A report from the

Deloitte AI Institute and the

Deloitte Center for Integrated Research

Becoming an AI-fueled
organization
Deloitte’s State of AI in the Enterprise, 4th Edition
About the Deloitte AI Institute

The Deloitte AI Institute helps organizations connect all the different dimensions of the robust, highly
dynamic, and rapidly evolving AI ecosystem. The AI Institute leads conversations on applied AI
innovation across industries, using cutting-edge insights to promote human-machine collaboration in
the Age of WithTM.

The Deloitte AI Institute aims to promote dialogue about and development of artificial intelligence,
stimulate innovation, and examine challenges to AI implementation and ways to address them. The AI
Institute collaborates with an ecosystem composed of academic research groups, start-ups,
entrepreneurs, innovators, mature AI product leaders, and AI visionaries to explore key areas of
artificial intelligence including risks, policies, ethics, future of work and talent, and applied AI use cases.
Combined with Deloitte’s deep knowledge and experience in artificial intelligence applications, the
institute helps make sense of this complex ecosystem and, as a result, delivers impactful perspectives
to help organizations succeed by making informed AI decisions.

No matter what stage of the AI journey you’re in—whether you’re a board member or C-suite leader
driving strategy for your organization, or a hands-on data scientist bringing an AI strategy to life—the
Deloitte AI institute can help you learn more about how enterprises across the world are leveraging AI
for a competitive advantage. Visit us at the Deloitte AI Institute for the full body of our work, subscribe
to our podcasts and newsletter, and join us at our meetups and live events. Let’s explore the future of
AI together. L
​ earn more.

About the Deloitte Center for Integrated Research

The Deloitte Center for Integrated Research (CIR) offers rigorously researched and data-driven
perspectives on critical issues affecting businesses today. We sit at the center of Deloitte’s industry and
functional expertise, combining the leading insights from across our firm to help leaders confidently
compete in today’s ever-changing marketplace.

Connect
To learn more please visit www.deloitte.com/us/cir.
Contents

Foreword: Becoming an AI-fueled organization 2

Executive Summary: Constantly transforming, never fully transformed 3

Strategy: What should your north star be? 7

Operations: How do you bring transformation into everyday work? 11

Culture & Change Management: Why is valuable change so elusive? 14

Ecosystems: How should you orchestrate your partnerships? 18

Our AI-fueled future: The pathway is clear 20

Methodology 21

Endnotes 23
Becoming an AI-fueled organization

Foreword: Becoming an
AI-fueled organization

R
APIDLY TRANSFORMING, BUT not fully Algorithms can independently balance financial
transformed—this is our overarching portfolios. Support centers often know customers’
conclusion on the market, based on the problems before they call.2 And these are still
fourth edition of our State of AI in the Enterprise early days.
global survey. Very few organizations can claim to
be completely AI-fueled, but a significant and In combination, these developments help enable
growing percentage are starting to display the businesses to increasingly liberate themselves from
behaviors that can get them there. the time constraints of human rhythms. Core
business operations can meet customer needs at a
To us, this is exciting and has reinforced our belief faster pace, while freeing up time and energy for
that now is one of the most opportunity-rich the workforce to use new tools to discover
periods in the history of AI technology. innovative avenues for value creation. Conversely,
Organizations are swiftly building capabilities and for those organizations lagging in AI capability
reaching enterprise scale: In fact, more than a development, it could pose an ever-increasing risk
quarter of our survey respondents have reached to their competitive viability in the not-too-distant
full-scale deployment of five or more types of AI future. The massive global disruptions over the last
applications within their organization.1 This year have only accelerated these trends beyond our
widespread enterprise experimentation has set a most aggressive predictions.3
solid foundation for many, making way for what
we believe could be a bumper crop of meaningful Fortunately, we’ve learned in recent years about
advancements and impact over the next few years. which practices can accelerate transformation, and
This is especially true for those who are already this knowledge can help fast-track outcomes. The
beginning to use AI to solve some of their most findings in this report aim to support organizations
business-critical and challenging problems. in navigating through the growing pains, in
whichever stage they may find themselves on the
Within just the last 18 months, AI capabilities have journey to becoming an AI-fueled organization.
advanced considerably, maturing from what was
often experienced as a bothersome critic—telling The Age of With is no longer on its way—it has
workers what to do or pointing out their mistakes— arrived. We hope you’ll join us as this story
to more frequently serving as a copilot, continues to unfold.
independently executing on insights and trends — Jason Girzadas, managing principal of
surfaced through the power and speed of cloud- Businesses, Global, and Strategic Services,
Deloitte LLP
based data hosting and computation. Today, some
supply chains are managing themselves.

2
 Deloitte’s State of AI in the Enterprise, 4th Edition

Executive summary:
Constantly transforming,
never fully transformed

S
INCE 2017, DELOITTE has documented the To learn how organizations across the globe are
increasing adoption of AI across the progressing toward this vision, we surveyed 2,875
enterprise. The third edition, published in executives from 11 top economies who have
2020, declared that we had entered the “age of purview into AI strategies and investments within
pervasive AI.” Pervasive AI adoption, however,
4
their organizations. We asked them about a wide
does not mean that AI is being used to its full variety of behaviors—from their overarching AI
potential. And so, with the fourth edition of our strategy and leadership, to their technology and
global State of AI in the Enterprise report, we data approaches, and how they are helping their
explored the deeper transformations happening workforce to operationalize AI. Then, to
inside organizations that are using AI to drive understand which behaviors lead to the greatest
value. In other words, we wanted to know: outcomes, we analyzed the survey responses based
What are today’s most AI-fueled on how many types of AI applications a company
organizations doing differently to has deployed full-scale and the number of
drive success? outcomes achieved to a high degree (figure 1).

AI-fueled organizations leverage data as an asset to

deploy and scale AI systematically across all types
of core business processes in a human-centered
way. They use the power of rapid, data-driven
decision-making to enhance workforce and
customer experiences to achieve competitive
advantage and continuously innovate.5

“Becoming an AI-fueled
organization is to understand
that the transformation process
is never complete, but rather a
journey of continuous learning
and improvement.”
— Nitin Mittal, Deloitte AI coleader,
principal, Deloitte Consulting LLP

3
Becoming an AI-fueled organization

FIGURE 1

An organization’s AI maturity can be profiled based on the number of

applications deployed and outcomes achieved
Transformers Pathseekers Underachievers Starters Total

AI application types fully

deployed

7> None 1–3 4–7 >7

"Pathseekers" "Transformers"
Low deployed/high High deployed/high
achieving: 26% achieving: 28%
N=753 N=794
7–4

Outcomes
achieved
(high degree)

"Starters" "Underachievers"
3–1

Low deployed/low High deployed/low

achieving: 29% achieving: 17%
N=832 N=496
enoN

6.8
6.2
5.9
5.5

4.0
3.5

1.9
1.6 1.4
1.0

Average number of AI apps in full deployment Average number of outcomes achieved to a high
degree
Source: The State of AI 4th Edition data analysis.
Deloitte Insights | deloitte.com/insights

4
 Deloitte’s State of AI in the Enterprise, 4th Edition

This analysis revealed four key profiles: strong outcomes became evident. They fell into the
following categories: Strategy, Operations,
• Transformers (High outcome and high
Culture and change management, and
deployed—28% of survey respondents):
Ecosystems.
Transforming but not fully transformed, this
group has identified and largely adopted
leading practices associated with the strongest Analysis of survey data and executive interviews
AI outcomes. They average 5.9 out of 10 revealed that success is built upon the foundation
possible full-scale deployments of different of a clear strategy that is communicated and
types of AI applications, and 6.8 out of 17 incentivized from the highest leadership—but that
possible outcomes achieved to a high degree. is not enough. With that clear strategy in place, two
They are the market leaders on their way to inter-related leading practices typically work
becoming AI-fueled organizations. together to support AI adoption and scale across
• Pathseekers (High outcome and low the enterprise: operations and culture plus change
deployed—26% of survey respondents): management. And finally, the support of a robust
Pathseekers have adopted capabilities and set of ecosystem partners was shown to provide the
behaviors that are leading to success, but on technical foundations and outside perspectives
fewer initiatives. They are making moves but needed to deliver and perpetually innovate at scale.
have not scaled to the same degree as
Transformers. They average 1.9 out of 10 Our analysis also revealed not just what those
possible full-scale deployments of different leading practices were, but how much of an effect
types of AI applications, and 6.2 out of 17
they had on organizational achievement:
possible outcomes achieved to a high degree.
• Strategy leading practice: AI-fueled
• Underachievers (Low outcome and high
organizations view AI as a key element of
deployed—17% of survey respondents): A
business differentiation and success, and
significant amount of development and
they set an enterprisewide strategy that
deployment activity characterizes this group;
is championed from the top. Organizations
however, they haven’t adopted enough leading
with an enterprisewide strategy and
practices to help them effectively achieve
leadership who communicate a bold vision are
meaningful outcomes. They average 5.5 out of
1.7 times more likely to achieve outcomes to a
10 possible full-scale deployments of different
high degree.
types of AI applications, and 1.4 out of 17
possible outcomes achieved to a high degree. • Operations leading practice: AI-fueled
organizations establish new operating
• Starters (Low outcome and low
models and processes that drive
deployed—29% of survey respondents):
sustained quality, innovation, and value
Getting a late start in building AI capabilities
creation. Organizations that document and
seems to characterize this group. They are the
enforce MLOps processes are approximately
least likely to demonstrate leading practice
two times as likely to achieve their goals to a
behaviors. They average 1.6 out of 10 possible
high degree. They are also about two times as
full-scale deployments of different types of AI
likely to report being extremely prepared for
applications, and 1.0 out of 17 possible
risks associated with AI and nearly two times
outcomes achieved to a high degree.
as confident that they can deploy AI initiatives
in a trustworthy way.
By analyzing these groups—the Transformers in
particular—the behaviors most associated with

5
Becoming an AI-fueled organization

• Culture and change management leading

practice: AI-fueled organizations
“By embracing AI strategically
nurture a trusting, agile, data-fluent and challenging orthodoxies,
culture and invest in change organizations can define a
management to support new ways of
working. Organizations that invest in change
road map for adoption, quality
management to a high degree are 1.6 times as delivery, and scale to create
likely to report that AI initiatives exceed or unlock value faster than
expectations and more than 1.5 times as likely
to achieve their desired goals, compared to ever before.”
the rest. — Irfan Saif, Deloitte AI coleader, principal,
Deloitte & Touche LLP
• Ecosystems leading practice: AI-fueled
organizations orchestrate dynamic
ecosystems that help build and protect
competitive differentiation. Overall, In the following report, we explore each leading
organizations with more diverse ecosystems practice in detail, sharing critical and often
were 1.4 times as likely to use AI in a way that overlooked actions that leaders can take to avoid
differentiates them from their competitors. pitfalls on their transformation journey.

6
 Deloitte’s State of AI in the Enterprise, 4th Edition

Strategy: What should

your north star be?

CORE LEADING PRACTICE: SET A CLEAR ENTERPRISEWIDE STRATEGY AT THE TOP THAT
ENABLES LEADERS TO HARNESS AI CAPABILITIES TO DRIVE NEW OPPORTUNITIES AND
COMPETITIVE ADVANTAGE.

Key findings:
• Set and communicate a bold vision. Organizations with an enterprisewide strategy and leaders who
communicate a bold vision are 1.7 times as likely to achieve outcomes to a high degree.

• Look for ways AI can help achieve a differentiated strategy. Only 38% of respondents believe their
use of AI differentiates them from competitors.

• Communicate your strategy transparently. Tell your workforce and the market about your strategy
and the implications and trade-offs along the way.

Pitfalls to avoid:
• Don’t ask data scientists or IT to drive your AI strategy. Senior business leaders should drive it based
on the core business strategy in partnership with data scientists.

• Don’t overindex on efficiency goals. Balance efficiency targets with growth- and
innovation-oriented goals.

O
NE OF THE most frequently cited leading Lost in AI use cases: Leaders
practices for AI transformation is the need can forget to put their
for a bold, enterprisewide strategy that is
business strategy first
set and championed by an organization’s highest
leadership. Our research confirmed this: To many leaders, it comes as a surprise to learn
Transformers are more than three times as likely to that the investment needed to develop AI solutions
have an enterprisewide strategy in place, and well cannot realize a return through the deployment of
over twice as likely as Starters to report their single, disconnected use cases, or even a handful.6
leaders communicate a vision for AI. However, This is why it’s so important to have an AI strategy
only 40% of our total survey respondents that is connected and coordinated across the
completely agreed that their company has one in enterprise, in tight alignment with the overarching
place. Meanwhile, even though a significant business strategy. All too often, however, business
majority (66%) of respondents view AI as critical to leaders get the planning process out of order,
success, only 38% believe their use of AI focusing too much on use cases or abdicating
differentiates them from competitors. What should leadership of the AI strategy to IT or data sciences.
organizations do differently to strengthen This can be a slippery slope, diminishing the
their approach? organization’s ability to use AI to create new ways

7
Becoming an AI-fueled organization

FIGURE 2

Leading AI strategy practices

Percentage of respondents who selected "completely agree" or "very important" to
these statements about strategy
Transformers Pathseekers Underachievers Starters Total

79%
68% 69% 66%
60%
55% 57%
48% 49%
44% 45%
38% 40% 40%
33% 33%
30%
22% 24%
19%

My company's use of My company has an My senior leaders Our AI initiatives are

AI differentiates us enterprisewide AI communicate a vision for important to our
from our competitors strategy AI that will significantly remaining competitive
change how we operate over the next five years

Source: State of AI 4th edition data analysis.

Deloitte Insights | deloitte.com/insights

of competing for customers, launching products, same key performance indicators (KPIs) that have
accelerating time-to-market, securing supply been crafted to incentivize and grow
chains, and beyond. competitive advantage.

In a now famous example from the early

To many leaders, it comes as a surprise 2010s, Jeff Bezos mandated that every
to learn that the investment needed leader across Amazon plan for how they
to develop AI solutions cannot realize would use AI and machine learning (ML) to

a return through the deployment of help the company compete and win. This
imperative drove unparalleled innovation
single, disconnected use cases, or even and was cited as the catalyst for the
a handful. Amazon’s rise to become an AI leader
today.7 Many of the strongest AI strategies
start in this same way: by pushing clear
The strongest AI strategies tend to begin without objectives down to business leadership, so they can
ever mentioning AI. Instead, they should begin identify gaps and opportunities within their
with the organization’s north star: the core divisions and work backward from there to
business strategy. From there, the process requires apply AI as a solution.
tight collaboration with engaged leaders across all
business divisions and the focus of workers at all These local plans should then be brought back to
levels. Ultimately, AI strategy should function as the top, so that mutual goals and initiatives can be
the fuel to the business strategy, aligning to the aligned and unified with the core business strategy.

8
 Deloitte’s State of AI in the Enterprise, 4th Edition

This step is often critical: It’s only when AI has “You have to go both for impact and build the
been integrated and proliferated throughout the foundations in parallel, and that is the most
enterprise that it can deliver the combination of challenging part,” advises Najat Khan, PhD, chief
efficiency- and value-creating outcomes needed to data science officer and global head of strategy &
fuel ongoing returns. operations for Janssen Research & Development.
“You have to pick the right questions, and have
what I call a diversified portfolio of questions to
Balance your goals: drive impact, ensuring that you can demonstrate
Overindexing on efficiency can early value to build momentum for achieving
longer-term, sustainable impact.”
lead to missed opportunities
It’s through the combination of both efficiency- and AI-fueled organizations can create durable
value-creation targets that organizations typically competitive advantage when the CEO and C-suite
achieve the most success. “When digitally collectively harness data, advanced analytics, and
transforming a company, you want greater degrees AI to shape strategic possibilities for both the near
of efficiency,” remarks Rajeev Ronanki, SVP and and long term in support of their
chief digital officer at Anthem. “But there is a corporate strategy.
second order of business: What new business
opportunities, what capabilities does AI open up
that allow for servicing adjacent or maybe entirely Communicate the
new areas?” vision: Publicly signaling
transformation can
Our survey results reinforced this, demonstrating
that lower-achieving organizations (Starters and
build market value
Underachievers) tended to focus more on efficiency Chief executives of high-achieving organizations
or “cost out” goals, while high-achieving typically serve as the AI communicator-in-chief.
organizations (Transformers and Pathseekers) According to our survey data, those organizations
were more likely to emphasize growth-oriented that communicate a clear vision are 1.5 times as
goals, such as: improving customer satisfaction, likely to achieve desired outcomes compared to
creating new products and offers, and entering new those who do not. The most effective leaders tend
markets. In other words, high-achieving to use their platform not only to communicate and
organizations are more likely to maintain an eye champion their plans; they also clarify the
toward the art of the possible and a growth implications and trade-offs required along the way.
mindset, which allow them to take advantage of This is often essential for maintaining focus and
opportunities often missed by those who overindex ensuring that decisions made at all levels of the
on efficiency or supporting business as usual. organization remain aligned to the vision.

Leaders should also remember that value

“Envision what is possible in your can be created by influencing perceptions

business, whether it’s been done before of the market and investors.
Communicating the company’s vision
or never been done before.” publicly can amplify success, signaling to
— Michelle Lee, VP of Amazon Machine Learning capital markets and the competitive talent
Solutions Lab
market that an organization is investing in
a bold and exciting future.8 If it’s not

9
Becoming an AI-fueled organization

important enough to merit such a forceful signal technology developments. As the organization’s
toward change, it’s highly likely that the core business strategy and AI capabilities mature
gravitational pull toward the status quo could over time, leaders should continually sharpen their
dampen outcomes for even the strongest strategy. goals, moving beyond staying competitive to
increasingly using AI and ML as competitive
differentiators.
Remain dynamic: Perpetually
iterate your AI strategy For more AI strategy recommendations:

Finally, developing an enterprisewide AI strategy • An innovation strategy powered by tech

that’s set up to fuel a differentiating core business
strategy is not a one-and-done exercise. • The AI Dossier: Top uses for AI in every major
Organizations should develop dynamic ways of industry — now and in the future
assessing their strategy to ensure it remains
responsive to ever-changing market and • A new language for digital transformation

10
 Deloitte’s State of AI in the Enterprise, 4th Edition

Operations: How do you

bring transformation
into everyday work?

CORE LEADING PRACTICE DRIVE ONGOING QUALITY, INNOVATION, AND VALUE CREATION
THROUGH NEW OPERATING MODELS, ROLES, AND PROCESSES.
Key findings:
• Reimagine business workflows and roles. Organizations that have undergone significant changes to
workflows or added new roles are more than 1.5 times as likely to achieve outcomes to a high degree.

• No excuses: Document and follow MLOps. Organizations that document and follow MLOps processes
are twice as likely to achieve their goals to a high degree. They are also approximately two times as likely
to report being extremely prepared for risks associated with AI and nearly two times as confident that
they can deploy AI initiatives in a trustworthy way.

Pitfalls to avoid:
• Business leaders should allocate more time to solution design. Effectively redesigning processes and
how AI tools fit into workflows requires thoughtful attention.

• Don’t underestimate the unique maintenance needs of AI solutions. Establish and document robust
MLOps procedures to ensure ongoing quality and ethical delivery.

T
ECHNOLOGY CANNOT DELIVER documenting AI life cycle publication strategies,
transformative results unless organizations and updating workflows, roles, and team structures
reimagine how work gets done. Most leaders across the business.
today understand this intellectually; however,
survey results show a disconnect in putting it into To ensure quality AI solution development,
action: Across a variety of operational activities— enterprise adoption, and the most successful
both on the business side and within IT or data outcomes, organizations should rethink their
science teams—only about one-third of those operations from two key perspectives: across the
surveyed report that they have adopted leading business workflow, and within their IT and data
operational practices for AI. This includes adhering science team processes.
to a well-calibrated MLOps framework,

11
Becoming an AI-fueled organization

FIGURE 3

Leading AI operations practices

Percentage of respondents who selected "completely agree" to these statements
about operations
Transformers Pathseekers Underachievers Starters Total

56%
53%
49% 50%
42% 44%
41% 41%
38% 38%
32% 33% 31% 30%
24% 24% 25%
20%
13% 15%

My functional group My functional group My functional group has My functional group

follows a documented follows documented undergone significant changes has created new AI job
AI model life cycle MLOps procedures in how we create teams and roles/functions to
publication strategy when developing an manage workflows to take maximize AI
AI solution advantage of new technologies advancements
in the last five years

Source: The State of AI 4th Edition data analysis.

Deloitte Insights | deloitte.com/insights
When this happens, Michelle Lee, VP of Amazon
A call for leaders: Machine Learning Solutions Lab, observes, “They
Business stakeholders then experience organizational inertia, because
either the use case being addressed wasn’t
should take ownership
important enough, or there is an unwillingness to
of AI-fueled solutions adopt a new and an unproven method.”
A successful AI solution should be conceived and
designed to fit within a new workflow created to Dr. Tian He, vice president and the head of JD
improve value delivery. To do this effectively, Logistics AI and Data Science, underscores that
business stakeholders should take a lead role, but “Most people learning especially machine learning
unfortunately, many misunderstand how to do this and deep learning just came out of school, and they
effectively. know the AI skills ... They’re technicians. But you
need to understand the business.”
This causes them to allocate too little
time to rethink the broader operational “We’ve seen a lot of AI projects where
shifts needed to support successful
people have implemented amazing
adoption of a value-creating solution.
All too often, AI and ML development
models, but they’ve never seen the light
teams are put in charge without a clear of day because the business rejects the
view into the business processes they process changes that go along with it.”
are tasked with transforming. — Rajen Sheth, VP of Google Cloud AI and Industry
Solutions

12
 Deloitte’s State of AI in the Enterprise, 4th Edition

Only with an engaged partnership between the and operational managers is important to align the
business and AI and ML development teams, can a necessary processes for AI and ML to take hold.
new way of working emerge. Even when business
leaders understand their role, a lack of AI fluency While developing these processes is generally the
can inhibit their ability to collaborate effectively responsibility of IT and data science leadership, all
with the AI and ML development teams. Some stakeholders and senior leaders should be
organizations have found success in creating new concerned that these processes and standards are
roles to help translate between business in place and observed across the organization.
stakeholders and model development teams. In They are key to ensuring the ongoing quality of
these circumstances, an individual well-versed in models that are fueling critical business processes.
both business and analytics can serve as the bridge Data from our survey bear out just how important:
between overarching business strategic goals and Organizations that strongly agreed that MLOps
AI technical requirements.9 Our survey processes were followed were twice as likely to
demonstrates that efforts in creating new roles like achieve their goals, compared to the rest.
this can pay off. High-achieving surveyed Furthermore, these surveyed organizations were
organizations (Transformers and Pathseekers) also approximately two times more likely to report
were significantly more likely to create new roles feeling extremely prepared for risks associated
and functions to maximize AI advancements. with AI, and nearly twice as confident that they can
deploy AI initiatives in an ethical, trustworthy way.

MLOps: New capabilities

require new processes Rethinking ops: A catalyst
for AI transformation
In the early days of enterprise AI, initiatives took
place within localized teams and were contained Establishing the appropriate structures, roles, and
within business divisions. Models were frequently working relationships across an organization can
built on data scientists’ desktops and required be one of the most important steps in bringing an
relatively simple and smooth processes to AI transformation to life: “If I were to give one
maintain.10 Today, models are being deployed in piece of advice to a C-suite–level person looking at
the cloud and running mission-critical workloads. how to get this right in their organization, I would
As organizations reach this scale, the level and say, ‘Look at the organizational structure, because
complexity involved in perpetually developing, that can really facilitate the change,’” advises Phil
training, testing, deploying, monitoring, and Thomas, executive vice president of Customer
maintaining models have caught many Insights Data & Analytics at Scotiabank. “That for
organizations by surprise: Only 33% of all survey us was a massive accelerant in our journey—getting
respondents completely agree they have MLOps the org structure right and creating a culture of
processes in place. being a data-driven organization that’s accepting of
the use of AI.”
Not all data scientists are skilled in taking on an
engineering or operational mindset to manage this For more AI operations recommendations:
at scale. This is why a strong collaboration across
• ML Oops to MLOps
data scientists, engineering, application developers,
• Taking AI to the Next Level

13
Becoming an AI-fueled organization

Culture and change

management: Why is
valuable change so elusive?
CORE LEADING PRACTICE BUILD A TRUSTING, AGILE, DATA-FLUENT CULTURE AND INVEST IN
CHANGE MANAGEMENT TO SUPPORT NEW WAYS OF WORKING.

Key findings:
• Data fluency pays off. High-achieving organizations (Transformers and Pathseekers) are nearly three
times as likely to trust AI more than their own intuition, compared to low-achieving organizations
(Starters and Underachievers).

• Prioritize change management. Organizations that invest in change management to a high degree are
1.6 times more likely to report that AI initiatives exceed expectations and more than 1.5 times as likely to
achieve their desired goals.

• Fear can be an indicator of positive change if paired with supportive culture and
change management.

Pitfalls to avoid:
• Don’t take a one-size-fits-all approach to change management. Tailor your efforts to key audiences
and ensure a variety of resources are available to support new behaviors.

• Don’t expect change management to fix a poorly designed transformation. Thoughtfully designing
a new solution from the beginning can set the foundation for positive change.

O
VER THE PAST few decades, the pace of Through interviews and survey data analysis, we
business and technology change has found the organizations with the strongest AI
quickened, requiring workers to adapt, outcomes tend to display some common
perpetually learn new skills, and make decisions characteristics, including high levels of
amid growing ambiguity. For many organizations, organizational trust, data fluency, and agility. And
these shifts have challenged a critical facet within to get there, investment in change management
their organization: their culture. has been key to successful AI transformation:

Executive interviewees repeatedly emphasized how

the cultural characteristics of their organizations “Not to say that technical model
either facilitate or hinder their AI-transformation building is easy, but the biggest
efforts. This aligned with a 2019 Deloitte Survey
challenge is culture change.”
that found that organizations with the most data-
— Phil Thomas, executive vice president
driven cultures were twice as likely to significantly of Customer Insights Data & Analytics at
exceed business goals.11 Scotiabank

14
 Deloitte’s State of AI in the Enterprise, 4th Edition

Organizations that invest in change management machines replacing humans. But high-achievers
are 1.6 times as likely to report that AI initiatives also reported little desire to reduce employee
exceed expectations and more than 1.5 times as headcount as well as high investment in training
likely to achieve outcomes than those that don’t. A and change management. When viewed through
recent study also made clear that by providing this lens, fear may be a positive indicator that an
workers with clear direction and support, change organization’s AI vision is bold. This can bear fruit
management can boost both trust and when paired with other supportive actions and
engagement. 12
cultural characteristics to drive success. A culture
that trusts, even if they fear, demonstrates agility.
Change management can help build that trust.
Ingredients of an AI-
ready culture: Trust, Executive interviews confirmed this interpretation,
calling out a variety of behaviors, such as
data fluency, agility
collaboration, relationship-building, and training,
Trust: Surprisingly, surveyed high-achieving which may collectively point to higher levels of
organizations (Transformers and Pathseekers) trust within the organization. Trust is based on
report more than twice the amount of fear competence and intent:13 If employees believe in
compared to low-achieving organizations the organization’s ability to build capable AI
(Underachievers and Starters). Typically, when we systems and its intent to use technology for their
consider AI-related fear, the focus is on job loss or benefit—not detriment—then trust can grow.

FIGURE 4

Leading culture and change management practices

Percentage of respondents who selected "completely agree" to these statements about
culture and change management
Transformers Pathseekers Underachievers Starters Total

54%

40% 40%
35% 37%
31%
27% 28% 26%
23%
19%
15%
12% 10%
8%

AI initiatives have created Employees trust AI-derived My functional group invests in

fear or concern among our insights more than their own change management,
workforce intuition incentives, or training activities
to help people integrate new
technology into their work

Source: The State of AI 4th Edition data analysis.

Deloitte Insights | deloitte.com/insights

15
Becoming an AI-fueled organization

“It’s really about working together, building emphasizes this point: “We have a widespread
collaborative, trusted partnerships,” advises Eileen culture of experimental validation. We don’t accept
Vidrine, chief data officer at the US Department of an answer of, ‘The model said so.’ No. Model
the Air Force. In organizations where that may be outcomes are continuously scrutinized through live
lacking, it’s imperative to support trust- and testing and validation.” In other words, data-
relationship-building to break down silos. focused organizations tend to require a more
profound understanding of data. Workers should
Data fluency: “In order for there to be AI success, be incentivized to explain and justify model
people will have to change their relationship with decisions; this serves to drive more creative
data,” says Andrew Beers, chief technology officer insights as well as faster detection of model errors
at Tableau. Part of this, of course, involves building if and when they arise.
advanced technical data capabilities; however,
that’s often a smaller piece of the puzzle than Upskilling is important in this effort. “Talent is
leaders realize. More foundational tends to be really one of the big challenges that we see,”
raising the base level of data literacy across all observes Ong Chen Hui, Biztech Group cluster
levels of the organization. This means encouraging director at Infocomm and Media Development
everyone to build the critical thinking skills needed Authority of Singapore. She continues, “it’s not
to ask the right questions, and then find the right strictly the AI scientists. We also see that, in
data to solve problems in their everyday work. adjacent competencies needed to support AI, there
is a talent shortage there also.”

“It’s our responsibility as a society, and Most organizations understand

the importance of including
our responsibility as business leaders to training or reskilling to support an
create new talent with AI skills—not only AI transformation—in fact, nearly
for the engineers and data scientists, but three quarters of all surveyed

also for every single role in an organization, organizations did not report a
strong preference for hiring
no matter how ‘technical.’ That’s how we externally over reskilling their
raise the level of AI proficiency for all of our current workforce.

people, and frankly for society at large. And

Agility: AI-fueled organizations
that’s when modernization will happen.” typically do more than trust data;
— Dan Helfrich, chairman and CEO, Deloitte Consulting LLP they demonstrate a willingness to
quickly turn insights into action
Developing data-literacy skills builds confidence and rapid experimentation. Anthem’s Ronanki
and a deeper trust in models and AI, which in turn agrees, commenting on the degree of change this
can help set organizations up for positive outcomes. can require for organizations that have grown
High-achieving organizations from our survey prioritizing safer and more secure investments: “A
(Transformers and Pathseekers) were lot of [the challenge] is getting comfortable with
approximately three times more likely to trust AI the fail-fast, pivot mindset when you take on and
more than their own intuition, compared to low- do new things,” he notes. “With AI investments
achieving organizations (Starters and and digital transformation in general, you need
Underachievers). Naturally, trusting AI doesn’t experimentation and learning from failures. It’s a
mean blindly following model outputs. Tulia big change.”
Plumettaz, director of data science at Wayfair

16
 Deloitte’s State of AI in the Enterprise, 4th Edition

Building an AI-ready addressing these elements for all workers, the

culture: The need for more likely desired change is to take hold. A strong
program typically spends time identifying a
change management
complete set of behaviors it wants to promote,
AI in particular is significantly altering the way which are then used to create a multilayered
work gets done, requiring a redefinition of work, 14
program of communication, training, support
and subsequently which skills and capabilities the resources, incentives, and “nudges”16 that
human workforce needs to deliver value.15 “Data ultimately drive the creation of new norms. In
science touches every single therapeutic area, other words, you can’t change people’s beliefs by
business unit, and the different functions,” says just telling them to think differently—but you can
Janssen’s Khan. “Therefore the change journey that share information, educate, incentivize, and
comes with it is significant. It can be support them to behave in different ways. This,
uncomfortable at first, certainly not optional, and over time, can change their beliefs.
completely worthwhile to have collective
transformational impact on the patients we serve.” Most organizations underinvest in these activities:
Only 37% of survey respondents reported
This level of change necessitates support to help significant investment in change management,
the workforce adjust, which our survey data incentives, or training activities to help their
reinforced: Transformers invest in change people integrate new technology into their work,
management at nearly twice the rate of Starters often resulting in a slower, less successful
and Underachievers. Not only does AI support transformation.
workers in adopting key skills and behaviors, it can
also be an important trust builder by Even when designed well, organizations should
demonstrating that the company values them. keep in mind that the most successful
transformations are typically based on workers’
Many organizations go astray even before they consent and buy-in, and this takes time. Leaders
initiate a change management initiative. If should seek ongoing measurement of KPIs, using
business sponsors haven’t taken the time to work them to track progress and iteratively hone the
closely with AI solution developers to carefully plan change program. Adding support where behaviors
how new technology will improve workers’ aren’t taking hold and celebrating achievements
experience and ability to deliver value, a change along the way is often key to ultimately arriving at
management program will not fix the inefficiencies a culture that can drive AI-fueled success.
or undesirable behaviors that poorly conceived
processes unintentionally introduce. For more change management recommendations:

• Analytics and AI-driven enterprises thrive in

Once confident in the transformation design itself,
the Age of With: The culture catalyst
successful change management comes down to
whether the transformation goals and • Humanizing Change: Developing more
requirements have been made clear, relevant, and effective change management strategies
achievable for the many different audiences within • Nudging for Good
an organization. The better a change program is at

17
Becoming an AI-fueled organization

Ecosystems: How should you

orchestrate your partnerships?
CORE LEADING PRACTICE ORCHESTRATE DYNAMIC ECOSYSTEMS THAT HELP BUILD AND PROTECT
COMPETITIVE DIFFERENTIATION.

Key findings:
• Build broad and diverse partnerships. Eighty-three percent of the highest-achieving organizations
surveyed (Transformers and Pathseekers) create a diverse ecosystem of partnerships to execute their
AI strategy.

• Use partners to improve your perspective on the market. Organizations with diverse ecosystems are
significantly more likely to have a transformative vision for AI, enterprisewide AI strategies, and use AI as
a strategic differentiator.

Pitfalls to avoid:
• Beware of vendor lock. Too few external partnerships can make it difficult to part ways with vendors if
needed in the future.

• Don’t sacrifice your competitive position. Build differentiating capabilities in-house to protect your
competitive advantage.

N
O COMPANY HAS all the needed talent, When an ecosystem strategy is robust and well-
algorithms, data sets, or breadth of orchestrated, it offers an organization the flexibility,
perspective in-house to innovate stability of resources, and informed perspectives
perpetually with AI. That’s largely why most of needed to navigate and compete in an everchanging
today’s AI-fueled organizations establish robust market. Survey data reinforced this point: Eighty-
technology ecosystems: Through a diverse set of three percent of high-achieving organizations
means, they build, partner, license, and access the (Transformers and Pathseekers) use at least two or
elements needed to execute their strategy over the more types of ecosystem partners—a significantly
long term. higher percentage than low-achieving organizations
(Starters and Underachievers). Organizations with
“Ecosystems are dynamic and more diverse ecosystem relationships also reported

coevolving communities of diverse stronger preparedness to address AI risk and more

confidence in delivering AI ethically.
actors who create and capture
new value through increasingly Most business leaders today understand the
sophisticated models of both importance of building strong ecosystems.
However, a common misconception can weaken
collaboration and competition.” business leaders’ approach and diminish long-term
— Eamonn Kelly, managing director of
value: Many mistakenly believe that simple and
Businesses, Global, and Strategic Services,
Deloitte LLP16 streamlined ecosystem strategies are more efficient
and thus stronger.

18
 Deloitte’s State of AI in the Enterprise, 4th Edition

FIGURE 5
trying to optimize interoperability and avoiding
Leading ecosystem practice what I would call ‘vendor lock’ as much as possible,”
Percentage of respondents with multiple advises the US Department of the Air Force’s Vidrine.

ecosystem relationships
A healthier ecosystem approach typically identifies
Transformers Pathseekers Underachievers a base platform and looks for a variety of
Starters Total
opportunities to integrate different vendors,
including those that may be emerging or niche.
83% 83% When this approach is executed well, it not only
70% 74%
protects from overdependence, but can also result
59% in a higher level of differentiation, flexibility, and
access to expanded perspectives on the market.
Survey data reinforced this, showing that
organizations with more diverse ecosystems were
much more likely to have transformative visions
Two or more ecosystem relationships for AI and use AI as a strategic differentiator.

Source: The State of AI 4th Edition data analysis. Of course, thoughtful orchestration of these
Deloitte Insights | deloitte.com/insights relationships is required to achieve such outcomes.
Time to diversify: Simple Deciding how to leverage each one is just as
and streamlined tech important as who and how many. “Emergent
technology domains, like artificial intelligence, are
ecosystems introduce risk
evolving too fast and in too many directions for
While it may seem counterintuitive, simple and most companies to keep pace. This demands a
streamlined ecosystems frequently introduce more more fluid and hybrid approach—buy, lease, invent,
risk than those that are diverse and complex. experiment, and partner with a variety of external
Understandably, many leaders attempt to organizations that compliment and extend your
minimize the time and expense of managing company’s competitive edge,” advises Keith Strier,
multiple relationships, and so they choose to work vice president of worldwide AI initiatives at Nvidia.
with the fewest partners needed to meet their goals.
As AI and ML tools become more deeply embedded Which elements your organization decides to build
into an organization’s core operations, it’s in-house versus buy or license externally is an
important to expand this view. Concentrating important strategic exercise that can’t be
resources on too few vendors or partners can lead overvalued. Each decision should come down to
to overdependence,17 so much so that an one central question: How does each relationship
organization could become virtually unable to support and protect your organization’s ongoing
move to a new vendor in the future. In other words, competitive differentiation in the market?
when too many critical business processes rely on a
single vendor’s platform and models, it can become For more ecosystem recommendations:
extremely difficult to disentangle from them
• Business ecosystems come of age
without significant disruption. This can ultimately
stymie innovation and growth. • Ecosystem-driven portfolio strategy

• In a world of ecosystems, what’s

“It needs to be part of the conversation at the your strategy?
beginning and through the whole life cycle about

19
Becoming an AI-fueled organization

Our AI-fueled future:

The pathway is clear

O
RGANIZATIONAL PROGRESS HAS always 2. Update your operations and document
depended on humans’ ability to imagine a new ways of working. Ask: Are your
new vision for the future and identify the business leaders (not just data sciences teams)
opportunity available within it. We seem to be driving the AI agenda? And are they allocating
rapidly approaching the day when AI could adequate time to design applications in a way
independently and reliably illuminate creative and that the business can adopt and deliver
strategic opportunity, releasing us from the transformational results?
confines of our limited perspectives. As we advance – Ensure business sponsors are taking the
further into that AI-fueled future, those lead in AI-fueled initiatives
organizations that lay the foundations now will – Redesign roles, processes, and workflows to
likely be rewarded manyfold. integrate and encourage adoption of
AI tools
Rapidly transforming, but not yet fully – Set enterprisewide standards for AI and ML
transformed, has been the underlying theme of this solution development to promote high-
report—very few organizations in the market have quality, ethical application design
achieved mature AI-fueled transformations across and delivery
the enterprise. Even among Transformers, who are
leading the way, there is room to improve: Forty- 3. Evaluate your culture in the Age of With.
five percent don’t have a significantly differentiated Ask: Does your organization display an
AI approach, 50% report partially mature MLOps AI-ready culture?
procedures, and 46% aren’t investing significantly – Build organizationwide data literacy, trust,
in change management efforts to drive deeper and agility
enterprise transformation. Most of today’s leaders – Ensure your workforce has adequate
have plenty of work to do to ensure strong support to learn new skills, capabilities,
foundations are set: and ways of working through multilayered
change management approaches targeted
1. Start with your enterprisewide strategy. to unique audiences
Ask: Have your organization’s leaders set a core,
overarching business strategy and incentivized 4. Diversify your ecosystems. Ask: Is your
AI development enterprisewide to achieve the vision? ecosystem strategy diverse, well-integrated, and
– Set bold, enterprisewide goals for AI that supporting your ability to differentiate and
tightly align to the core business strategy remain agile?
– Remember the art of the possible; don’t let – Design your ecosystem to always protect
“cost out” goals cause you to miss and drive your competitive differentiation
big opportunities
– Champion the plan and transparently The pathways have been made clear. For those
communicate about required trade-offs leaders with the ambition, these leading practices
across all levels of the organization and can be a road map to future—currently unimaginable
even externally —AI successes as the Age of With matures.

20
 Deloitte’s State of AI in the Enterprise, 4th Edition

Methodology

T
O OBTAIN A global view of how AI is outcomes achieved through AI initiatives. For full-
transforming organizations, Deloitte scale AI deployments, we calculated the cumulative
surveyed 2,875 IT and line-of-business frequency of respondents who selected “deployed”
executives between March and May 2021. Eleven (achieved at least one full-scale deployment)
countries were represented: Australia (109 among the 0–10 types of AI applications. Similarly,
respondents), Brazil (218 respondents), Canada we calculated cumulative frequency by counting
(216 respondents), China (219 respondents), the number of outcomes achieved to a “high degree”
France (188 respondents), Germany (218 among the 0–17 potential outcomes achieved by
respondents), India (225 respondents), Italy (97 respondents. This established the following profile
respondents), Japan (110 respondents), the United groups of respondents:
Kingdom (218 respondents), and the United States
• Transformers: (28%, N=794) have achieved
(1,057 respondents).
four or more high full-scale AI deployments
and at least four outcomes to a high degree
All participating companies have adopted AI their AI initiatives. They are considered the
technologies and are AI users. Respondents were leader group, the most “AI-fueled,” within our
required to meet one of the following criteria: survey respondents.
responsible for AI technology spending or approval
• Pathseekers: (26%, N=753) have achieved
of AI investments, developing AI technology
fewer than four high full-scale AI deployments
strategies, managing or overseeing AI technology
but still achieved at least four outcomes to a
implementation, serving as an AI technology
high degree through their AI initiatives.
subject matter specialist, or making or influencing
decisions around AI technology. • Underachievers: (17%, N=496) have
achieved more than three high full-scale AI
deployments but still achieved fewer than three
To complement the blind survey, Deloitte
outcomes to a high degree through their
conducted qualitative telephone interviews with 17
AI initiatives.
AI experts from various industries.
• Starters: (29%, N=832) are still developing or
Analysis model. To identify “AI-fueled” leader exploring AI deployments and have deployed
groups, we developed an analysis model defining fewer than four high full-scale AI deployments.
four profiles of organizations based on the They have achieved fewer than four outcomes
frequency of full-scale AI deployments and the to a high degree through their AI initiatives.

21
Becoming an AI-fueled organization

FIGURE 6

Global respondent profile

2,875 IT and business executives
Industry totals

Technology, media, and telecom 24%

Energy, resources, and industrials 20%

Consumer 20%

Financial services 19%

Life sciences and health care 10%

Government and public services 7%

Company size Annual revenue

18% 14%
21%
26% 100–999 $10B+
$50–500M
10,000+

14%
$5–10B

20%
18% $500M–$1B
38%
5,000–9,999 1,000–4,999
30%
$1–5B

Title Role
7% Other
4% Owner/partner
31% 6% Board of directors
C-suite
11%
Head of business unit
51% 49%
LOB IT

13%
Head of department

Director/VP 28%

Source: The State of AI 4th Edition data analysis.

Deloitte Insights | deloitte.com/insights

22
 Deloitte’s State of AI in the Enterprise, 4th Edition

Endnotes

1. Unless otherwise cited, data in this report comes from the State of AI 4th Edition survey fielded between March
and May 2021.

2. Mike Bechtel, Seeing the forest for the trees, and the forests beyond: The future of AI, Deloitte, accessed
September 27, 2021.

3. Frank Konkel, “Survey: Pandemic accelerating AI adoption,” Nextgov, March 26, 2021.

4. Beena Ammanath, Susanne Hupfer, and David Jarvis, Thriving in the era of pervasive AI: Deloitte’s State of AI in the
Enterprise, 3rd Edition, Deloitte Insights, 2020.

5. Deloitte definition of an AI-fueled organization.

6. Deloitte analysis (Dave Kuder interview).

7. Steven Levy, “Inside Amazon’s artificial intelligence flywheel,” Wired, February 1, 2018.

8. Deloitte analysis.

9. Deloitte, “An innovation strategy powered by tech,” accessed September 27, 2021.

10. Rohit Tandon and Sanghamitra Pati, ML-Oops to MLOps, Deloitte, accessed September 27, 2021.

11. Tom Davenport et al., Analytics and AI-driven enterprises thrive in the Age of With: The culture catalyst, Deloitte
Insights, July 25, 2019.

12. Paul J. Zak, “The neuroscience of trust,” Harvard Business Review, January 1, 2017.

13. Deloitte, The future of trust: A new measure for enterprise performance, 2021.

14. John Hagel, John Seely Brown, and Maggie Wooll, Redefine work: The untapped opportunity for expanding value,
Deloitte Insights, 2018.

15. John Hagel, John Seely Brown, and Maggie Wooll, Skills change, but capabilities endure: Why fostering human
capabilities first might be more important than reskilling in the future of work, Deloitte Insights, August 30, 2019.

16. Shrupti Shah et al., Nudging for good: Using behavioral science to improve government outcomes, Deloitte Insights,
June 24, 2019.

17. Eamonn Kelly, Business ecosystems come of age—Part of the Business Trend Series, Deloitte University Press,
April 16, 2015.

18. Justice Opara-Martins, Reza Sahandi, and Feng Tian, “Critical analysis of vendor lock-in and its impact on cloud
computing migration: a business perspective,” Journal of Cloud Computing 5, no. 4 (2016).

23
Becoming an AI-fueled organization

Acknowledgments
A research effort of this size requires the collaboration of many, without whom this project would not
be possible. The authors would like to thank our external interviewees: Keith Strier, NVIDIA
Corporation; Andrew Beers, Tableau; Dr. Tian He, JD.com; Eileen Vidrine, Department of the Air
Force; Michelle K. Lee, Amazon Web Services; Najat Khan, Janssen Pharmaceuticals; Ong Chen Hui,
Singapore Infocomm Media Development Authority; Phil Thomas, Scotiabank; Rajeev Ronanki,
Anthem; Rajen Sheth, Google; Tulia Plummettaz, Wayfair. We would also like to thank Deloitte
leadership and subject matter experts: Ben Stiller, Rich Nanda, Jim Rowan, Manoj Suvarna, Tasha
Austin, Dave Kuder, Rohit Tandon, Sangha Pati, Michael Griffiths, and Mike Bechtel. And last but
certainly not the least, the core research project team: Abha Kulkarni, Brenna Sniderman, David
Novak, David Levin, Jonathan Holdowsky, Negina Rood, Andy Bayiates, Kate Schmidt, Mike Nash,
Scott Woodward, Jaya Gopalan, James Boise, and Bridget Maloney.

About the authors

Beena Ammanath | bammanath@deloitte.com

Beena Ammanath is a managing director with Deloitte Consulting LLP and is an award-winning senior
executive with extensive global experience in artificial intelligence and digital transformation. She is the
founder and CEO of Humans for AI Inc and has coauthored the book AI Transforming Business.

Nitin Mittal | nmittal@deloitte.com

Nitin Mittal is a principal with Deloitte Consulting LLP. He currently serves as the US Artificial
Intelligence (AI) Strategic Growth Offering Consulting leader. He specializes in advising clients to
achieve competitive advantage through data and cognitive-powered transformations that promote
amplified intelligence and enable our clients to make strategic choices and transform ahead
of disruption.

Irfan Saif | isaif@deloitte.com

Irfan Saif is a principal at Deloitte & Touche LLP and chief strategy officer for Deloitte Risk & Financial
Advisory. Throughout his professional services career, he has delivered and overseen the strategy and
implementation of transformational cyber risk and disruptive technology solutions for leading global
enterprises. He serves large, multinational clients in the technology sector as well as leading brands in
media, financial services, and retail.

Siri Anderson | sianderson@deloitte.com

Siri Anderson is a researcher and writer at the Deloitte Center for Integrated Research, where she also
serves as chief of staff. Previously, Anderson led the sponsored and contributed content program for
Deloitte Digital. Before joining Deloitte, Anderson worked in editorial and strategic communications
roles in technology and health care industries.

24
 Deloitte’s State of AI in the Enterprise, 4th Edition

Contact us
Our insights can help you take advantage of change. If you’re looking for fresh ideas to address your
challenges, we should talk.

Industry leadership

Nitin Mittal
Principal | Deloitte Consulting LLP
+1 617 947 7500 | nmittal@deloitte.com

Nitin Mittal is a principal with Deloitte Consulting LLP. He currently serves as the US Artificial
Intelligence (AI) Strategic Growth Offering Consulting leader.

Irfan Saif
Principal | Deloitte Risk & Financial Advisory
+1 415 269 8276 | isaif@deloitte.com

Irfan Saif is a principal at Deloitte & Touche LLP and chief strategy officer for Deloitte Risk & Financial
Advisory. In his more than 25-year career in professional services, Saif has delivered and overseen the
strategy and implementation of transformational cyber risk and disruptive technology solutions for
many leading global enterprises.

Beena Ammanath
Executive director of Deloitte AI Institute | Deloitte Consulting LLP
+1 925 474 7139 | bammanath@deloitte.com

Beena Ammanath is a managing director with Deloitte Consulting LLP and is an award-winning senior
executive with extensive global experience in artificial intelligence and digital transformation.

Research leadership

Brenna Sniderman
Executive director | The Deloitte Center for Integrated Research | Deloitte Services LP
+1 929 251 2690 | bsniderman@deloitte.com

Brenna Sniderman is the executive director of the Center for Integrated Research, Deloitte Services LP,
where she oversees a cross-industry research team focused on global shifts in technology, mobility,
climate, and workforce—developing work that ultimately helps guide how organizations can operate
and strategize in an age of rapid technology and cultural change. Sniderman’s work is cited in
prominent media platforms such as Forbes, GE Digital, and Wired, and frequently gives keynote
speeches at industry and client conferences.

25
Sign up for Deloitte Insights updates at www.deloitte.com/insights.

Follow @DeloitteInsight

Deloitte Insights contributors

Editorial: Andy Bayiates, Denise Weiss, Emma Downey, Ribhu Ranjan, and Rupesh Bhat
Creative: Kevin Weier and Jagan Mohan
Audience development: Nikita Garia, Alexandra Kawecki, and Hannah Rapp
Cover artwork: Kevin Weier

About Deloitte Insights

Deloitte Insights publishes original articles, reports and periodicals that provide insights for businesses, the public sector and
NGOs. Our goal is to draw upon research and experience from throughout our professional services organization, and that of
coauthors in academia and business, to advance the conversation on a broad spectrum of topics of interest to executives and
government leaders.
Deloitte Insights is an imprint of Deloitte Development LLC.

About this publication

This publication contains general information only, and none of Deloitte Touche Tohmatsu Limited, its member firms, or its
and their affiliates are, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other
professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be
used as a basis for any decision or action that may affect your finances or your business. Before making any decision or taking
any action that may affect your finances or your business, you should consult a qualified professional adviser.
None of Deloitte Touche Tohmatsu Limited, its member firms, or its and their respective affiliates shall be responsible for any
loss whatsoever sustained by any person who relies on this publication.

About Deloitte
Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its
network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent
entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. In the United States, Deloitte refers to
one or more of the US member firms of DTTL, their related entities that operate using the “Deloitte” name in the United States
and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public
accounting. Please see www.deloitte.com/about to learn more about our global network of member firms.

Copyright © 2021 Deloitte Development LLC. All rights reserved.

Member of Deloitte Touche Tohmatsu Limited