Magazine

t
Article

os
Technology and Analytics

rP
yo
op
tC

How AI Can Power Brand

Management
No

It can automate creative tasks and improve the customer experience.

by Julian De Freitas and Elie Ofek
Do

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
How AI Can Power Brand
Management

rP
It can automate creative tasks and improve the customer experience.
by Julian De Freitas and Elie Ofek
From the Magazine (September–October 2024) / Reprint R2405G

yo
op
tC

Mojo Wang
No

Few brands are more iconic than Nike. From its swoosh logo to its
slogan “Just Do It,” the company has mastered the artistry necessary to
build a renowned brand. So when Nike asked Obvious, a trio of Parisian
artists who make AI-inspired designs, to develop new iterations of the
Air Max sneaker in 2020, it wanted to be sure the designs wouldn’t
deviate too dramatically from Nike’s signature style. Obvious trained
Do

its generative AI model by feeding it pictures of the Air Max 1, the Air
Max 90, and the Air Max 97 and used the model to create a vast array
of design ideas. Then, drawing on their own knowledge and perception

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 1

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
of broader fashion trends along with Nike’s marketing objectives, the
trio iteratively tweaked the model until it produced a design that struck
the right balance between novelty and staying on brand. The design
incorporated many of the stylistic elements of the classic Air Max but

rP
blended them with new colors, shapes, and patterns to achieve a fresh,
cool feel. The limited edition shoes sold out in less than 10 days.

Unsurprisingly, marketers have begun experimenting with AI to

improve their brand-management efforts. But unlike other marketing

yo
tasks, such as A/B testing and bidding on search words, brand
management involves more than just repeatedly executing one
specialized function. Long considered the exclusive domain of
creative talent, it encompasses multiple activities designed to build
the reputation and image of a business—such as crafting and
op
communicating the brand story, ensuring that the product or service
and its price reflect the brand’s competitive positioning, and managing
customer relationships to forge loyalty. A brand is a promise to
customers about the quality, style, reliability, and aspiration of a
purchase. AI can’t fulfill that promise on its own (at least not anytime
tC

soon). But it can help shape customers’ impressions of a brand at every

interaction. And it can automate expensive and complex creative tasks
—including product design.

Mixing brands and automation is a delicate affair. AI has the potential

No

to adversely affect a brand, so successfully implementing it in this

context often involves confronting resistance and backlash from both
customers and employees. Nevertheless, AI is becoming an integral part
of brand management. To succeed with it, you must understand how
it is perceived by stakeholders and what can be done not simply to
mitigate their concerns but to make them avid supporters. You must be
Do

sure not to overautomate by removing any sense of human control or

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 2

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
making AI the face of the brand. And you should always keep in mind
that AI and creative pursuits aren’t opposing forces.

On the basis of examples from Intuit, Caterpillar, LOOP, and Jasper AI,

rP
along with in-depth scholarly research, we propose a framework for
thinking about the key roles that AI plays when it comes to managing
brands effectively. The most successful approaches blend the best of
human and machine intelligence to augment, not replace, human
creativity.

yo
The Four Ps of AI’s Brand Impact

AI can improve performance at each stage of the customer-management

life cycle, from acquisition to development and even retention. Those
performance improvements, in turn, can reinforce and extend a
op
brand’s equity. They can be grouped into four basic categories of
impact: Productivity—AI increases the efficiency and convenience of
accomplishing marketing tasks, improving the customer experience and
driving brand loyalty. Prediction—AI reduces uncertainty, augmenting
tC

what the brand can promise and thereby building confidence and
trust in the product and the company. Personalization—AI increases
engagement and relevance for the firm’s offering by tailoring elements
to each customer, thus forging the image of a brand that cares about the
customer’s needs on an ongoing basis. Proposals—AI offers new creative
solutions and value drivers while staying true to the brand’s essence.
No

Although new forms of AI are constantly being developed, this

framework accommodates the main roles that it can and will play. For
example, classification algorithms, like the ones that sort for spam in
your email, contribute to each of the first three Ps, whereas generative
AI can contribute to all four and is especially suited to personalization
Do

and proposals. Brands should use this framework as a simple guide for
navigating a complex and expanding industry. If an AI program doesn’t

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 3

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
contribute to any of the four Ps, it’s probably not worth the risk to the
brand associated with the technology.

Now let’s dig deeper into the framework to see how some companies are

rP
already using artificial intelligence to improve their brand management.

1. Productivity

Customer service reps are your frontline brand ambassadors. And,

arguably, the most important step in brand management is retaining

yo
expensively acquired and developed customers. The risk that a
customer will be unable to resolve an issue with a product, a service,
or a payment satisfactorily—and will then abandon the brand—is one of
the biggest challenges a company faces.
op
When customers have problems, they contact customer support, and
most would rather wait in line for a human agent than get help
immediately from a chatbot. Their biggest complaints about bots are
a lack of understanding and an inability to solve complex issues.
tC

However, unlike chatbots, humans are not eternally attentive, patient,

and cheerful—especially when faced with a relentless queue of angry
callers. Long wait times and frustrating interactions may corrode a
company’s brand and lead customers to depart.

Intuit, a global financial-technology platform that makes software

No

for personal finance, small-business operations, and tax prep,

offers products including TurboTax, Mailchimp, Credit Karma, and
QuickBooks. At one point it was dealing with a barrage of customer
questions and complaints regarding the use of its software. To improve
its customer service, Intuit wanted to provide its agents with frequent
feedback on their performance. But only about 10% of callers answered
Do

customer experience surveys (a typical response rate for call centers),

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 4

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
and managers could listen in on only a small subset of calls. So agents
had little opportunity to receive robust feedback in a timely manner.

To overcome that challenge, Intuit used transcripts of the calls that

rP
customers had rated to train an AI model to detect which interactions
were most likely to result in customer satisfaction. Because all calls were
recorded and could readily be transcribed, Intuit could use the trained
AI model to provide personalized daily feedback to all its agents based
on all calls, whether they’d been customer-rated or not.

yo
Thus the company improved customer satisfaction at a fraction of
the cost of typical human supervision or expensive training programs
for its agents. And because the employees were more effective, they
felt more fulfilled. Furthermore, because Intuit engaged employees at
op
all levels and solicited their input on how to design a dashboard to
display feedback for them, it eased concern that agents were going to
be replaced by AI and ensured that they remained the face of the brand
when customers called in to get support.
tC

2. Prediction

Caterpillar, a maker of heavy-duty construction and mining equipment,

uses AI to deliver additional value by literally foreseeing the future.
Its subdistributor Borusan Cat, based in Turkey, faced this problem:
When a customer’s equipment broke down, the repairs were often
No

very expensive because a part had deteriorated to the point where it

damaged the rest of the machine. In some cases an engine overhaul
was needed, and the downtime to get the necessary parts and conduct
the extensive repairs was costly for both the customer and Borusan
Cat. Furthermore, when their machinery did break down, customers
sometimes turned to unauthorized third-party vendors, resulting in lost
Do

business for Borusan Cat. The company believed that it could deliver
significant value by detecting part failures before they rendered the

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 5

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
equipment inoperable, much the way a medical checkup can detect a
clogged artery and prevent a heart attack.

The first step was to establish the infrastructure necessary to harvest

rP
data. The company embedded sensors in the machines to continuously
harvest information on the state and use of the machines’ parts. After
amassing enough failure incidents, the company trained AI to combine
signals from various parts of the equipment with past data patterns
to predict—with 97% accuracy—which piece of equipment was at risk

yo
of breaking down and what the exact problem would be. Using these
predictions, the company alerted customers and selectively sent its
technicians to validate diagnoses and determine the level of service
or repair required. If the customer agreed, the equipment would be
repaired with minimal if any downtime and more cheaply than if the
op
customer had waited for the machine to malfunction.

However, despite the benefits of the AI-

based technology, willingness to pay for its
insights was initially lacking. Customers
tC

thought that proactive calls from

technicians were a marketing gimmick to
sell more parts or service. Salespeople were
skeptical of the technology, especially
when it flagged equipment on a “healthy”
No

site that they had been unconcerned about.

Mojo Wang
And because each repair was a low-ticket
item compared with selling equipment,
salespeople had little incentive to pursue
repairs. For these reasons they were waiting too long to call customers
whose machines had been flagged as being at risk of breaking down—
Do

thereby failing to deliver on the company’s promise of preemptive

maintenance and in the process doing damage to the brand.

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 6

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
After a few twists and turns, management chose to abandon the uphill
battle of explaining the capabilities of the technology to skeptical
customers. Instead it absorbed the value of the technology into
its maintenance contracts. The company guaranteed customers no

rP
downtime, or else Borusan Cat would provide replacement equipment.
Thus the AI’s predictive capabilities augmented the brand’s promise
and also removed the possibility that customers would turn to third
parties if their equipment broke down. The company captured the
added value provided by the AI without drawing attention to its

yo
presence.

That fix also addressed the sales team’s financial incentives. Folding AI
insights into the maintenance contracts resulted in a higher-ticket sales
item worth prioritizing. Because it was critical that salespeople follow
op
up on AI alerts in a timely manner, the company also created a central,
cross-organizational team dedicated to that service.

With those changes in place, the company could finally realize the
technology’s potential: to repair parts before they sustained significant
tC

damage. In fact, because the parts removed from repaired equipment

were often salvageable, they could be refurbished and resold. Effectively
providing these added services each time a repair was needed gave
Borusan Cat more opportunities to interact positively with customers,
increasing the attractiveness of its AI solution while reinforcing its
No

image as a trustworthy brand that cares about customers’ business and

long-term success.

3. Personalization

Getting customers through the door is an achievement. But they may

make only a single purchase, sign up for the most basic service, or buy
Do

a limited quantity of the company’s products. In that case the company

generates modest revenues from customers it expended considerable

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 7

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
resources to acquire. And customers may become less excited about
an offering over time and feel that they are overpaying—severely
restricting their lifetime value.

rP
To see how AI can enable a company to provide tailored offerings
that keep existing customers engaged with the brand, let’s turn
to an unorthodox car-insurance app: LOOP. It doesn’t use several
standard insurance-premium criteria, such as credit score, income
level, and occupation, which tend to introduce bias against certain

yo
minority groups. It can afford to omit them because, as an AI-powered
smartphone app, it constantly collects risk-relevant data about where
customers drive (type of road, traffic volume, weather) and how
(speeding, hard braking, talking on a cell phone). Its unique approach
combines that data with extensive information on road accidents to
op
predict, using AI, whether a customer who tends to drive in a particular
way on particular roads is at high, medium, or low risk of filing a claim.

Most customers would rather wait in

tC

line for a human agent than get help

immediately from a chatbot. But unlike
chatbots, humans are not eternally
attentive, patient, and cheerful.
No

Using the AI’s predictions, LOOP makes additional offerings to

customers, including much cheaper rates. Just imagine how a LOOP
customer feels when she receives a notification like this: “You’re on a
roll, Jacky! You’ve unlocked a surprise for driving safe: lower rates!”
LOOP further incentivizes safe driving by providing a customized,
Do

continually updated driving score (say, 8.18 out of 10), along with safety
insights and tips gleaned from its treasure trove of customer and road
data combined with what it learns about individual drivers’ behavior.

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 8

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
Customers see, on a weekly basis, meaningful feedback and suggestions
about how to drive and what roads to avoid—plus actual results in
terms of their scores and rates. That goes a long way toward removing
any skepticism they may have had about whether an insurance app

rP
that eschews many of the conventional measures can accurately assess
their risk. Although several competitors offer customers some benefits
in exchange for gaining access to their telematics data (typically only
driving behavior), these are usually comparatively small discounts—and
they employ the data more as an excuse to send users telemarketing

yo
communications than to set their rates.

LOOP not only helps create safer roads; it lowers the chances that its
customers will get into an accident and file a claim, increasing their
lifetime value. In short, AI helps achieve a win-win for the customer and
op
the company while building the image of a brand that is fair and cares
about customers’ well-being beyond the initial point of sale.

4. Proposals (Lots of Them)

tC

A brand has a unique tone or personality that cuts across all company
communications, from social media and email messages to blogs and
other long-form content. A well-defined and consistent brand voice
reinforces brand image, fosters a deeper connection with the audience,
and helps the brand stand out from competitors.
No

Many managers have resisted automating brand communications,

despite the advent of large language models. First, they worry that using
generative AI means surrendering their brand’s unique voice in favor of
generic, cookie-cutter output. Second, they are wary of the tendency of
large language models to “hallucinate” facts in their responses, hurting
a brand’s reputation. Third, they are rightfully concerned that whatever
Do

they put into the system will be used to train the models and thus will
become accessible to competitors. In short, they see a trade-off between

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 9

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
marketing efficiency and staying in control of the brand’s image and
integrity.

Consider how Jasper AI, a marketing-content generator, navigates that

rP
trade-off. Let’s say you want to create a marketing campaign. You begin
by helping Jasper learn your brand’s unique tone of voice. You can
upload a style guide or link Jasper to a few examples of previous
posts that you believe best reflect your brand. Jasper will learn your
brand’s personality (attitudes and feelings about a topic), style (word

yo
choice, sentence structure, rhetorical devices), and other aspects of the
language you typically use in branded communications. For instance,
it can learn that your brand’s tone is more casual than formal, more
funny than serious, or more irreverent than respectful. Then you upload
information about your company, products, services, and audience, and
op
the goal or objective of the communication.

Jasper will help generate marketing materials for communication

campaigns. It will conjure a solid first draft of, say, a blog post that is
not only optimized for search engine visibility but is also written in
tC

your brand’s unique voice and accurately incorporates facts about your
company. Beneath the hood, Jasper leverages a family of large language
models (OpenAI, Bard, Stability.ai, and Anthropic) and searches Google
for the latest news to ensure that its proposals are up-to-date, because
most large language models have been trained on “old” data, perhaps
No

to 2019. Most important, it layers facts about your business with your
brand’s distinctive qualities while ensuring that those facts are not used
to train the underlying AI models.

You are involved throughout the process. You can ask Jasper to help
you edit the blog to better align with your desired brand attributes (for
Do

example, “infuse more excitement,” “convey a more upbeat attitude,”

“emphasize global appeal”) or to keep generating entirely new blog

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 10

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
posts until you see one you like. Jasper can also morph the blog into
other types of marketing content for your campaign, such as a landing
page, a LinkedIn post, or an email, creating a full gallery that conforms
to the format of each asset type while preserving the essence of the

rP
brand. And if you need to replicate your campaign for a different
product or audience (say, consumers in the developed world), simply
upload information for that audience, switch out audiences in a drop-
down menu, and let Jasper regenerate all the materials.

yo
In these various ways Jasper can eliminate the trade-off between
efficient marketing and controlling your brand identity, allowing you to
creatively communicate at scale without losing your brand’s distinctive
voice, veering away from the facts, or giving up your trade secrets.

...
op
AI cannot and should not automate all aspects of brand management.
Our examples emphasize the importance of integrating it into
conventional marketing efforts without making it the face of the
brand or relying on it so much that it usurps your control. It should
tC

complement rather than substitute for the roles that human brand
builders play.

As Nike’s partnership with Obvious proved, managers can combine

human intelligence with artificial intelligence to push the boundaries
No

of what their brands can do. Successful brand managers will be those
who master bringing out the “art” in artificial intelligence.

A version of this article appeared in the September–October 2024 issue of Harvard Business
Review.
Do

Julian De Freitas is an assistant professor in the marketing unit at

Harvard Business School.

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 11

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860
 HBR / Magazine Article / How AI Can Power Brand Management

t
os
Elie Ofek is the Malcolm P. McNair Professor of Marketing at Harvard
EO Business School.

rP
yo
op
tC
No
Do

Copyright © 2024 Harvard Business School Publishing. All rights reserved. 12

This document is authorized for educator review use only by Tien Nguyen, The University Of Da Nang until Sep 2024. Copying or posting is an infringement of copyright.
Permissions@hbsp.harvard.edu or 617.783.7860