What it takes to rewire a CPG

company to outcompete in
digital and AI
Answering six specific questions holds the key to successful digital and AI
transformations for CPG companies.

by Abdul Wahab Shaikh and Shruti Lal

with Hannah Mayer and Spurthi Gummadala

© Getty Images

June 2024
 There’s a race on to capture value from digital Our analysis of digital and AI maturity has shown
and AI, and consumer packaged goods (CPG) that CPG companies are among the poorest
companies are in danger of falling behind both performers, while retailers are near the top
retailers and consumers. It’s not for lack of trying. (Exhibit 1).1
Like most sectors, CPG companies have been on
some form of digital and AI transformation journey. This halting progress is all the more frustrating
But most of them are stuck in the pilot purgatory and worrisome in view of the huge value at
stage characterized by plenty of subscale activity stake. We analyzed the potential of digital and
and little at-scale value. AI transformations to drive top- and bottom-line

1
The scores are based on Digital Quotient (DQ) and AI Quotient (AIQ) assessments that measure digital and AI maturity across core
capabilities and management practices essential to capturing value. The DQ and AIQ gap between leaders and laggards was 10.3 percentage
points in the period from 2016–19; the gap increased to 16.3 percentage points in the period from 2020–22.

Exhibit 1
Digital and
Digital andAIAI maturity
maturityscores
scoresforfor individual
individual companies
companies show
show significant
significant spread
spread within each
within each sector. sector.

Digital and AI Quotient scores by sector¹ Maximum

x Average
Minimum
100

80

60

40 40 41
35 37 37 37
31 32 34
30 30
25
20

0
Public Indus- PMP² CPG³ Energy Health- Trans- Banking Insur- Media Retail High
sector trials and care port ance tech
materials

1
Digital Quotient (DQ) and AI Quotient (AIQ) assessments measure digital and AI maturity across core capabilities and management practices that are essential
to capturing value.
2
Pharmaceuticals and medical products.
3
Consumer packaged goods.

McKinsey & Company

2 What it takes to rewire a CPG company to outcompete in digital and AI

impact along the full value chain.2 Our analysis centralized stage-gate process with clear decision
revealed a potential 6 to 10 percent incremental rights to track, advance, and fund solutions to help
revenue uplift and corresponding growth of 3 to maintain momentum.
5 percentage points in EBITDA over three to five
years, depending on the subcategory. To support the technology solutions, they moved
aggressively into cloud to help create more scale
Furthermore, the increased adoption of generative and flexibility to use in building modular, customized
AI (gen AI) could increase the economic impact applications. They also invested in a data lake and a
of traditional AI by 15 to 40 percent, unlocking an governance model with clear areas of responsibility.
additional $160 billion to $270 billion annually in For example, IT defined and managed the data
profit (measured in EBITDA) for CPG companies architecture while the business defined use cases
globally. 3 for data products. Realizing talent was an acute
issue, they upskilled their own people through
The CPG sector faces some unique challenges. tailored learning curricula that included on-the-job
The proliferation of data, for example, and its learning, formal training sessions, and individual
complexity—sources are scattered across retailers, coaching.
suppliers, manufacturers, and consumers—have
created massive issues in terms of harnessing
the data to find, track, and capture value. At its Six questions to help outcompete with
core, the reason for this low success rate is that digital and AI
companies fail to perform the deep organizational In working with more than 200 large companies
surgery required to affect the broad-based change across industries, including 25 of the top 30 CPG
that’s needed. It’s never “just tech” when it comes companies, we found that six enterprise capabilities
to successful digital and AI transformations. are critical for companies to rewire themselves and
Companies need to rewire how they work. 4 achieve sustainable competitive advantage from
digital and AI (Exhibit 2). In rewiring how they work,
But even while CPG as a sector performs poorly, CPG companies need to answer six key questions.
some companies are high performers, as Exhibit
1 shows. One beverage company that embarked 1. Where is the value?
on a digital and AI transformation unlocked 18 Making sure that a digital and AI transformation
percent EBITDA uplift over two years. It was able delivers meaningful value starts with prioritizing
to make and sustain these improvements by being and focusing efforts on domains where meaningful
comprehensive in the scale of, and commitment to, value exists. While the distribution of value varies
the change needed. across CPG sectors, our analysis shows that the
greatest payoff for most sectors is concentrated
A committee of senior leaders started by heavily in two areas: consumer insights and demand
prioritizing domains that had significant growth creation, and customer and channel management
opportunities and were feasible, given their (Exhibit 3). There is one notable exception: the
capabilities. They then developed a detailed road beauty industry, where the direct-to-consumer
map and established more than 50 cross-functional area takes center stage. New technologies can
pods—joint teams spanning markets, regions, and have a big impact on that interaction model, as
enterprises—with specific goals and the autonomy well as on the e-commerce process and fulfillment
to deliver necessary solutions. They put in place a

2
Analysis based on McKinsey’s CPG digital and AI impact calculator tool (latest version June 2023), which consists of a repository of more
than 60 digital and AI use cases across 11 domains in front (precision revenue growth management; data-driven marketing; sales and in-store
excellence; portfolio innovation and design; e-commerce and direct to consumer), middle (autonomous planning; plant of the future; digital
logistics; digital procurement and supply management and supplier collaboration), and back (talent analytics; general and administrative
services).
3
“The economic potential of generative AI: The next productivity frontier,” McKinsey, June 14, 2023.
4
Eric Lamarre, Kate Smaje, and Rodney Zemmel, “Rewired to outcompete,” McKinsey, June 20, 2023.

What it takes to rewire a CPG company to outcompete in digital and AI 3

 Web 2023
McKQ-RewiredToOutcompete
Exhibit 22
Exhibit 2 of

Six enterprise capabilities are critical for successful digital and AI

transformations.
Six enterprise capabilities are critical for successful digital and AI transformations.

Transformational value comes from careful and coordinated execution across all areas of focus

Alignment 1. Business-led digital road map

on value Align senior leadership team on the vision, value, and road map for the transformation;
reimagine business domains to deliver outstanding customer experiences and to lower unit costs.

Delivery 2. Talent 3. Operating model 4. Technology 5. Data

capabilities Ensure that you Increase the Make technology Continually enrich data
have the right skills metabolic rate of easier for teams to and make it easily
and capabilities to the organization by use so they can accessible across the
innovate and execute. bringing business, innovate at pace. organization to help
operations, and improve customer
technology together. experience and
business performance.

Change 6. Adoption and scaling

management Maximize value capture by ensuring the adoption and enterprise scaling of
digital solutions and by tightly managing the transformation progress and risks.

McKinsey & Company

management. In general, these are the domains the marketing function, with a value between 5 and
CPG companies should prioritize for their 15 percent of total marketing spending.6
transformation programs.
One consumer company, for example, implemented
As CPG companies assess the value, they will a gen AI large language model (LLM) to improve the
need to be thoughtful about understanding gen manual process of financial planning and analysis
AI’s impact. McKinsey analysis identified just (FP&A) research. An initial proof of concept
four areas—customer operations, marketing showed a reduction of up to 30 percent in time
and sales, software engineering, and research spent on research.
and development—that could account for
approximately 75 percent of the total annual Some companies are already actively educating
value from gen AI use cases. 5 The majority of their organizations, especially leadership,
that value comes from increased productivity in about gen AI, the capabilities it unlocks, and its
the form of, for example, better and faster issue applications in business. In fact, since the second
resolution in customer service, more personalized quarter of 2023, almost all CPGs we analyzed
communications, and more-effective product have had an immersion session on AI for C-level
discovery. Gen AI could increase the productivity of executives.

5
“The economic potential of generative AI,” June 14, 2023.
6
Ibid.

4 What it takes to rewire a CPG company to outcompete in digital and AI

Exhibit 3
Digital
Digitaland
andAI
AIimpact
impactvaries
variesbyby
sector.
sector.

Share of impact by value stream and sector,¹ % Low High

Personal care and home
Value stream Food and beverage and personal health Beauty
Consumer insights and demand creation 19 37 23
Product and innovation 8 10 7
Manufacturing and operations 15 4 3
Supply chain planning and logistics 14 7 11
Customer and channel management 30 23 25

Direct-to-consumer 11 18 30
Core 2 1 2

1
Figures may not sum to 100%, because of rounding.

McKinsey & Company

2. Are leaders from the business side actively The core operational unit is the agile “squad.”
part of the transformation? Comprised generally of five to eight people across
CPG companies often underestimate the role business, data, technology, and design functions,
of business functions in a successful digital and these teams are responsible for building the
AI transformation, relegating the initiative to IT. solutions on the road map (Exhibit 4). They are
Case in point: At a large food manufacturer, the crucial for scaling. Acting autonomously based
supply chain domain was one of the company’s on clear guidelines, these agile squads are the
most successful domains, in part because the only way companies can enable hundreds or even
leadership, including the COO and CFO, made thousands of teams to deliver transformational
supply chain a priority. The business dedicated change.
a senior director as product owner for not only
the build but also rollout and adoption, and these 3. Are you an attractive long-term employer for
business leaders joined biweekly sprint reviews, digital talent?
engaging with the supply chain digital leader and Without top in-house technical talent, CPG
the working team in designing a completely new companies will struggle with their transformation.
way to optimize customer service. Having a hiring strategy and competitive
compensation will only get you so far in attracting
Top business talent should act as product owners and retaining digital talent, especially when you
of the transformation for their given product, are competing with digital natives. The core issue
working closely with technology leaders to define is that work at CPG companies often doesn’t
the road map of solutions, managing the pipeline attract top tech talent. Companies will need to
of use cases to build the solution, influencing offer meaningful missions, learning opportunities,
technology infrastructure decisions, governing and an environment where tech talent can thrive.
data, and being the voice of the transformation to
their wider teams.

What it takes to rewire a CPG company to outcompete in digital and AI 5

 Exhibit 4
Agile squads
squads are
are the
the core
core operating
operatingunit
unit of
of the
the transformation.
transformation.

Illustrative product and platform squads

Talent (to be assigned to various squads) Product squads

Product
owner

Software
engineer

Data
engineer

Agile
coach

Data
scientist

Design
lead Platform
squads
Subject-
matter
experts

Support and control functions for all squads

Typical squad structure

Product owner Software Data engineer Agile coach Data scientist Design lead Subject-matter
defines and engineer builds data supports and mines business creates experts
prioritizes develops code, pipelines to coaches the data to identify customer- bring expertise
product road writes unit drive analytics squad team patterns and centric design, and knowledge
map and tests, and solutions from on agile build predictive develops user of business,
backlog drives different data development models engagement functions,
integrations sources practices plan, and operations, legal,
conducts user risk, and
testing compliance

McKinsey & Company

A good place to start is upskilling leaders to better It’s important to focus on hiring talent with
understand how tech creates value. P&G instituted some proficiency in relevant areas. Competent
a reverse-mentoring program in which junior tech developers are significantly more productive than
people worked with senior leaders to help them inexperienced ones, and that trend carries over
understand how to use tech.7 Leaders also went to into gen AI as well. 8 To find the right talent, CPGs
leading digital businesses to observe operations should look to suppliers or retailers that are already
and speak with leaders to understand what skills progressing on their digital journey. Adjacent
are important to bring into an organization. sectors such as hospitality and telecom can also
be sources of expert talent with broader skills that

7
“How P&G’s 4D culture helps with successful digital transformation,” Human Resources Online, May 12, 2020.
8
“The economic potential of generative AI,” June 14, 2023.

6 What it takes to rewire a CPG company to outcompete in digital and AI

are transferable to CPG companies. Offering (RGM) decisions. Getting to this state requires
promotions and compensation based on skills companies to develop a modular architecture
mastery and establishing engineering-specific (Exhibit 5).
career tracks can also help in retaining your top
people. For this approach to work in practice, companies
need a global strategy and a team made up of
A large beverage CPG knew it needed to experts from central IT (including enterprise
upgrade its talent if it wanted its digital and AI architects, cloud developers, and engineers) and
transformation to succeed. But established leaders from target markets to design the system
recruiting practices were slow and not geared to together. They should focus on creating modules
the leading technical talent the business needed. to support three to five market archetypes (for
A talent win room came together with a new example, hypermarkets in the United States and
plan. First, they invested in developing clear new more-traditional trade stores in Latin America),
role descriptions, tailored to the specific skills develop a road map for building them, and then
required. Second, they focused on identifying test them in the market. Typically, this process
new recruiting sources rather than turning to includes making key decisions, such as what
more general jobs and networking venues. Third, domain tools to deploy, what cloud infrastructure
they sped up the evaluation process with coding to develop, and whether to build or buy specific
tests, which allowed them to quickly narrow technologies.
down more than 7,500 initial applicants to high-
potential candidates. And fourth, they put in 5. Are you developing data products?
place batch days organized around the full set Data in the CPG industry is notoriously
of decision makers, which allowed them to make fragmented across retailers, distributors,
decisions on candidates and extend an offer consumers, syndicated data providers, marketing
within 24 hours. Over this time, the talent win platforms, contract manufacturers, third-party
room had a set of KPIs that they referred to often logistics providers, and more. This issue has
to track their progress and correct any issues become even more acute with generative AI, with
quickly. In 90 days, they were able to fill the 25 the introduction of huge amounts of unstructured
critical digital and analytics roles they needed—a data. Without a centralized, well-coordinated
tenfold increase in the speed of hiring. data strategy, teams end up using raw data for
their one-off needs, wasting data scientist and
4. Are you deploying your technology engineering capacity on creating inconsistent
investments to optimize for reuse? data sets that can’t be accessed by other teams
Many CPG companies have key elements of a or systems.
core infrastructure in place—such as cloud, data
lakes, and planning software—but they are often To redress this issue, companies need to focus on
not set up to operate at scale. This situation is three things:
largely the result of a complex tech stack where
systems are built to support a specific function or — Build data products. Data products are
market. This makes it challenging to share data reusable building blocks of data that can be
and reuse solutions, leading to costly replication easily consumed by teams and systems. One
of applications and difficulty in scaling. CPG focused on developing a data product on
store-level customer attributes, which teams
Enabling solutions that serve multiple domains across sales, RGM, planning, and supply chain
has significant benefits, such as allowing a used for their own use cases. To get this right,
company to use a broader range of data to the company had to understand what data
make better revenue growth management it had and which parts of it were important

What it takes to rewire a CPG company to outcompete in digital and AI 7

 Exhibit 5
A multilayer
A multilayer technology
technologyarchitecture
architecturecancanenable
enablean
anRGM
RGMsolution
solutiontotoreuse
reuse than
more more60than 60 percent
percent of its components.
of its components.

Illustrative components
Market layer
Tool kit
configuration 25% standard
for markets Integration of RGM¹ to User interface workflow Customized for individual market by
TPM/ERP² system integrating price/promo delivery teams
modules

Use-case layer
Modules built
to solve specific 50% standard
use case Pricing Promotion Assortment Trade User-centric design-based user interface;
customized by product and deployment
teams based on channel/market

AI and data layer

Use-case-
agnostic
70% standard
components Databricks Python Power BI Gurobi Modular heavily automated components
pipeline Analytics reporting optimization serving all RGM¹ use cases
Workbench templates engine

95% standard
Base layer
Infrastructure
and software Cloud GitLab Terraform Postgres Docker Customized by product team depending
tooling platform DevOps infra- database container- on infrastructure development choices
structure ization
as code

1
RGM = revenue growth management.
2
TPM = trade promotion management; ERP = enterprise resource planning.

McKinsey & Company

to each function’s use cases. Out-of-stock — Put in place a strong data product owner.
data, for example, was important for sales The dedicated data product owner should
and planning functions, while the promotions be from the business, not the IT, side of the
calendar, promotions ROI, and volume uplift company. One of the key roles of the data
data were important for RGM (Exhibit 6). product owner is to establish and closely
monitor meaningful KPIs to track the most
— Standardize APIs and invest in data important determinants of value to the
pipelines. It’s important to create standard business, such as fulfillment, on-time delivery
APIs and systems (with the right access to warehouses or shelves, or data-product
privileges) that developers can use to access usage rates.
data sources. Invest in data pipelines to deliver
data to the right data products. Because the A global CPG was able to rapidly transform its
company often doesn’t own all the data from RGM domain end-to-end in just two years by
retailers or manufacturers, for example, that building modular technology components that
data needs to be constantly monitored for enabled it to scale solutions across its more than
changes. 30 category/market combinations. Instead of
building pricing and promotions analytics around

8 What it takes to rewire a CPG company to outcompete in digital and AI

Exhibit 6
A
A CPG
CPGcompany
companyset setup
updata
dataproducts
productstotodetangle
detangleand
andsimplify
simplifydata
datause
useand
unlock significant efficiencies in scaling.
and unlock significant efficiencies in scaling.
Illustrative CPG example

Before setting up data products

Systems Data Use-case-specific Use-case-specific Use

of record platform data sets technologies cases

Digital picture
of success
Core Data
processing warehouse
systems Personalized
marketing

Perfect order
recommendation
External Raw data
data lake
Promo pattern
recognition

Customer
Unstructured Operational performance
data data store
Order tracking

Data for each domain, such as the customer, is inefficiently reworked for every use case.

After setting up data products

Systems Data Data Consumption Use

of record platform products archetypes cases

Digital picture
Customer 360 Digital applications
of success
Core Data
processing warehouse Personalized
systems Product 360 marketing
AI
Perfect order
recommendation
Financial budget
External Raw data and actuals Promo pattern
Reporting
data lake recognition
Customer
Salesforce performance
External data
sharing
Unstructured Operational Order tracking
data data store Fleet
Discovery sandbox New use cases

McKinsey & Company

the data available for a particular region/category, underlying analytics engine. Because the data
for example, the company built a standardized product was modular, specific local factors, such
RGM data product that teams could easily use. as currencies and units of measure, could easily be
Each market category only had to input its own swapped in.
data into the product to take advantage of the

What it takes to rewire a CPG company to outcompete in digital and AI 9

 6. Are you anticipating and preparing for the The key to tackling scaling is to “assetize”
most critical scaling challenges? solutions by packaging them as modular assets
The fragmentation in CPG companies—accounts, that teams can easily reuse. The focus should be
categories, brands, geographies, and functions— on technologies, such as APIs; processes, such
makes adoption and scaling technology as solution rollouts, operational guidelines, and
challenging. Too often, companies have to redo a training; and support, such as subject-matter
lot of work to tailor solutions to local environments. experts who understand how to deploy the
solution and adapt it to different environments.
While the adoption of technology relies on
many factors, in practice it is most crucial to
involve potential future users early in solutions
development; give end users incentives to use A digital and AI transformation is a complex
the technology; minimize the effort required of journey. But for CPG companies willing to make
Find more content like this on the
them, by building solutions into existing tools, for the commitment to change at scale, the value can
McKinsey Insights App
example; and track their uptake over time. be transformative and a competitive necessity.

Abdul Wahab Shaikh is a partner in McKinsey’s Atlanta office, Shruti Lal is a partner in the Chicago office, Hannah Mayer is
an associate partner in the Bay Area office, and Spurthi Gummadala is a consultant in the Seattle office.

Scan • Download • Personalize The authors wish to thank Roger Roberts for his contributions to this article.

Copyright © 2024 McKinsey & Company. All rights reserved.

10 What it takes to rewire a CPG company to outcompete in digital and AI