Cloud's Trillion-Dollar Prize Is Up For Grabs
Cloud's Trillion-Dollar Prize Is Up For Grabs
Cloud's Trillion-Dollar Prize Is Up For Grabs
This article was a collaborative effort by Will Forrest, Mark Gu, James Kaplan, Michael Liebow,
Raghav Sharma, Kate Smaje, and Steve Van Kuiken, representing views from McKinsey Digital.
February 2021
Moderna CEO Stéphane Bancel made the decision to build his mRNA research-and-development
platform on public cloud to create what he calls “software for life.”i He used the cloud as a means to
accelerate therapeutic discovery and development. When the COVID-19 pandemic hit, this strategy
proved prescient. The company was well positioned to quickly design research experiments and
to harness its automated laboratory and manufacturing processes and enhanced drug-discovery
pipeline.
Moderna runs its Drug Design Studio, a proprietary web application, on cloud and leverages cloud’s
scalable compute and storage infrastructure to analyze and quickly design mRNA sequences for
protein targets. Scientists and engineers also use fully managed cloud data-warehousing services
to integrate insights from multiple experiments running in parallel and quickly refine the design and
production cycle.ii Moreover, the adoption of cloud principles, such as infrastructure as code (IaC) and
security as code, helped to automate good-practice (GxP) compliance processes so the organization
can move quickly while staying secure and compliant.iii Thanks in part to cloud, Moderna was able
to deliver the first clinical batch of its vaccine candidate (mRNA-1273) to the US National Institute of
Health for phase one trials just 42 days after the initial sequencing of the virus,iv “because you don’t
have to reinvent everything, you just fly,” Bancel said.v
More companies are starting to see the real benefits of cloud,vi which has been long heralded as a
catalyst for innovation and digital transformation, thanks to its ability to increase development speed
and provide near-limitless scale. While Moderna’s success illustrates the business opportunities that
cloud makes possible, it only scratches the surface of the potential value at stake. A detailed review of
cloud cost-optimization levers and value-oriented business use cases foresees more than $1 trillion
in run-rate EBITDA across Fortune 500 companies as up for grabs in 2030 (see sidebar, “About the
research”), a number that will grow as cloud facilitates the adoption of emergent technologies such as
augmented reality and blockchain. This $1 trillion is less a prediction than an estimate of what should
be possible, provided companies aggressively pursue the cloud opportunity—and a call to action,
because early adopters will capture a disproportionate share of the total value.
The emergence of this immense value pool comes at a time of increasing competitive pressure on
companies. Fast-moving digital players are creating a fluid business landscape and accelerating the
pace of change. For CEOs, cloud adoption is not just an engine for revenue growth and efficiency.
Its speed, scale, innovation, and productivity benefits are essential to the pursuit of broader digital
business opportunities, now and well into the future. Yet an overly narrow view of cloud-value
economics and where value exists often keeps companies from achieving the desired outcomes.
The good news is that many companies across a range of industries have successfully implemented
public cloud to achieve impressive results. These companies follow three best practices. First, they
execute a well-defined, value-oriented strategy across IT and businesses and install a cloud-ready
operating model. Second, they develop firsthand experience with cloud and adopt a much more
technology-forward mindset than their peers. And finally, they excel at developing a cloud-literate
workforce.
Our research identifies the pools of value for cloud adoption across three dimensions—rejuvenate,
innovate, and pioneer—as well as the drivers of that value across the first two dimensions. It also
highlights likely avenues for growth in the pioneer dimension. CEOs can begin their journey by working
with their tech leadership to focus on four actions: set ambitious targets, pursue a hard-headed
economic case, adopt cloud-native ways of working, and invest in standardized, automated cloud
platforms.
1. Rejuvenate 2. Innovate
$430 billion $770 billion
IT cost optimization Innovation-driven growth
Cost optimization of application development and Business growth from new and enhanced
maintenance and IT infrastructure use cases in analytics, IoT, and automation
Source: Independent third-party research data (OmnicomGroup and Known), industry and McKinsey expert interviews, McKinsey D2020 IT cost benchmarking,
McKinsey Global Institute research, team analysis
Companies in every industry can capture substantial value from cloud, but it isn’t distributed evenly. High
tech, oil and gas, retail, healthcare systems and services, insurance, and banking are positioned to generate
the most value as measured by EBITDA impact in 2030, although almost all industries across the Fortune
500 show potential for an average rise in EBITDA of more than 20 percent (Exhibit 1).
# of EBITDA impact as %
companies EBITDA impact, $ billion of 2030 EBITDA
High tech 30 110–160 28–40
Oil & gas 45 80–160 29–60
Retail 64 90–140 31–53
Healthcare systems & services 30 70–140 35–74
Insurance 45 70–110 43–70
Banking 36 60–80 13–17
Automotive & assembly 23 40–60 31–54
Telecom 12 40–60 12–19
Advanced electronics & semiconductors 25 30–50 12–25
Consumer packaged goods 43 20-40 11–20
Transport & logistics 20 20-40 24–41
Pharmaceuticals & medical products 12 20-40 9–19
Media & entertainment 14 20–30 12–18
Travel 11 10–30 28–44
Aerospace & defense 12 10–20 14–22
Basic materials 23 ~10 11–20
Chemicals 15 ~10 10–18
Electric power & natural gas 14 ~0 5–7
Infrastructure 13 ~0 12–21
Source: Independent third-party research data (Omnicom Group and Known), industry and McKinsey expert interviews, McKinsey D2020 IT cost benchmarking,
McKinsey Global Institute research, team analysis
This value distribution is likely to change as the impact of cloud evolves. Democratized access to
computational power and infrastructure could reshape the landscape in industries that have historically
not been highly competitive. Like several previous technology disruptions, cloud shifts barriers to entry
in many markets from scale to skill, enabling smaller companies with the right skills to scale businesses
on the latest infrastructure without worrying about up-front costs—thus creating a threat to slower-
moving incumbents.
Use cases also differ by industry. Solutions that unlock the value of cloud include inventory optimization
in retail, automated forecasting in oil and gas, chatbot support for high tech, and customer-call-center
optimization in banking (Exhibit 2).
Healthcare
Retail systems & services High tech Oil & gas Banking
Inventory Analytics-driven Chatbots: Provide Parameter Digital advertising:
optimization: procurement: first-level support and optimization: Customized web
Large sets of data are Document analysis answer most FAQs via Data-driven analytics service based on
analyzed daily to converts scanned phone, email, and chat adjust well settings to digital advertising and
increase accuracy and documents into increase output customer browsing
maximize availability searchable text, and behavior
while minimizing risk machine-learning (ML)
of waste models identify clauses
of interest
Total 700–1,200
Rejuvenate
Rejuvenation describes a break from traditional legacy approaches by using cloud to lower costs and
risk across IT and core operations.
The economics of cloud computing is both controversial and complicated. On the one hand, cloud
provides access to automated capabilities that enterprises could never afford to develop on-premise,
and cloud service providers (CSPs) leverage inverse correlation of workload usage patterns to run
their assets at much higher utilization. On the other hand, CSPs charge based on consumption, and
companies must remediate existing applications for them to run efficiently in the cloud.
Research also indicates that effective cloud usage can improve application development and
maintenance productivity by 38 percent and infrastructure cost efficiency by 29 percent for migrated
applications. As a result, increasing the share of Fortune 500 applications in the cloud from 10 percent
to 60 percent would yield benefits of $56 billion in application development and maintenance and
$12 billion in infrastructure expenditures.
As it recovered after a cyberattack, Maersk, the largest container-shipping line and vessel operator
in the world, used the cloud to build out new capabilities at half the cost of doing it on-premise. The
company implemented a new IT operating model to enable user self-service and put the responsibility
for resource management in the hands of users. Recognizing the need for centralized governance,
Maersk created the tools and processes to allow for real-time chargebacks at a project level, visibility
into license management, and better analytics to understand consumption patterns and potential cost
savings.
In India, Easy Pay provides local neighborhood shops with a point-of-sale (POS) system that facilitates
their payments to a variety of suppliers. Today, the company has a presence in six cities, with a footprint
of 650-plus retail points servicing six million unique customers. It is also involved in government-led
smart-city projects in cooperation with major Indian banks. Easy Pay launched in 2016 with its own
data centers, but its customers felt that processing was too slow, and the system was plagued with
unforeseen downtime. Since migrating to cloud, Easy Pay provides close to 100 percent availability, and
transaction times have been reduced from 12 seconds to just five. Easy Pay expects at least a fivefold
increase in its existing merchant base, from 300,000 to 1.5 million, in fiscal year 2021 and considers
cloud a critical enabler of an exponential increase in traffic.viii
California Design, an online fashion-bedding brand, depended upon a myriad of systems to track its
complex forecasting and reordering processes. Team members typically planned inventory manually
using desktop spreadsheet software, which could lead to excess inventory. Accurately forecasting
demand and supply was essential to the company’s financial success, but it was also a challenge. The
organization migrated its database to a cloud platform and has started leveraging cloud-based vision
and machine-learning solutions to reduce inventory carryovers by more than 50 percent, improving
the accuracy of demand planning quarter over quarter and gaining granular insights into how individual
SKUs are performing.
AON Securities needed a quantum leap in compute power to efficiently process complex financial
modeling. By using on-demand graphical processing units (GPUs) in cloud, the organization gets easy
access to large numbers of GPUs and the ability to spin them up quickly and inexpensively, an innovative
capability that is a challenge to construct on its own. Now its clients can run and rerun Monte Carlo
simulations millions of times with different variables, all in parallel, recalculating policies and trades in
minutes rather than hours or days. That frees the business to ask a lot more questions without needing
to schedule workloads or stand up infrastructure.ix
Innovate
The next dimension involves harnessing cloud to accelerate or enable innovation using technologies
such as advanced analytics, IoT, and automation at scale. These provide companies with ways to pursue
innovation-driven growth and optimize costs for business operations. The range of potential value is
large and reflects the fact that not all organizations have the cloud maturity to achieve a similar degree
of innovation. We analyzed 700 use cases to determine the impact of cloud in unlocking value. The value
was allocated across a range from 100 percent in select cases, 30 percent in the bulk of cases, and null
in a small number of cases (Exhibit 4).
700+ use cases were split into three categories by their ... and then their respective share of value was attributed
level of public-cloud dependency ... to the value of public cloud
Use-case value
Use cases per for the industry,3 Public-cloud
category, %2 Category Definition Use-case example $ billion allocated value, %
Unlocked Use cases that can- Design to value in retail: Full value allocated,
~15 100 given scale of data
not be implemented Using advanced analytics
16 without public cloud through sentiment analysis, volume and
due to complexity of trend modeling, and social compute power
the algorithm and listening to predict which would becost
data volumes products are likely to prohibitive to stand
succeed in private labels up on-premise
Accelerated Use cases that can Predictive maintenance in Partial value
6–11 ~30
be implemented on- advanced industries: allocated, given
premise but will Predicting defects by cloud will expand
benefit from cloud managing quality of value via
capabilities—namely, products and analyzing accelerated
time to market and what drives performance product
scalability of vehicles development
76 and scalability
On-premise Use cases that will Demand forecasting in None of the value
not likely benefit healthcare: Forecasting ~1 0 of the use case is
incrementally from demand for healthcare allocated to
cloud capabilities services and reducing public cloud
and would most readmissions through
likely be imple- targeted discharge setting,
mented on-premise, due to strong regulations
sensitive to (eg, HIPAA compliance)
regulation, privacy, and relatively low data
bandwidth, and volume
7 latency
1
Also includes core operations in rejuvenation dimension.
2Figures may not sum to 100%, because of rounding.
3Adjusted to Fortune 500 scale.
Source: Industry and McKinsey expert interviews, McKinsey Global Institute research, team analysis
Logistics-and-shipping giant UPS was able to use cloud-based data processing, artificial intelligence
(AI), and machine-learning tools to design optimized routing software. When delivering more than 20
million packages a day around the world (more during peak times), UPS drivers make an average of
120 to 125 pickups and drop-offs. The number of possible routes is nearly 200 digits long. Machine-
learning models capable of processing one billion data points a day examine package weight, shape,
and size, as well as facility capacity across the network, to save UPS up to $400 million a year and
reduce fuel consumption by ten million gallons a year.
Melbourne-headquartered Hanes Australasia sells its products through its approximately 550 stores,
14 websites, and extensive wholesale network. The organization is moving away from a manual, labor-
intensive way of recommending products and toward using cloud to offer AI-based recommendations.
It has integrated a cloud-based recommendation engine into pages for 10,000-plus products for its
popular brands, and initial A/B testing has identified double-digit uplift in revenue per user session.
Hanes Australasia plans to extend integration to additional sites within its portfolio and to personalize
the marketing emails it sends to customers.
To expand into consumer banking, Goldman Sachs launched both its Marcus consumer-lending
product and its inaugural credit card with Apple in cloud. The card product required close integration
of three different ecosystems to support collaboration and a seamless user experience. Within six
months, the offering had already attracted several million customers and had scaled to meet demand.
CEO David Solomon stated, “The only reason we were able to deliver these capabilities digitally and at
scale is because of cloud technology.”x This success has sparked leadership’s imagination, leading to
an ambitious expansion of new offerings as well as enhanced transparency into its operations.
Chevron set up a cloud-readiness acceleration program to expedite its move to cloud. In 18 months,
the program expanded from 40 to 450 people across 40 scrum teams. With an automated CI/CD
pipeline, Chevron accelerated its cloud migration, the release of new applications, and the delivery of
its digital platforms, accommodating around 3,000 active users and 400 live projects. It now delivers
code in seven minutes and new infrastructure in as few as 30 minutes, making it faster, easier, and
cheaper to deliver new business functionality.
Cloud infrastructure allowed Zoom to efficiently add capacity at the rate of 5,000 to 6,000 servers each
night to meet demand during the early days of the COVID-19 pandemic. Says CEO Eric Yuan: “When the
pandemic crisis started, our own data centers could not scale fast enough to handle the unprecedented
traffic.”xi
Indonesian e-commerce giant Tokopedia needed a solution for issues with network scalability and
reliability. The platform was frequently strained during high-profile online events, most notably for
its Tokopedia Play product, which could support only 55,000 concurrent users. Cloud enabled the
company to rebuild Tokopedia Play as a microservice in five weeks to support 1.5 million concurrent
users. At the same time, overall operating cost declined due to more effective scaling.
Pioneer
As its name suggests, pioneer, the third dimension of cloud adoption, is where an enterprise can extend
cloud’s value once it has reached a certain level of cloud maturity. At this stage, companies can harness
cloud to experiment with new technologies, such as blockchain, quantum computing, augmented and
virtual reality, and 3-D printing.
Nestlé is experimenting with blockchain to achieve unparalleled transparency into its supply chain, from
origin to store shelves. The company uses a cloud-native blockchain solution to store supply-chain
transactions in ways that are transparent, immutable, and verifiable. Similarly, BMW has implemented
blockchain to track a vehicle’s history from manufacture to registration to maintenance to resale.
Quantum computing is expected to provide significant performance improvement and thus potentially
disrupt existing business models. By moving infrastructure to cloud and adapting the operating model,
companies will be better positioned to benefit from cloud-based quantum computing when relevant use
cases emerge. CSPs already offer these computing services, which allow organizations to run hybrid
quantum and classical algorithms.
Other emerging technologies, such as augmented and virtual reality and 3-D printing, also have
immense promise. For example, the health-tech company Axial3D provides clinicians with patient-
specific 3-D anatomical models, using a cloud-native integrated development environment for machine
learning.
Experience matters, and companies with high cloud maturity exhibit different adoption mindsets
compared with their peers. Third-party primary research on 705 users of public cloud indicates that
companies with higher cloud maturity share a number of traits: they are early adopters of cutting-edge
technology (71 percent), aggressively innovate (72 percent), and view technology as a competitive
differentiator and key enabler for launching and building new businesses (79 percent). By being the first
ones to move, these organizations gain considerable experience on cloud, outstripping their peers in
cloud outcomes (Exhibit 5).
Building capabilities yields tangible results. Highly mature cloud companies pursue excellence in
knowledge and skill sets, which translates into a cloud-literate workforce. To build capabilities within the
workforce, these companies create tracks for role and career progression specifically for cloud experts,
and they build tailored learning programs to develop cloud-specific skills and competencies. They also
ensure that all workers across the enterprise receive on-the-job training about relevant cloud capabilities.
Exhibit 5
Mature cloud
Mature cloudadopters
adoptersread
reaplarger
largerbenefits
benefits than
than average
average cloud-embracing
cloud-embracing
companies.
companies.
Average
Mature cloud adopters
0 20 40 60 80 100
Downtime (minutes) 57 58
The acceleration in digital engendered by COVID-19 is likely to continue far beyond the COVID crisis,
and companies must be prepared to respond and adapt rapidly. Cloud can not only help organizations
move faster and reduce IT costs but can also support innovation and the integration of powerful,
disruptive, emerging technologies. However, companies can capture their share of the trillion-dollar
prize only when they develop a clear view of the value at stake and the business cases they need to
prioritize.
To quantify the total potential value that companies could development and maintenance and infrastructure
generate by adopting cloud, we conducted detailed spending, drawing on double-blind surveys of more
analysis based on three reports from the McKinsey than 1,000 cloud-adopting companies conducted
Global Institute (MGI); McKinsey D2020 benchmarking by Omnicom Group, an independent third-party
for IT spending structure based on more than 1,000 IT market-research firm. The cost baseline has been
diagnostics worldwide; and independent third-party assessed by drawing on McKinsey’s proprietary
surveys of more than 1,000 organizations that have D2020 knowledge base, which encompasses holistic
adopted cloud to pursue potential operational efficiency IT-performance diagnostics conducted at more than
gains. In applying the MGI research, we assessed more 1,000 clients spanning more than 20 industries and all
than 700 use cases across 19 industries. We used industry- geographies, and IHS Markit and Oxford Economics
level, historical, real-revenue growth rates (4 percent market projections. In addition, we have analyzed
weighted average across industries) to forecast revenues reports by Ponemon Institute, IBM Security, and
to 2030. To forecast 2030 EBITDA and costs, we applied CyberEdge Group to estimate the value in reduced risk
the 2019 EBITDA margin (17 percent) to the projected of breaches and lower spending on cybersecurity and
2030 revenues. We estimated the revenues and EBITDA of breach management. Last, we assessed the potential
Fortune 500 companies to 2030 to establish a baseline for improvement in back-office performance from cloud’s
the assessment. The analysis identified three dimensions ability to accelerate and unlock technology use cases.
for cloud adoption (Exhibits 1 and 2).
For the second dimension, innovate, we assessed
For the first dimension, rejuvenate, we calculated the revenue uplift (the margin impact of revenue increases)
potential value from IT cost efficiency across application and cost savings from business operations. We
Exhibit 1
Valuewas
Value wasestimated
estimated from
from thethe bottom
bottom up, based
up, based oncases,
on use use cases, benchmarks,
benchmarks, surveys,
surveys, and expert
and expert input. input.
quantified the potential value of more than 700 use cases any reasonable precision. We anticipate being able to
involving advanced analytics, IoT, and automation. Here, we calculate the impact of this dimension within the next three
categorized use cases as not requiring public cloud and to five years as case evidence matures.
as accelerated or unlocked by public-cloud technology,
attributing respective share of value to public cloud. The For further reading, on McKinsey.com:
output was a detailed estimate of value by dimension, “Three actions CEOs can take to get value from cloud
benefit driver, and industry. computing”
“Debunking seven common myths about cloud”
The third dimension, pioneer, involves exploring business “Unlocking value: Four lessons in cloud sourcing and
models by experimenting with new and emerging consumption”
technologies such as blockchain, quantum computing, “How CIOs and CTOs can accelerate digital transformations
augmented and virtual reality, and 3-D printing. Given the through cloud platforms”
nascent stage of these technologies, it is far too early to “Capturing value in the cloud” (collection)
quantify their potential impact over the next decade with
The authors wish to thank Anupama Agarwal, Venky Anant, Ankush Chauhan, Vijay D’Silva, Ewan Duncan, Andreas Langholz,
Erik Rignér, Roger Roberts, Sophus Svarre Rosendahl, Nikita Smagin, and Mariss Tubelis for their contributions to this article.
This article draws on multiple sources of insight and data across sectors, geographies, and technologies, including case
examples and data from Amazon Web Services, Google Cloud, and Microsoft Azure. Amazon Web Services also provided
access to third-party research.
i
Tyler Clifford, “How Moderna is using Amazon cloud to produce personalized cancer vaccines,” CNBC, January 15, 2020, cnbc.com.
ii
“AWS powers Moderna’s digital biotechnology platform to develop new class of vaccines and therapeutics,” Business Wire, August 5, 2020,
businesswire.com.
iii
AWS for Industries, “Automating GxP compliance in the cloud: Best practices and architecture guidelines,” blog entry by Susant Mallick and
Chris McCurdy, November 19, 2019, aws.amazon.com/blogs.
iv
“Moderna announces positive interim phase 1 data for its mRNA vaccine (mRNA-1273) against novel coronavirus,” Moderna, May 18, 2020,
investors.modernatx.com.
v
Clifford, “How Moderna is using Amazon cloud.”
vi
Throughout this article, “cloud” refers to public cloud.
vii
Assuming 50 percent of the estate migrated to cloud.
viii
“EasyPay aims to be USD 100mn revenue company by FY22,” Outlook, September 23, 2020, outlookindia.com.
ix
“Aon Securities Inc. case study,” AWS, 2014, aws.amazon.com.
x
“AWS re:Invent 2019: David Solomon of Goldman Sachs talks about using AWS to innovate,” AWS, December 4, 2019, video, 3:43, youtube.com.
xi
“Zoom first quarter 2021 earnings webinar: Prepared remarks,” Zoom, June 2, 2020, webinar, investors.zoom.us.
xii
For more, see Jayne Giemzo, Mark Gu, James Kaplan, and Lars Vinter, “How CIOs and CTOs can accelerate digital transformations through
cloud platforms,” September 2020, McKinsey.com; and Chhavi Arora, Tanguy Catlin, Will Forrest, James Kaplan, and Lars Vinter, “Three
actions CEOs can take to get value from cloud computing,” July 2020, McKinsey.com.