How Remote

Work Changes
Design Thinking
SPRING 2025
ISSUE Replacing onsite design-thinking sessions with virtual ones
fundamentally changes the innovation process and outcomes.

Daniel Wentzel
Alice Minet
Stefan Raff-Heinen
Janina Garbas

Vol. 66, No. 3 • Reprint #66312 • sloanreview.mit.edu

 T H E I N N OVATO R’ S E D G E
[ Collaboration ]

How Remote Work

Changes Design Thinking
Replacing onsite design-thinking sessions with virtual ones fundamentally
changes the innovation process and outcomes.

D
By Daniel Wentzel, Alice Minet, Stefan Raff-Heinen, and Janina Garbas

ESIGN THINKING IS A POWER- To facilitate such interactions, observational

ful method for understanding cus- workshops are typically conducted onsite in
tomer needs and developing new end users’ familiar environments or in carefully
solutions to meet them.¹ It has been arranged design studios. In recent years, however,
used by innovators to invent consumer with the rise in hybrid work, we have seen some
products like electric toothbrushes and innovation processes shift to the digital realm.²
to develop business-to-business ser- Design-thinking practitioners now frequently
vices such as customer relationship management watch consumers use products through videocon-
software. A key advantage of the design-thinking ferencing and discuss their observations on digital
process over other innovation methods is its conference boards and in group chats. Using these
emphasis on the user experience. Whether a team kinds of digital tools is certainly more convenient
is imagining a car dashboard, a tax declaration than getting people into the same room. But by
app, or an electric lawnmower, each step relies on shifting away from in-person interactions, are com-
repeated, personal interactions among team mem- panies sacrificing the essence of what makes design
bers, end users, and other stakeholders. thinking so powerful in the first place?

Taylor Callery spring 2025 sloanreview.mit.edu 29

THE
I N N OVATO R’ S Collaboration
EDGE

To answer that question, we conducted 41 semistruc- chological sense compared with having that same con-
tured interviews with design-thinking experts from lead- versation face-to-face. This, in turn, affects how people
ing companies, innovation consultancies, and academia think about the conversation: They are more likely to
and also drew on our own extensive experience running form an abstract mental representation of what is being
design-thinking projects.³ One key insight that emerged said, devoid of specific details, and focus instead on the
from our research is that the transition from physical to big picture. Moreover, the body and mind are intimately
virtual is much more than just a change of medium. It connected: The way we think is strongly influenced by
fundamentally changes team members’ experience of the all of the information acquired by our bodily senses, our
design-thinking process and the outcomes they generate, physical movements, and our embodied interactions with
in both positive and negative ways. Understanding these the world.⁷ Being able to touch a product leads to a dif-
changes is essential for innovation leaders to determine ferent understanding of what the product may or may
how they will apply the best features of both the physical not do compared with just seeing it or reading a product
and the virtual design-thinking formats. description. Simply put, people think not only with their
minds but also with their bodies.
Virtual Environments Change Perception Consequently, when design thinkers work in virtual
and Understanding environments, they are likely to form a different impres-
Design-thinking practice is often thought of as a process sion of end users and their latent needs than they would
that begins with team members gaining a deep under- if they were physically present with the users. They are
standing of customer problems and then imagining or also deprived of important ways to express their ideas
refining products or services as they move through dis- about solutions and iterate them with their team. In onsite
tinct phases in a sequential, iterative manner. environments, design teams often engage in collabora-
The popular framework Stanford University uses in its tive experimentation, where they use physical materi-
design program breaks the design-thinking process into als and artifacts to arrive at a common understanding of
five distinct phases: empathize (understand user needs), the design problem. They’re able to generate and han-
define (frame the design challenge), ideate (explore the dle rough prototypes of potential solutions. This kind of
potential solution space), prototype (build tangible rep- experimentation is severely hampered in virtual environ-
resentations of potential solutions), and test (gather feed- ments. Hence, such environments not only affect how
back and refine the prototypes).⁴ The phases differ in the design thinkers interact with end users but also how they
extent to which each relies on divergent thinking (nonlin- work together as a team.
ear, spontaneous, and free-flowing) or convergent thinking
(linear, structured, and rule-based).⁵ Generally speaking, How Virtual Environments Affect Each
research on design thinking shows that empathizing with Design-Thinking Phase
end users and ideating novel solutions is better served by Using digital tools in a virtual environment affects all
divergent thinking, whereas defining the design challenge, phases of design thinking but in different ways. It’s par-
prototyping, and testing benefit from convergent thinking. ticularly important to understand its limitations and
We wanted to understand how the shift from a physi- opportunities.
cal to a virtual environment affects the individual phases In the empathize phase, digital tools constrain
as well as the design-thinking process as a whole. Cer- insights. To build a rich and comprehensive understand-
tainly, the most important change concerns the loss of ing of users’ needs, teams often choose qualitative, eth-
direct, physical experience, both when observing poten- nographic research methods, which involve observing
tial customers and, later, when figuring out what those users in their natural environments and fully immersing
observations mean. In a virtual environment, design themselves in users’ day-to-day experiences.
thinkers interact with users and one another through Our findings show that the application of digital tools
screens. Instead of writing down their thoughts on phys- provides poorer insights in this phase. When trying to
ical Post-it notes that they can arrange on a wall, they understand users’ problems, observing what users do is
type them on virtual notes, which sit on shared com- equally important, or even more important, than listening
puter screens; instead of building physical prototypes, to what they say. Nonverbal information — an enthusias-
they sketch them with digital tools. tic nod, a bored posture, a weary sigh — is largely missing
Research has shown that the loss of physical experi- in a virtual interaction.
ence has profound implications for one’s psychological Design thinkers rely heavily on what users tell them.
experience.⁶ For example, having a conversation with This may be problematic in virtual environments for two
someone through Zoom will feel more distant in a psy- reasons. First, users can only verbalize problems they

30 MIT Sloan Management Review spring 2025

 understand, not those they either don’t see or don’t rec- not only users’ verbal statements from the interviews but
ognize as areas for improvement. Consequently, some also information about their body language, pictures of
opportunities for innovation may go completely unno- their home or their possessions, and video recordings
ticed because design thinkers cannot infer these needs of them using a product, service, or software. Bring-
through the screen. A senior innovation manager from ing together many disparate data points into a coherent
Ford Motor Co. noted, “It is so much harder to get a whole and separating the wheat from the chaff is often a
grasp on the actual usage situation in a digital setting. challenge. Design-thinking teams must be careful not to
You depend so much on what the user will tell or show get bogged down in any single observation.
you compared to a setting where you can see things for Our interviews across various product and service
yourself and then start asking questions.” contexts — such as building materials and software sub-
Second, talking through a screen not only affects scriptions— show that digital tools help design teams
what is said but how it is said. Users open up to team transition through the define phase more effectively and
members less frequently when they cannot interact with efficiently. In this phase, design teams need to struc-
them directly. “It’s really difficult to fully see the other ture the data, cluster key findings, and synthesize their
one’s perspective,” said one of our interviewees, a leading ideas in succinct problem statements. Shared digital
design-thinking researcher from the Delft University of whiteboards help visualize these workflows and facil-
Technology (TU Delft). “How can you immerse fully if itate documentation, allowing for easier collaboration
there is just a digital screen in front of you and you can’t in identifying common themes and patterns in the data.
sense, smell, or see the full person?” In these scenarios, In comparison, organizing this information on physi-
design thinkers gain a less rich and more abstract under- cal whiteboards or pinboards is more time-consuming
standing of the users’ problems. because it requires the data to be available in a physical
These limitations of digital tools may undermine format. When all of the data gathered in the empathize
the goal of the empathize phase, which is to build a deep phase is available at one digital location, team mem-
understanding of the user. Both the breadth of insights bers can also retrace their own thought processes more
(how many unmet needs are identified) and the depth easily and move more quickly into repeated learning
of insights (how well individual needs are understood) cycles. Consequently, as one design-thinking coach from
often deteriorate. Virtual formats fall short in providing the Hasso Plattner Institute School of Design Thinking
the same rich, contextualized insights that onsite inter- told us, it’s easier to get an accurate understanding of
actions do during the empathize phase, especially for the design challenge: “What people love about digital
problems that users themselves do not fully understand. whiteboards is that the whole journey is right in front
In the define phase, digital tools help in struc- of you — the whole learning experience, all the Post-its,
turing and synthesizing messy data. At this point, everything the participants have shared and thought
design thinkers begin to make sense of the vast amount about from Day 1 to the last. It’s just really good for learn-
of material collected in the empathize phase in order to ing that you can go back and reflect.”
correctly frame the design challenge — and start defining Digital tools not only affect how design-thinking
potential solutions. The data gathered typically includes teams work through their data but also how they look
at that data in the first place. During the synthesis pro-
cess, design thinkers often need to take a step back from
their findings and make sure they understand the entire
THE RESEARCH problem, not just parts of it. What’s essential in this step
▪ The authors conducted 41 semistructured interviews with design-think- is the ability to abstract from individual observations and
ing experts in 2021 and 2022. The experts were from various industries to integrate all observations into a coherent whole. As
(including Ford Motor Co., Henkel, IBM, Pfizer, Salesforce, and SAP), noted above, working with digital tools — having video
consultancies (including IDEO, Deloitte, and HYVE), and academic discussions with other team members or sketching out
institutions (including the Hasso Plattner Institute School of Design
ideas on a whiteboard on an online platform — likely fos-
Thinking and Delft University of Technology).
▪ They also conducted a field study in which design-thinking teams alter- ters greater psychological distance and a more analytical,
nated between physical and virtual formats and then reported on their abstract mindset. This can be very beneficial in the define
experiences. phase because it focuses the team’s attention on the pro-
▪ The authors have participated in design-thinking projects as part of verbial forest rather than the trees.
consulting engagements with companies including Deutsche Post DHL,
In the ideate phase, being in the same room
Deutsche Telekom, Ford, and T-Systems, which informed their analysis.
matters. When design thinkers begin exploring a vari-
ety of ideas and solutions, they often rely on techniques

spring 2025 sloanreview.mit.edu 31

THE
I N N OVATO R’ S Collaboration
EDGE

such as brainstorming and storyboarding. Visualizing that have not yet been refined.
ideas is particularly important in this phase to overcome Our research shows that shifting the prototype phase
the ambiguity of abstract, verbal explanations. However, to a virtual setting can have both adverse and beneficial
the effective use of creativity-inducing tools is severely effects. Building and exploring a physical prototype —
restricted in digital sessions. Our findings show that the whether it’s the basic interface of an app for preparing
creative thought process and the ideas that emerge from tax declarations or the complete interior of an airplane
it are adversely affected. cabin for a redesign of the long-distance flight experience
Creativity is not purely an activity of the mind. Design- — allows design thinkers to “think with their hands” and
ers often pace around a room, use pen and paper to write helps them understand whether a prototyped solution is
down ideas, and rearrange sticky notes on a board in the viable. In virtual settings, such physical explorations are
process of thinking. Physical activity can kick-start cre- severely restricted. An innovation manager from Ford
ativity, but in virtual sessions, physical activity is curbed noted, “Unless an idea has been developed into a concrete
and mostly reduced to typing. product or prototype, different people can interpret it in
Creative solutions are also more likely to emerge when completely different ways. I think this challenge is even
team members feed off of one another’s energy. The envi- greater in the digital world.” Even advanced digital tools
ronment where such interactions take place can encour- such as virtual reality systems cannot adequately repro-
age or discourage playfulness and inspiration. For team duce the experience of physically interacting with a pro-
members to figuratively think outside the box, they phys- totype. In a virtual space, it can be more difficult, too, to
ically need to leave their own box. Working from home get a read on a team member’s hesitancy. A consequence
and interacting with team members through a screen are of this lack of physical interaction is that it is more difficult
hardly inspiring and playful experiences. “Random, spon- for design-thinking teams to gain a shared understanding
taneous conversations are much less common online,” and converge on an effective solution.
noted an innovation manager at Saint-Gobain, a company However, digital tools can help to spur along the pro-
that conducts design thinking for innovative building totype phase. Modifying physical prototypes across mul-
materials, such as sustainable glass and insulation. “While tiple iterations is often a lengthy process that requires
online meetings can be convenient, they can undermine the efforts of many designers. Digital prototypes have
this process. In fact, a little inefficiency can be beneficial an advantage here because they can be rapidly amended.
for developing ideas.” This is particularly important in the early stages of the
One may be tempted to downplay the importance of prototyping process, when design-thinking teams want
the physical environment in the process of ideation. But to get quick feedback on the basic viability of an idea.
our research shows that virtual settings adversely affect The bottom line in the prototype phase is that
the results. An innovation manager from the Munich- design-thinking teams must balance effectiveness and
based consultancy HYVE observed that physical idea- efficiency: Physical prototypes allow for deeper explo-
tion sessions generate a greater number of ideas: “Not ration and more meaningful insights, while digital pro-
every idea can be brilliant, but through the law of large totypes enable quicker iteration. Therefore, for rapid
numbers, through the flow, through the quantity, you feedback on the basic viability of ideas — especially in
have a lot more possibilities to build on things and think early prototyping stages — virtual prototypes are ideal. In
about them further. This happens more often in physi- later stages, when the designers are refining the details,
cal sessions.” transitioning to physical prototypes and in-person ses-
Our findings show that participants in virtual sessions sions is more appropriate.
not only generate fewer ideas but also tend to produce In the test phase, virtual and physical tools
more abstract ones. These effects mirror those observed are both important. When the design-thinking team
in the initial empathize phase. Given that the core aim assesses the prototypes for their practical usefulness,
of design thinking is to create effective, actionable, and often in direct interaction with end users, experimenta-
human-centered solutions, a shift to virtual formats is a tion and iteration are very important.
significant drawback for ideation. Again, our findings point to positive and negative
In the prototype phase, designers need to han- effects in shifting the test phase to a virtual format. Similar
dle physical objects; at the same time, digital tools to the experience in the prototype phase, digital testing
are good for iterating. The prototype phase focuses environments limit the potential for physical exploration.
on creating tangible representations of the generated A user’s experience with a digital prototype is not as rich
ideas. A key advantage of prototyping is that it enables as it would be with a physical one. For instance, a proto-
design thinkers to visualize and experiment with ideas type for a new automotive head unit — a dashboard touch

32 MIT Sloan Management Review spring 2025

How to Set Up Hybrid Design-Thinking Processes
Effectively combining physical and virtual formats throughout the design-thinking process allows
innovation leaders to harness the distinct advantages of each setting. A strategy that incorporates
both modes is more likely to yield innovations that are truly focused on user needs.

DESIGN-
THINKING DOMINANT
PHASE THINKING STYLE VIRTUAL VERSUS PHYSICAL ACTIVITIES HOW TO WORK

Empathize Divergent Although they bridge geographical distances and If including geographically dispersed users is not
incorporate a more diverse range of users, virtual essential for the innovation challenge, go physical.
formats fall short in providing the same rich insights
as onsite interactions, especially for problems that
users cannot verbalize.

Define Convergent Compared with physical environments, virtual Go virtual.

environments can enhance the effectiveness and goal
orientation of this phase by helping to separate the
important from the unimportant and allowing the
team to see the big picture.

Ideate Divergent Virtual sessions not only generate fewer ideas but Go physical.
also tend to produce more abstract ones. Given that
the core aim of design thinking is to create effective,
actionable, and human-centered solutions, a shift to
virtual formats is a significant drawback for ideation.

Prototype Convergent Design-thinking teams face a trade-off between For quick feedback on the basic viability of an idea,
effectiveness and efficiency: Physical prototypes allow especially in the early prototyping stages, go virtual.
for a richer exploration, whereas digital prototypes In later prototyping stages, when refining details,
allow the team to move faster through different go physical and switch to tangible prototypes and
iterations. onsite sessions.

Test Convergent While virtual testing limits the depth of feedback, it To allow users to deeply experience the prototype
also enables a more analytical, goal-oriented testing in real-world conditions, go physical. To process
procedure and the collection of unbiased responses. user feedback, go virtual, because it better
facilitates structured and honest analysis.

Source: Daniel Wentzel, Alice Minet, Stefan Raff-Heinen, and Janina Garbas

screen to control the audio system and apps — could be preted differently from what you intended.”
tested virtually, with users clicking through digital mock- At the same time, digital tools can help address a chal-
ups to navigate through the system as they would use lenge teams often face in testing prototypes: Our research
the screen in a car. However, a digital prototype would has found that onsite testing is often affected by factors
not allow users to understand whether the head unit that have little to do with the prototype itself. In onsite
would be easy and safe to operate while driving. Hence, presentations, design thinkers may describe a prototype
users — and designers themselves — may find it difficult very enthusiastically or defend a prototype against users’
to tell whether and to what extent the prototype could criticisms, which can lead to biased user assessments of
be the basis for an effective solution. Moreover, virtual the prototype’s actual desirability. Because virtual set-
settings restrict the type of information design thinkers tings typically trigger a more analytical and goal-driven
can collect. As in the empathize phase, the feedback that mindset — or at least one less influenced by an in-person
teams will get from digital testing will be mostly verbal presenter — digital testing procedures may help designers
in nature, and users’ nonverbal reactions will largely be overcome this bias and gain a more objective assessment
lost. A design-thinking researcher from TU Delft told of a prototype’s appeal.
us, “Digital testing is very shallow compared to real-life So the data about which mode is best in the test phase
testing. When you can’t see someone using a prototype is nuanced. Our research shows that while virtual testing
in context, you will never see how it is misused or inter- can limit the depth and meaningfulness of feedback, it

spring 2025 sloanreview.mit.edu 33

THE
I N N OVATO R’ S Collaboration
EDGE

also enables the collection of responses that are less likely tured and goal-oriented analysis. Hence, combining phys-
to be biased by the presenter. ical and virtual formats during the test phase may yield
better results than relying on a single format throughout
The Right Mix of Virtual and Physical the phase. (See “How to Set Up Hybrid Design-Thinking
The digitization of innovation processes is increasing at Processes,” p. 33.)
a rapid pace. But our research has revealed that reducing To get the most out of their design-thinking processes,
the question of virtual or physical to an either-or choice companies must also look at them dynamically. Effective
is too simplistic. Each mode has specific advantages and design thinking is not a linear process — it goes through
disadvantages, opportunities and pitfalls. Neither format multiple iterations. Its iterative nature offers design think-
is inherently superior to the other, and both offer some ers the chance to benefit from the unique advantages of
advantages that benefit the process. The key question is physical and virtual formats. In the initial rounds of iter-
not whether to prioritize one format over the other but ation, designers may be better served by building digital
rather how to combine physical and digital design-think- prototypes to gain quick feedback on the basic viability
ing practice for maximum impact. of an idea. In the later rounds of iteration, the viability of
Our general guideline is that innovation leaders should an idea is already understood, and design-thinking teams
set up hybrid design-thinking processes. Overall, the data may be well advised to build physical prototypes to gain
indicates that the phases requiring divergent thinking a better understanding of an effective solution.
(empathize and ideation) are generally best conducted Design thinking is a powerful tool for innovation and
in physical formats whereas phases involving convergent can help companies develop products and services that
thinking (define, prototype, and test) may be more effec- address untapped customer needs. This human-centered
tively completed through virtual formats. approach is not at odds with an increasingly digitized
Effective hybrid processes optimize how design teams approach to innovation. Quite the contrary: Effectively
interact with end users and how team members interact combining physical and digital tools will allow compa-
with one another. For example, understanding a user’s nies to harness the distinct advantages of each approach
needs, motivations, and emotions in the empathize phase and come up with innovations that are truly focused on
requires some physical immersion in the user’s environ- user needs. ▪
ment. Running this phase only through digital tools such Daniel Wentzel is a professor of marketing at RWTH Aachen
as Zoom or Teams risks not understanding the full scope University. Alice Minet is a doctoral student at RWTH Aachen
of a problem or, of greater concern, missing a problem University. Stefan Raff-Heinen is a professor of technology man-
altogether. While digital tools may be the only viable agement at the Bern University of Applied Sciences and a research
affiliate at the MIT Sloan School of Management. Janina Garbas is
option in some cases (such as when users are geograph-
an assistant professor of marketing at ESCP Business School.
ically dispersed), companies should generally attempt
to run the empathize phase in face-to-face settings. In REFERENCES
1. T. Brown, “Design Thinking,” Harvard Business Review 86, no. 6 (June
contrast, in the define phase, digital tools such as Miro 2008): 84-93.
or Mural can help design-thinking teams synthesize the 2. E. Tippmann, P.S. Scott, and M. Gantley, “Driving Remote Innovation
fuzzy data from the empathize phase — interview tran- Through Conflict and Collaboration,” MIT Sloan Management
Review, April 15, 2021, https://sloanreview.mit.edu; L. Thompson,
scripts, personal notes, photographs, and video record- “Virtual Collaboration Won’t Be the Death of Creativity,” MIT Sloan
ings — and recognize meaningful patterns more readily Management Review 62, no. 2 (winter 2021): 42-46.
compared with having to organize this information with 3. A. Minet, D. Wentzel, S. Raff, et al., “Design Thinking in Physical and
Virtual Environments: Conceptual Foundations, Qualitative Analysis,
whiteboards or pinboards. Analytical and goal-oriented and Practical Implications,” Technological Forecasting and Social
mindsets are tremendously helpful here, and digital tools Change 207 (October 2024): 1-14.
will facilitate such mindsets. 4. “An Introduction to Design Thinking Process Guide,” PDF file
(Stanford, California: Hasso Plattner Institute of Design at Stanford
There are caveats, of course. On a more granular level, University, n.d.), https://web.stanford.edu.
companies need to recognize that some phases consist of 5. D. Dunne and R. Martin, “Design Thinking and How It Will Change
subactivities that may each require a different approach. Management Education: An Interview and Discussion,” Academy of
Management Learning & Education 5, no. 4 (December 2006): 512-523.
In the test phase, for example, design thinkers need to
6. Y. Trope and N. Liberman, “Construal-Level Theory of Psychological
provide users with some form of a prototype and they Distance,” Psychological Review 117, no. 2 (April 2010): 440-463.
need to process user feedback on that sample. While the 7. A.M. Glenberg, “Embodiment as a Unifying Perspective for
Psychology,” Wiley Interdisciplinary Reviews: Cognitive Science 1, no. 4
first activity (interaction) is best realized through phys-
(July/August 2010): 586-596.
ical prototypes that users can explore and interact with,
the second activity (data collection and review) may be Reprint 66312. For ordering information, see page 4. Copyright ©
better served through virtual tools that allow for a struc- Massachusetts Institute of Technology, 2025. All rights reserved.

34 MIT Sloan Management Review spring 2025

PDFs ■ Reprints ■ Permission to Copy ■ Back Issues
Articles published in MIT Sloan Management Review
are copyrighted by the Massachusetts Institute of
Technology unless otherwise specified.

MIT Sloan Management Review articles, permissions,

and back issues can be purchased on our website,
shop.sloanreview.mit.edu.

Reproducing or distributing one or more MIT Sloan

Management Review articles requires written
permission.

To request permission, use our website

sloanreview.mit.edu/store/faq or
email [email protected].