Practical AI

for CEOs
A playbook for realizing value
from your AI initiatives

When it really matters. 1

“AI will transform every
industry, but getting the
most from your investment
in AI takes discipline and
a pragmatic approach
aligned with your business
objectives.”
Hoyoung Pak
Global Co-Leader, AI and Data Practice Group,
Partner and Managing Director

2
At a glance

Practical AI for CEOs

Creating value through your AI initiatives requires
alignment and focus across 3 critical areas

Focus on the right business problems to identify

where AI should be applied

1. Strategic alignment
Align and shape business strategy with AI

Strategy 2. Stakeholder engagement and change management

Secure stakeholder buy-in to drive change
Blueprint for Al-driven transformation
3. Use case selection
Target high-impact use cases that deliver measurable results and meet ROI goals

Effective execution turns strategic vision into

measurable results

1. Model building
Build tailored, practical Al models

Execution 2. Model deployment

Deploy into ongoing operations and workflows to begin delivering results
Tangible value creation
3. Continuous learning and improvement
Refine models with new attributes and data, adapting to shifting patterns

4. Performance and ROI tracking

Track progress along key metrics to deliver ROI

None of this works without the right foundations

Technical Organizational
1. Data 1. AI skills
Foundational Secure and expand data assets—
the lifeblood of AI initiatives
Hire, retain, and engage talent with
AI skills
pillars 2. Org structure
2. Technology
Building blocks for Structure AI organization to drive
Build scalable AI infrastructure and
Al excellence strategic initiatives and foster
forge strategic partnerships
collaboration
3. Operations 3. Risk & compliance
Streamline for AI efficiency and Embed an appropriate control
performance environment to identify and manage
risks, ensure compliance, and enable
trustworthy innovation
3
Contents
Page Section

06 Why do you need a playbook?

07 Strategy:
A blueprint for AI-driven transformation

09 Execution:
Turn vision into tangible value creation

12 Foundational pillars:
Building blocks for AI excellence

16 How can AI inform your strategy

going forward?

4
 CEOs say AI is their most
important opportunity
1

But
More than 80% of all AI
projects fail
2

1. 2024 AlixPartners Disruption Index

2. AlixPartners estimate 5
Why do you need
a playbook?
Common traps The AI race is on. CEOs recognize it as the new
force multiplier. AI has the potential to both
reshape productivity and drive commercial
– Lack of alignment with business effectiveness. In just the past year:
strategy
– $27 Billion invested in AI startups
– Failure to build strong foundations
– $150 Billion invested by corporations in AI
– Poor use case selection applications and projects

– Cultural impediments to change – Model capabilities evolving at an

unprecedented pace
– Gap to operationalization
– As their effectiveness increases, competitive
alternatives are proliferating

How do we avoid risks and pitfalls?

This is intended as a guide to help you AI

manage a broad array of competing Basics
stakeholder demands, as well as the For a summary of the
fundamentals of AI,
development of your own understanding please click here

and capabilities.
CEO:
Building a fundamental understanding of AI and its capacity to
change the trajectory of your business

Investors/Board:
Ensuring AI investments will make the highest impact

Customers:
Anticipating shifting demands and outpacing competitors

Employees:
Leading to a brighter, AI-empowered future

6
 Strategy
Do you need AI? Are you using it
in the right places?
Developing the right AI strategy is essential for integrating these technologies with business
objectives to drive innovation, create value, and gain a competitive advantage. Starting with
that business objective is essential. Where we see failure—time and again—is when leaders
lose sight of the business problem they are trying to solve and chase (and by that we mean
pour precious resources into) the shiniest new thing. Only by focusing on the right business
problems can you ensure that you will receive a meaningful return on these investments, drive
long-term value for customers and shareholders, and chart a sustainable trajectory into an
AI-empowered future.

1. Strategic alignment Key elements:

AI strategy: Align AI strategy with corporate strategy,
financial goals, value proposition, and customer needs.
Ensures your AI initiatives are tied to your value proposition
and business goals, informed by the competitive landscape, Competitive intelligence: Analyze competitors’ AI
and adequately funded. strategies to guide strategic positioning and areas for
differentiation.
Critical question: AI investment strategy: Determine funding required to align
Is the AI strategy aligned with business goals and AI strategy with business objectives to focus investments
backed by strategic investment decisions? And are our where they can create value and ensure long-term success.
AI initiatives reshaping our corporate strategy?

2. Stakeholder engagement Key elements:

Focus on people, not tech: AI, like any technology, only has
and change management an impact when it is adopted and used by people across the
organization.
Identify the right stakeholders: Select influential
Ensures integration of AI initiatives into business practices.
stakeholders who can drive AI adoption and manage change
Getting early buy-in from key stakeholders can help avoid
effectively.
cultural or personal roadblocks. Can also provide key insights
into use case selection. Training and empowerment: Equip stakeholders with
fundamental AI skills through training and support to
Critical question: cultivate AI adoption.
Are we effectively managing change and engaging
stakeholders in our AI journey?

3. Use case selection Key elements:

Strategy alignment: Ensure each use case is linked to
strategic business and financial objectives.
Selecting the right use cases is essential for focusing AI Feasibility analysis: Assess the feasibility of AI use cases,
efforts on high-impact areas and resource allocation. considering data availability, technological maturity, and
implementation complexity.
Critical question:
Near-term results: Pick use cases that have the strongest
Are AI use cases in line with expectations for impact,
foundational pillars and can be executed quickly to show
feasibility, and ROI?
results.
Value, investment, and ROI assessment: Evaluate
business impact and investment required for use cases to
ensure they deliver value and meet ROI goals.

7
Client story
Finding the right use case
can drive material results
When your strategic
imperative is growth
aligned to strategic goals.

Supercharging sales at a large-scale retailer

47%
higher revenue from
through hyper-personalization
A major retailer with over 300 brick-and-mortar stores and a strong digital operation faced
major growth challenges in spite of their well-rounded in-person and online presence. Sales
contacted customers
and marketing teams needed to use AI to supplement limited resources and develop new
strategies that could unlock stronger sales.

45%
AlixPartners was brought in to overhaul business as usual—and put in motion a suite of
customized AI solutions that personalized the customer experience.

Solving the seemingly unsolvable took a 10-week sprint. We swiftly developed proprietary
higher clickthrough AI models to estimate customer lifetime value; built response propensity models to
identify which customers were most likely to reply to sales team contact; created micro-
segmentation models to hone marketing messages to specific sub-groups; and used
generative AI and Bayesian optimization techniques to test dozens of messaging strategies
and optimize each one per micro-segment. Each AI solution helped executives make
confident, calculated moves—and unlock big gains.

The retailer quickly saw an increase in sales, as the AI models predicted. The sales team
improved revenue by 40% among customers who were contacted, compared to customers
who weren’t. The marketing team sent optimized messages, which had 45% higher click-
through rates than previously circulated language. Ultimately, our team delivered the vision
and know-how to deploy AI solutions with 47% higher revenue outcomes from contacted
customers than the retailer’s incumbent methodology.

8
 Execution
Developing a good proof of concept is one
thing. Deploying it, and achieving business
results, is another. Your technology is only as
good as your execution.
Execution is the critical phase for realizing these technologies’ potential. Only through effective
execution can you turn strategic plans into operational successes that enhance innovation,
financial performance, and competitive positioning.

1. Model building Key elements:

Data processing: Collect, clean, and process data to feed
Model building—developing a program that analyzes datasets into the AI models, ensuring quality and relevance.
to find patterns and make predictions—lays the groundwork Model development: Design and develop AI prototypes to
for AI success, translating data into predictive models that address specific business challenges or opportunities.
drive business value. Model testing/refinement: Rigorously test AI models to
validate their predictions and refine their accuracy before
Critical question: deployment.
Does our model building process effectively
leverage available data to construct accurate
and scalable AI models?

2. Model deployment Key elements:

Integration planning: Prepare for integration of AI models
Model deployment operationalizes AI solutions, integrating into business operations, considering both technical and
them into existing systems and workflows to start delivering change management factors.
business results and creating value. Deployment management: Oversee the rollout of AI models,
monitoring for issues and ensuring smooth implementation.
Critical question: User training and support: Provide training and resources
Are we ensuring rapid and seamless integration and to end-users to facilitate adoption and maximize the utility
user readiness in AI deployment? of the AI models.

9
 Execution

3. Continuous learning and Key elements:

Continual model refinement: Regularly enhance AI models
improvement by incorporating new data and adapting to shifting patterns
for sustained accuracy.
Continuously refining AI models is crucial to adapting to new Trend analysis and response: Monitor emerging trends
data, trends, and market shifts and to ensuring effectiveness to anticipate adjustments, maintaining relevance and
and sustained value. performance of AI models.
Performance benchmarks: Review performance metrics to
measure improvement and guide continuous enhancement
efforts.
Critical question:
Have we got the right processes in place to
ensure necessary adaptation and agility?

4. Performance and Key elements:

ROI tracking: Track ROI for AI initiatives, guiding decisions on
ROI tracking progression and additional investments.
Performance monitoring: Assess AI models against
Measuring performance and ROI is essential when it comes business goals to confirm they contribute to targeted value
to validating the impact of AI, guiding strategic decisions, and creation.
optimizing resource allocation

Critical question:
Have we identified the right KPIs to measure returns
from our AI initiatives?

10
Client story
Tech expertise alone is
not sufficient. Knowing
which ML tasks solve
When you’re struggling
to meet customer
which particular problems,
and then deploying them
at speed, is what drives

demand
results.

point increase
Customers want what they want, when and
in “on-time and
in full” rates for
how they want it. Meeting their expectations
some clients
can prove difficult and expensive.
Everyone wants to keep their customers happy, but doing so in an age of ever-increasing
demands for personalized and rapid service requires rethinking supply chains and inventory
management, and eats into the bottom line.

Even companies that are rising to these challenges find their ability to deliver orders “on-
time and in full” increasingly difficult. But what if you knew in advance which orders are likely
to fail?

We have worked with clients to build machine learning models to do just that, based on the
number and nature of SKUs, inventory levels, fulfillment channels, and other factors. These
models have proven highly effective at identifying those features of an order (which, of
course, vary by company) that indicate a higher likelihood of failure and can therefore
predict with high accuracy whether an incoming order is likely to fail.

But predicting failure is not enough. Armed with this insight, we have helped companies
pursue actionable solutions, which typically include moving high-risk orders to a different
fulfillment process, perhaps by splitting orders or moving certain orders to a different
fulfillment location.

Technical understanding alone is not enough. Deep experience dealing with the business
problem, combined with the insights that AI can provide, drive the kind of actionable results
that customers demand.

11
 Foundational Pillars:
Technical
There’s no AI without the right foundations.
But don’t let the perfect be the enemy of
the good.
These foundational pillars provide the necessary infrastructure and environment to support
the strategy and execution of value-creating use cases. Without these building blocks, no
AI program can be successful. Do you have the right foundations in place to drive value at
pace? Perhaps most critically, have you invested in the quality and accessibility of your data,
because your model is only as good as the data on which it depends.

1. Data Key elements:

Data accessibility: Ensure data is accessible to relevant
Data is the lifeblood of AI initiatives, and ensuring the teams and systems, facilitating a seamless flow of
availability and quality of data powers AI insights and information for AI processes.
solutions. Many organizations discover that their underlying Governance: Establish a data governance framework to
data isn’t good enough, but that doesn’t necessarily mean manage data assets, ensuring quality and compliance.
a long delay in getting started. A good data strategy is Infrastructure: Develop a resilient data infrastructure capable
about making the right data readily available for its required of handling large volumes of data efficiently and securely.
purposes when it is needed.

Critical question:
Is our data strategy comprehensive and aligned
with our AI goals?

2. Technology Key elements:

Strategic partnerships: Cultivate partnerships that augment
AI capabilities and provide additional leverage.
The technology ecosystem is a strategic asset, providing
tools and platforms essential for AI deployment, and AI platform: Build a scalable platform encompassing
enhanced by partnerships to stay at the forefront of infrastructure, system integration, and a versatile tech stack
innovation. to support comprehensive AI applications.
Tool and model repository: Select a suite of tools
Critical question: and maintain a model repository to facilitate AI model
How are our technology choices and partnerships development, ensuring easy access and fostering innovation
strategically positioning us for AI advancement and through reuse and customization.
market leadership?

3. Operations Key elements:

MLOps: Adopt MLOps to streamline the lifecycle
Robust operations are essential for the seamless performance management of AI models, from development to deployment
of AI systems, with Machine Learning Operations (MLOps) and maintenance.
providing the framework for management, quality control, and Monitoring and maintenance: Implement monitoring
continuous enhancement. systems to ensure AI solutions are performing as expected
and to address issues.
Critical question: Continuous improvement: Foster a culture that continuously
How are we enhancing our AI operations with MLOps refines AI operations.
and continuous improvement practices?

12
 Foundational Pillars:
Organizational
The right people, structure, and guardrails
enable the realization of value at speed.
Do you have the right people and the right organizational and legal ecosystem to
build a successful AI program? Are you supporting a culture that embraces change,
collaboration, and continuous learning?

1. AI skills Key elements:

Talent acquisition: Recruit and retain ML/Gen AI experts to
build and scale AI solutions.
Building a dedicated AI team equips the business with
the specialized expertise necessary for developing and Continuous learning culture: Establish continuous
implementing advanced AI solutions effectively. professional development programs to keep the team up-to-
date with the latest AI trends and techniques.
Critical question: Collaborative integration: Encourage collaboration between
Do we have the right talent in place to execute the AI team and other business units to ensure AI solutions
on our vision? are aligned with broader company goals and user needs.

2. Organization structure Key elements:

Leadership and vision: Position AI leadership roles to drive
strategy, champion AI initiatives, and serve as a bridge
Establishing a clear organizational structure for AI drives
between technology and business outcomes.
strategic initiatives and fosters cross-functional collaboration.
Cross-functional alignment: Design organizational
structures that promote cross-departmental communication
Critical question: and project execution, reducing silos and fostering agility.
Does our org structure support both building the right
strategy, and effective execution?

3. R
 isk and compliance Key elements:
Governance framework: Implement a governance structure
that supports AI scalability and ethical standards, and aligns
Appropriate risk management, compliance, and
with regulatory compliance for AI deployments.
governance ensure AI is used responsibly and ethically,
aligning innovation with legal standards, minimizing Regulatory compliance: Stay abreast of and comply with
reputational, financial, and other risks, building trust in current and emerging regulations and community standards
AI solutions, and maximizing long-term value for the relevant to AI and data usage.
enterprise. Effective governance is necessary to make Risk management: Identify and mitigate legal, ethical,
effective investment decisions and build trust. reputational, financial, and other risks associated with AI
deployment and usage.
Critical question: Privacy and security: Enforce strict data privacy and security
Are we maintaining legal and ethical integrity measures to protect sensitive information and comply with
and holding ourselves to appropriate standards global data protection laws.
in our AI practices? Dashboards: Use the right metrics to monitor performance
and risk along the product development lifecycle.

13
Client story
Your data is one of your
most valuable assets. When you need to take
the guesswork out of
Make it work for you.

pricing
Price is the biggest lever most companies
24%
higher revenue over 5 years
have to affect their bottom line
Every businessperson is familiar with the notion of a demand curve, and the relationship
between the price charged and the quantity sold. Establishing that relationship, though, can
for one client be tricky, given the number of other factors—including marketing, promotions, seasonal
fluctuations, and competitive alternatives (to name a few)—that influence consumer
behavior.

Luckily, companies today have access to unprecedentedly rich data and to effective
tools (particularly machine learning) to interrogate that data and separate out the noise.
AlixPartners has developed sophisticated AI-driven pricing models tailored for a broad
spectrum of B2B and B2C clients, delivering targeted recommendations that enhance
profitability and revenue growth.

These recommendations inform overall marketing spend, promotional discounts, loyalty

programs, home delivery, and financing options. Too often, these decisions continue to be
made based on intuition and rules of thumb, instead of hard data. And in a world of sticky
inflation and higher cost of capital, getting these pricing decisions correct has never been
more critical.

The key is not to overcomplicate the process. Start simply, then test and validate. You rarely
need perfect data for this work; almost all companies already have sufficient information to
gain real insight.

Once the model is built and validated, we then deploy an intuitive, user-friendly interface to
provide pricing recommendations for specific time periods, and “what-if” scenarios with
estimated impacts on revenues and gross margins. AI-based demand forecasting tools can
even provide dynamic pricing recommendations at the store-SKU-level.

Gaining more visibility and control over your pricing decisions, and ultimately your bottom
line, should be at the top of every CEO’s agenda.

14
How can AI inform your
strategy going forward?
Using this playbook, you can begin to rate your AI readiness in these critical areas. What does your
specific “maturity pattern” reveal? What are the implications for any AI prerequisites or critical action
steps? Would this self-assessment look different if you asked your entire leadership team?

After taking this self assessment, use case identification and near-term remediations should
become more apparent, paving the way for swifter action and results.

Self-assessing the AI environment: An illustrative example

Key:
3 = Currently being implemented and/or established and ongoing
2 = E arly-stage planning and/or currently being implemented
1 = Not yet started and/or being considered
Maturity level:
Strategic alignment 1 2 3
Is our AI strategy aligned with our business strategy?

Stakeholder engagement and change management

1 2 3
Are stakeholders committed to supporting and driving change?

Use case selection

Have we selected the right priority use cases? 1 2 3

Model building 1 2 3
Have we yet built successful models for these use cases?

Model deployment
1 2 3
Have models been fully deployed and adopted by the business?

Scaling models
How adaptable and scalable are the models? 1 2 3

Data
1 2 3
How readily can data be applied to AI opportunities/use cases?

People
1 2 3
Do we have the necessary AI skills and competencies?

Operations
1 2 3
How strong are MLOps and continuous improvement capabilities?

Technology
1 2 3
Does our technology truly support our AI ambitions?

Risk & compliance

Do we have tailored risk, compliance, and governance roles, 1 2 3
responsibilities, and procedures in place?

15
How can AI inform your
strategy going forward?
A fear of missing out and a herd mentality can disproportionately influence technology
investments in the current market, leading to poor returns.

AI is going to transform every industry. But that doesn’t have to mean losing sight of the
business rationale for any investment in these technologies. Start with the business rationale,
then determine whether today’s AI is even capable of solving the problem at hand. Developing
a practical, straightforward plan to apply AI within your business is the only way to realize
value from these investments.

Think big, start with purpose, learn fast,

and scale with confidence.
This playbook is your guide to AI success.

For more information contact:

Hoyoung Pak
Global Co-Leader, AI and Data Practice Group, Partner and Managing Director
hpak@alixpartners.com

Luca Ridolfi
Global Leader, AI and Machine Learning, Partner and Managing Director
lridolfi@alixpartners.com

Catherine Brien
EMEA AI and Data Practice Group, Partner and Managing Director
cbrien@alixpartners.com

Greg Adams
Director, AI and Machine Learning
gadams@alixpartners.com

About us
For more than 40 years, AlixPartners has helped businesses around the world respond quickly and decisively to their most critical challenges –
circumstances as diverse as urgent performance improvement, accelerated transformation, complex restructuring and risk mitigation. These are
the moments when everything is on the line – a sudden shift in the market, an unexpected performance decline, a time-sensitive deal, a fork-
in-theroad decision. But it’s not what we do that makes a difference, it’s how we do it. Tackling situations when time is of the essence is part of
our DNA – so we adopt an action-oriented approach at all times. We work in small, highly qualified teams with specific industry and functional
expertise, and we operate at pace, moving quickly from analysis to implementation. We stand shoulder to shoulder with our clients until the job is
done, and only measure our success in terms of the results we deliver. Our approach enables us to help our clients confront and overcome truly
future-defining challenges. We partner with you to make the right decisions and take the right actions. And we are right by your side. When it really
matters.

The opinions expressed are those of the authors and do not necessarily reflect the views of AlixPartners, LLP, its affiliates, or any of its or their respective professionals or clients. This article
A practical AI playbook for CEOs (“Article”) was prepared by AlixPartners, LLP (“AlixPartners”) for general information and distribution on a strictly confidential and non-reliance basis. No
one in possession of this Article may rely on any portion of this Article. This Article may be based, in whole or in part, on projections or forecasts of future events. A forecast, by its nature, is
speculative and includes estimates and assumptions which may prove to be wrong. Actual results may, and frequently do, differ from those projected or forecast. The information in this Article
reflects conditions and our views as of this date, all of which are subject to change. We undertake no obligation to update or provide any revisions to the Article. This Article is the property of
AlixPartners, and neither the Article nor any of its contents may be copied, used, or distributed to any third party without the prior written consent of AlixPartners.

16
AlixPartners 2024
17