in a completely new way.

A digital twin is a combination of sensors, AI, and automation, where you create a digital, virtual replica of any operation. For example, you could have a factory, a warehouse, a hospital ward, or a system of conveyor belts at an airport. You place sensors in that operation and collect data in real time. Then you use AI applications to predict how to run that operation. Once you do that, many of the workers who used to be in the factory or the warehouse or the hospital don’t need to be there [to do their jobs]. In the book, I write about my research with Unilever in a Brazilian detergent factory and with a Turkish power generation company that has created a digital-twin headquarters to operate ten power plants.

This is not only a technological solution. It also provides a reskilling/upskilling complement. In the example of the Turkish power plant company, many of the technicians are not running turbines and boilers. They’re sitting in this digital-twin headquarters, working and interpreting with machine learning data. They are becoming “indigo workers”—a mix of blue-collar and white-collar workers. They need to know some machine learning, some data analytics, in addition to engineering and the science of running a power plant. But it’s an exciting moment where work from anywhere can expand from office-based settings to desk, deskless, and semidesk settings.

What prompted you to research this topic?

I’ve had different life experiences. I grew up partly in the UK and partly in India. I followed a traditional path in India, which is to study computer science engineering. Then I obtained my MBA from one of the Indian Institutes of Management and worked with McKinsey for five years across different industries. After that, I was responsible for a business in Southeast Asia for Microsoft. Given these life experiences, I traveled a lot. I observed how organizations and companies have managed the decision rights to geography for millions of people.

For years, companies have told people where to live, even if it means leaving aging parents behind, living in a city where you know no one or you don’t like the food or the cost of living is too high.

We’ve been constrained with respect to our geographic choices for far too long. The only logic has been that you must work in an office or you must work in a factory [located where you live]. That’s one way to organize work.

But I had been thinking for many years, “What if instead of the worker moving to the office, what if work moved to where the worker is?” And that was the genesis of the idea of work from anywhere. Around 2018, I stumbled upon work from anywhere in, of all places, the US Patent Office. I completed a study for the Patent Office that had been requested by the US Congress. That led me to a series of explorations on work from anywhere. Since the pandemic, this has become my life.

Tell us more about the evolution of work from anywhere.

The first wave of work from anywhere was an experiment conducted by companies such as GitLab and Zapier, because they wanted to access talent who did not live in their proximity. They wanted awesome, talented workers all around the world.

That phase has run now for seven, eight, ten years, and we have learned a lot. Now, there’s a second wave of work from anywhere coming, and there are three factors prompting it. The first factor is the digital-twin technology, which is now expanding the frontiers of work from anywhere to factories, warehouses, airports, hospitals, and any physical operation.

The second factor is that countries around the world are now providing incentives to remote workers, such as the digital-nomad-visa revolution. So there is policy driving this work-from-anywhere movement.

The third factor driving work from anywhere is that now, after a decade, we have documented best practices. We don’t have to reinvent the wheel with work from anywhere. Today, if a manager, a senior leader, or a consultant listening to this discussion is thinking about work from anywhere, they already have a starting base of knowledge that didn’t exist a decade ago.

How did the pandemic change the trajectory of your research?

The pandemic was instrumental in accelerating remote work, distributed work, and work from anywhere. There’s no question about that. One important effect of the pandemic was to challenge the biases and deep-rooted perceptions of managers that their company or their industry could not work in a distributed or remote fashion.

I’ve known this for a while, because I’ve studied companies like GitLab, Atlassian, and Deel, where the entire company practices work from anywhere. It’s not just the technologists. The sales, marketing, finance, and accounting teams also work from anywhere, as well as the C-suite. The pandemic made that a reality for some large, traditional companies, and many of them have stuck to that model. Others are practicing a hybrid work model, to a large extent. So the pandemic definitely accelerated the curve of adoption quite a bit.

What surprised you during your research or writing?

As an individual researcher, I was really surprised by the findings when I began this research. My initial focus was on productivity: Does work from anywhere make workers more productive or less productive?

That is the question the patent office asked me to study, and that is the question the US Congress was asking the patent office. Over the first few years, I realized that beyond productivity gains and cost savings, the most important and exciting aspect of work from anywhere is the talent strategy.

It opens up access to talent who are not able to live in that company’s location. When I realized that, I understood why this would be a permanent change, because the labor market would correct for any temporary fluctuations in policy. Even if there are two more companies forcing RTO tomorrow, there will also be competitors that will not do the same.

The competitors will poach some of the talent from the company forcing RTO. They will ultimately force an adjustment in the talent policy at the RTO company, because no company can avoid attracting and retaining talent for too long. As a labor economist, I view this as a sustainable, permanent change in how companies manage workers, because it’s a talent strategy.

At a time when RTO mandates seem to be gaining momentum globally, what is the future of work from anywhere?

The research is still emerging, but it suggests that RTO mandates do not lead to positive outcomes. The research I’ve seen from the University of Pittsburgh and other places is that RTO mandates are usually followed by a negative reaction on the stock price and by large-scale attrition of talent.

One of the interesting studies looks at a very large percentage of company employees looking for jobs after an RTO mandate. Many of these RTO mandates are in response to other incentives, not talent attraction incentives. They might be responding to incentives around layoffs without paid full benefits. There could be incentives around real estate holdings and investments of certain stakeholders in companies.

Ultimately, any organization must come back to a talent strategy. Again, as a labor economist, my prediction is these RTO mandates can be a blip, but they cannot sustain in the long term. No company can say, “We don’t want talented people and diverse talent.”

The final reason is that my research has shown that the unit of analysis is not the company. The right work arrangement for the sales team may not be the right arrangement for the IT team, marketing team, legal team, or compliance team, which makes these top-down, one-size-fits-all mandates meaningless. They don’t account for the differences between teams. The solution is to have some very open-ended policies top down and then let teams decide, “What is the frequency of our meeting and what is the venue?”

Watch the full interview

Visit Author Talks to see the full series.

Learn more