Journal of Retailing and Consumer Services 69 (2022) 103119

Contents lists available at ScienceDirect

Journal of Retailing and Consumer Services

journal homepage: www.elsevier.com/locate/jretconser

Retail media networks

Darrell E. Bartholomew a, *, Mark Williamson b
a
Associate Professor of Marketing, Penn State Harrisburg, School of Business Administration, 777 West Harrisburg Pike, Middletown, PA, 17057-4898, USA
b
SVP Client Strategy & Development, CitrusAd, Harrisburg, PA, USA

A R T I C L E I N F O A B S T R A C T

Keywords: This study explores large retailers’ exchanges of customer data and personalization through targeted advertising
Retail media networks on retail media networks. The Retail Media Networks Model is introduced, and a case illustration is provided
Personalization with examples from the model illustrating personalization/privacy exchanges from the retailers’ perspective.
Data monetization
Secondary data and primary data featuring interviews with retail media network managers and technology ad
platform providers from large retailers in the US and EU. The size of retail media networks is discussed along
with other current managerial implications relevant to personalization and data monetization, including, walled
gardens, data privacy controls, and public policy issues.

1. Introduction 2014); however, there is a lack of understanding from the consumers’

perspective of how their data are managed and used in retail exchanges
AdAge: Can you give us a sense of how important advertising is to collect, store, and exchange consumer data.
Amazon? This leads us to focus on the following research questions:
I think the way Jeff [Bezos] would answer that is, if we think about
1. What are retail media networks and how do they work?
Amazon in two worlds, one world is an Amazon with ads and lower
2. Is it possible for large retail media networks to protect their con­
prices. Another world is an Amazon with no ads and higher prices.
sumer data and provide personalization?
Which one would we choose? I think nine times out of 10, or 10 times
out of 10, we would take Amazon with ads and lower prices. —Lisa
The purpose of this paper is to address these important questions. In
Utzschneider, Former Head of Amazon Media (Learmonth, 2012)
the next sections, we review the literature on data privacy and person­
Retailers have always been advertisers, but they are recently growing alization. We then address our research questions by providing case il­
their advertising businesses to pay for improved omnichannel retailing, lustrations from both primary and secondary sources to examine retail
and home delivery services. Online retailers such as Amazon, whose media networks. We then conclude with a discussion about these retail
“Amazon effect” has led traditional retailers to improve omnichannel media networks, the limitations of our research, and offer ideas for
retail offerings and provide consumers with better retail experiences future research.
through improved digital options, improved customer shopping expe­
riences, and home delivery (Fromm, 2016; Tyrväinen et al., 2020). To 2. Literature review
fund this retail transformation and pay for these omnichannel environ­
ments, retailers have followed Google and Facebook’s advertising 2.1. Data privacy and personalization
models by creating retail media networks to monetize their own con­
sumer data. For the consumer and the retail marketer, however, this Customers are very different when it comes to their willingness to
becomes somewhat of a paradox. On the one hand, consumers want exchange data with retailers, and their own data privacy preferences and
privacy, on the other hand, they desire improved omnichannel retail desired levels of protection (Pallant et al., 2022). Customers are willing
experiences through online shopping, in-store, and home delivery. to make the information exchange if they benefit from the risk and if
Recent research focuses on how companies use their consumer data there are adequate privacy controls provided (Phelps et al., 2000). Such
to target customers online with personalized offers (Schumann et al., privacy controls have been shown to be positively related to consumers’

* Corresponding author.
E-mail address: [email protected] (D.E. Bartholomew).

https://doi.org/10.1016/j.jretconser.2022.103119
Received 8 June 2022; Received in revised form 22 July 2022; Accepted 16 August 2022
Available online 23 August 2022
0969-6989/© 2022 Elsevier Ltd. All rights reserved.
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

willingness to provide information to trusted service providers if the these circumstances, privacy becomes a policy issue that must be
disclosures are clear in terms of what is being shared and who it is being addressed.
shared with (Markos et al., 2017). Retailers are relying more on con­
sumer data as a resource for personalized products, services, and com­ 2.2. Retail media networks
munications. Such data is collected from multiple sources through
on-ramping to provide a detailed view of consumers’ preferences and Large retailers are now growing their ad businesses, becoming retail
behaviors (Pallant et al., 2022). There is a tradeoff “between informa­ media networks by providing new advertising platforms that include
tion to implement personalization and the potential violation of privacy display advertising, responsive banner ads, video advertising, branded
that comes with this information” (Montgomery and Smith, 2009 p. pages, email ads, and in-store screens. All of this would be owned media
134). While incredibly powerful as an advertising tool, interest- or that retailers market through their own ad agencies or together with
behavior-based targeting creates a unique challenge for advertisers who technology partners that provide cloud-based platforms to the retailers.
are expected to walk a tightrope between relevance and privacy. On one Retailers then market this media to manufacturers of consumer pack­
hand, consumers desire relevant content and want marketers to “act like aged goods (CPGs). Traditional retail media include servicescape
you know me”—personalization. At the same time, too much advertiser advertising such as in-store advertising such as radio, print, digital dis­
knowledge can lead to consumers feeling that marketing has become plays, newspapers, and magazines. Retailers can also provide location-
pushy, invasive, uncomfortable, or creepy. Personalization provides based signage, window displays, endcaps, shelf-talkers, floor and win­
customers with content, services, products, and interactions to please dow signs, posters, and outdoor advertising. The retailers’ brick-and-
the customer (Tyrväinen et al., 2020). For many retailers, data sharing mortar business, which is different from retail media networks, also
and loyalty programs are tools to award their best customers with a includes trade dollars that include product placement and slotting fees.
better experience (Leppäniemi et al., 2017). As part of their Many retailers also own other related businesses such as banking, res­
customer-oriented strategy, marketers often tailor their advertising ef­ taurants, gas stations, liquor stores, floral, etc., and run their own loyalty
forts and use personalization as a tool to target the right consumers with cards and programs as well. retail media networks are becoming more
the right message at the right time (Aguirre et al., 2016). Retailers use important to manufacturers because they want to influence brand choice
personalization to address their customers by name and anticipate their closer to the point of sale within the store (Nguyen et al., 2020). This
needs (Shen and Ball, 2009). According to one expert on customer market is attractive to retailers with national manufacturers spending
engagement: more at the retail level (Karray et al., 2022). For example in 2020
advertising from the top 200 companies was estimated to be $165 billion
… the two keywords here are "repeated" and "investment". A simple
US, and Proctor and Gamble spent $4.7 billion on ads (Ad Age, 2021).
focus on maximizing conversions can, in some circumstances,
Examples of retail media networks include Amazon, Walmart, Kroger,
decrease the likelihood of repeat conversions. Think of that pushy
Target, and Ahold Delhaize who have all invested heavily in their retail
door-to-door salesman. You might buy something from him but what
media network (Klecker, 2020). In 2020, Walmart left its partnership
is the likelihood that you will repeat that purchase, or recommend
with the WPP agency to launch its own retail media network called the
his services to a friend?
Walmart Advertising Partners Program which is part of the Walmart
Customer engagement places the strategic emphasis on the creation Media Group (Klecker, 2020). Walmart has focused on growing its retail
of valuable relationships and encourages both parties to see mutual media network by acquiring its retail media network from Polymorph
advantage in that relationship. Frequently this will mean that your Labs in 2019 (Katz, 2019). In 2019, Target rebranded its retail media
tactics require a multi-channel approach incorporating the best of network as Roundel and entered into a partnership with Index Ex­
digital and traditional media. I see the ability to effectively blend change, another advertising exchange to become Roundel/Index Ex­
media in its type, quantity, and timing (known as Right-touching) as change (Katz, 2019). Ahold Delhaize, which ranks fourth among large
one of the key challenges for those interested in engagement in the grocery stores, launched its advertising exchange PeaPod Digital Labs
coming years. —Richard Sedley of cScape (Chaffey, 2007) Media Partnerships in 2018 (Peapod, 2019a, 2019b). They partnered
with Quotient a digital online and promotions company and can provide
Such right-touching (delivering the right message to the right con­
consumers information about the products that they buy that are
sumers at the right time) through personalization not only helps ad­
currently on sale in the stores and make customers aware of these
vertisers reach the right consumers more accurately but also benefits
through their retail media network (Becker, 2021). Kroger had been in a
consumers by presenting products and services that better match their
partnership with DunnhumbyUSA, which provided the retailer with
preferences (Vesanen, 2007). More specifically, right-touching messages
insights about its customers. They acquired this firm in 2015 and
are effective for companies because they support consumers in their
rebranded its agency 84.51◦ (Watkins, 2015). They are now called
purchase process and consumers simultaneously allow the collection of
Kroger Precision Marketing and are in a new partnership with Micro­
information regarding their needs and preferences (Blom et al., 2017). In
soft’s PromoteIQ to track brand ROI from advertising to in-store and
turn, consumers benefit from right-touching personalized ads because of
online sales. Kroger’s Precision Marketing ad revenue was expected to
the increased convenience and reduced cognitive load (Ansari and Mela,
grow by 50% in 2020 (Klecker, 2020).
2003).
Personalized ads are often delivered through programmatic adver­
3. Methodology
tising, which is “computer-based advertising buying using a real-time
bidding process” (Belch and Belch, 2018, p. 339). This allows a
The authors conducted interviews with retail media experts by
marketer to buy advertising using criteria based on budget, the audience
employing the phenomenology method. Interviews were conducted
they would like to reach, and its key performance indicators (Paulson
using this methodology to attain a first-person description of a specified
et al., 2018). Buying advertising programmatically is on the rise: Ac­
domain of experience (Thompson et al., 1989). The format of the in­
cording to eMarketer, three out of every four ad dollars are spent on
terviews follows the process of attaining a dialogue where the inter­
digital ad execution with programmatic digital advertising. By 2020,
viewer guides the discussion in an unobtrusive way, leading to lengthier,
spending on digital advertising is expected to reach $65.63 billion, of
more detailed descriptions of the phenomena being examined (p. 139).
which 86.2% will be through programmatic ad placement (eMarketer,
This approach has been used successfully to study large retailers’ op­
2018a). While both consumers and advertisers benefit from such
erations by interviewing store managers (Gruber et al., 2016). It also is
advertising, this heavy reliance on personalized consumer data has
an effective exploratory method for analyzing digital advertising and
become one of the major concerns, especially in terms of privacy. Under
websites (Bunker et al., 2016). Since this research focuses on retailers

2
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

Table 1 with the focus on the exchanges taking place within the retailer’s tar­
Description of the sample interviewed. geted advertising program. The US interviews were conducted by phone
Interviewee Retail Number of Location Responsibilities with and in person at the retailers’ US headquarters. Pseudonyms were used
Experience Employees Retail Media Networks to protect the privacy of key informants along with using generic terms
(Years) for actual names of businesses.
Wade 8+ 500+ US Led an effort to create Interviews were conducted before and after the Cambridge Analytica
and run retail media event and prior to and after the implementation of the European Union’s
networks for two (EU) General Data Protection Regulation (GDPR). This allowed us to see
separate large retailers.
Responsible for
firsthand how large retailers are addressing these important issues sur­
working with rounding consumer privacy and personalization. Through these discus­
implementation and sions, the author created a Retail Media Networks model for large
design of advertising retailers that allows for both better protection of consumers’ personally
for all stores.
identifiable information (PII) and data monetization through new forms
Responsible for all
third-party of advertising.
relationships.
Nash 20+ 3600+ US Works with dominant 4. Results
advertising platforms of
Google and Facebook
as well as large retailers
4.1. Data privacy and personalization
throughout the US and
EU to provide Prior research focused on the personalization-privacy paradox of
advertising technology data exchange from the perspective of the customer (Pallant et al.,
capabilities to retailers’
2022). Our research focuses on this same exchange but from the
media networks.
Quinn 4+ 300,000+ EU EU Counterpart to perspective of the retailer. Our second research question asks “is it
Wade. Responsible for possible for large retail media networks to protect their consumer data
all advertising within and provide personalization?“, in short the answer is similar to that
the EU for multiple given in the past … it depends, “the result is a personalization-privacy
retail operations and
paradox, such that personalization can both enhance and diminish
GDPR compliance.
consumer engagement with the firm” (Aguirre et al., 2016 p. 98). Our
research adds to this discussion by contributing what we know about
and their use of consumer data, this study also used secondary data from how large retailers are engaging in these exchanges with their customers
retail media. while providing some privacy controls to these customers.
Because retailers care about the relationship that they have with
their customers, we found that retailers’ media networks are providing
3.1. Research approach and sampling data privacy, as well as customer information and controls through
transparency and choice when it comes to customer data. Our model
In-depth semi-structured interviews were conducted with key in­ illustrates how consumer data is exchanged for personalization by large
formants from large retailers in the US, as well as a technology/privacy retailers. Our model also shows how media targeting and personaliza­
expert from a market research firm that works with Google, Facebook, tion work in such exchanges. Retail media networks engage in data
and Pinterest, and a company similar to Amazon in the EU (see Table 1). monetization through a two-step process of data on-boarding, and media
Each interview was conducted to attain first-person viewpoints focusing targeting. Creating retailer media platforms protects consumer data and
on what retail media networks are and how retail media networks provides an opportunity for marketers to engage in retargeting without
operate in terms of their collection, storage, and use of consumer data,

Fig. 1. Retail media networks model.

3
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

Fig. 2a. Retailers evolving exchanges as media companies.

Step 1. Retargeting Expansion through Data Onboarding

disclosing any PII by creating anonymized IDs used for this purpose. The points for suppliers and their agencies to leverage anonymized data of
retailer can then engage in a pull marketing strategy (making a con­ consumer segments made available within the retailer’s media network
sumer actively seek a product) by retargeting the consumer with (see Fig. 1). Retail media networks accomplish this in a variety of ways
personalized targeted ads. These ads are pushed out of the retailers’ using push and pull advertising methods, including using media trading
media platforms via advertising exchanges through the retailers’ desks, data management platforms (DMPs), and data gateways. These
advertising technologies (AdTech). tools are offered to consumers through either a self-serve or managed
During the authors’ interviews, the following Retail Media Networks service model, and in most cases, the data and the interfaces are cloud-
model emerged: based to increase speed and agility. According to the Interactive
As shown in Fig. 1, the first step in creating a retail media network is Advertising Bureau, “DMP is a system that allows the collection of
when a retailer “walls off” first-party shopper data and/or loyalty data audience intelligence by advertisers and ad agencies, thereby allowing
that the retailer collects about its consumers. First-party data is defined better ad targeting in subsequent campaigns” (Riordan-Butterworth,
as “Information collected directly and stored by website publishers, re­ 2012). Within the retail media network, retailers use the DMP to create
tailers and other types of companies about their site visitors or cus­ audience segments and then push those audiences to demand-side
tomers. Because companies with this information have a prior platforms where targeted media opportunities are purchased. Instead
relationship with their customers, they are able to use this first-party of sharing/selling/provisioning the data directly to CPGs, retailers use
data—which may include names, addresses, phone numbers, site- funding from the CPGs to buy media that is enhanced by their
interaction data, and information about products purchased—to closed-loop retail media network ecosystem within their walled garden.
communicate directly with them. First-party data is then what is stored We have adapted the definition of a walled garden similar to Microsoft’s
in customer-relationship-management and loyalty-program databases” patent definition (see Kelly et al., 2005 p. 15) and define it as walling off
(Ad Age, 2013). Restricting access to this data is new in retailing, as the data within one’s network by keeping it secure using levels of privacy
data were previously exchanged more freely between the retailers and and or/security allowing the retailer access to the data but keeping it
their suppliers. By restricting access, retailers are now seeing the private from everyone else. Within the retailer walled garden, these
increased potential to monetize the data using the retailer’s private retail media networks are especially powerful because it allows for
marketplace. anonymized shopper data to be matched or mapped to digital IDs
Data monetization occurs when the retailer creates defined access (cookies, device, social). These two steps are illustrated below using a

Fig. 2b. Retailers evolving exchanges as media companies.

Step 2. Media Targeting - Increased Personalization and Ad Relevance

4
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

consumer example as shown in Figs. 2a and b. revealed or be part of any process. We have eliminated the PII
(personally identifiable information) from our platform and utilized
4.2. Step 1. Data on-boarding a third party that specializes in personal information to create a
unique ID, which is then used for all the transactions. Any research
Sarah, a fictitious shopper, is used to help illustrate how exchanges regarding personal information is always confidential because the PII
on these retail media networks work. By agreeing to use apps such as is not connected to us [the retailer] in any way. This is a better
Angry Birds or Candy Crush, Sarah is allowing her information to be approach to collaborate on without errors or lapses. —Nash
exchanged for the use of the app. She has essentially opted-in, and the
For coupons and promotions used in a loyalty program, it makes
personal information she gives up includes what apps she is using on her
sense that this information would have to be shared with the companies
phone, her location, and her device IDs (Egelman, 2019). Through a
paying for these manufacturers’ rebates and incentives i.e., the supply
process of data on-boarding, the data are both enriched by industry data
chain partners. It seems that transparency in the retailer’s privacy
and anonymized into shared consumer segments where they are kept
disclosure is the main approach to addressing these issues in the US. For
private by the retailer within its retail media platform to enhance the
retailers in the EU, this model is not allowed under GDPR to share data
existing first-party data the retailer already owns using past purchase
with service partners or even between their own companies and the
history. “Data on-boarding, sometimes referred to as data on-ramping, is
parent companies without explicit permission. Even if a customer
the transfer of data gathered offline to the digital realm” (Ad Age, 2014
wanted to opt-in to step 1 of data onboarding it would not be allowed
p. 1). This represents a shift in the marketplace for the retailer, who now
without explicit permission for each sharing point.
has more of a media agency role.
We are only using cookies for our own tools. We are using it for our
The approach that most of us are taking … is kind of a walled-garden
own advertising. We are not able to combine data with a third-party.
approach where we want them to run their media against our cus­
If you want to do that, then it’s an extra tick in the box. According to
tomers and to run it on our platforms, to run it in a way where we can
GDPR, you have to be clear who those third parties are. In our
track the ad impressions so that we can connect those impressions to
country, we have retail chain and an online pure-play and we belong to
store sales and prove efficacy. And, so, the old model would be for us
the same parent company. We are not allowed to share data between
to take our data and rent it out or sell it. So, P&G would say, “Okay.
these two companies. If we were to share data between those two
I’m going to do a media buy with Google and then I’m gonna buy
companies, which are legally third-party to each other, then we
some data over here from Nielsen or IRI that have some retailing data
would have to have the third tick in the box. We must always be clear
in it, and I’m gonna come up with a sophisticated attribution model
who the third parties are. If we share the data further with another
to estimate what sales list was based on my advertising.” And so they
partner, then it’s another tick in the box. We have to define the
kind of do this panel—a panel match type of approach and I think
parties. —Quinn
that’s really what traditional media has been about. You look at
Nielsen Home Scan that looks at different things—that looks at The authors next asked Quinn to explain how the exchange is
media mixed modeling and that sort of thing. And what we’re saying different for consumers in the EU vs. the US and what happens if con­
is that this flips the script a little bit and says we can do it more sumers opt-out of all sharing?
directly. If you take your media dollars and you invest against our
I think in the US, the consumer doesn’t have to agree upon the fact
known customers in a way that we can track exposure, what we’ll do
that you can use my data. Within Europe, the consumer has to give
is that we’ll follow that path all the way through to transaction and
consent, and also, they have every right to know how the data will be
repeat transaction and long-term loyalty using all of the first-party
used, and what will be done with it. For instance, if I collect card data
data that we have within our walled gardens. So I think that’s
from you, I should be able to read in the privacy statement, what you
really the model, like how—like the retailer being more in control of
are doing with that card data. If you don’t like what you see, then you
the media program itself, acting as an agency to P&G, which is a
should have the option of opting-out.
paradigm shift for the retailer. Because we’ve all been monetizing to
a certain extent for a while because there’s some value there, but it’s They will get access to the website, but it will be very basic. If they
been hands-off. The new model is being much more deliberate, much want to order something, they will have to fill out everything manually.
more buttoned-up, and professional in the services we provide to There will be stuff there but, very limited. —Quinn.
advertisers. —Wade
How then is PII data kept from being disclosed during the on- 4.3. Step 2. Media targeting
boarding process, and how are consumers informed of this? These are
important issues that are actually addressed by creating a walled garden The second step is for the retailer to engage in media targeting or
within the retailer’s media network. The authors conducted additional retargeting. This process modifies not only the exchange between the
interviews after the Cambridge Analytica event to see how this may have retailer and its suppliers in terms of trade but also the communication
impacted retailers’ use of data monetization and their efforts to establish exchange between the consumer and the retailer:
privacy by design:
There is pushback when it comes to customer experience there’s kind
We won’t share your data with third parties; however, we would of a clean-store mentality. If you shop at a higher-end grocer, like a
share your data with our business partners, which would help us Wegmans or a Whole Foods, it’s a much cleaner store and you don’t
deliver services to you. For example, our loyalty program data will be have a lot of this ancillary noise that is circling around. But if you go
shared with the supply-chain partners. We won’t send your PII file to to a Walmart or a Dollar General or a Target or someplace else, you
a third party, who can then do whatever they want with it. —Wade do get a little bit more CPGs messaging coming at you … We can’t
put up ads everywhere; I don’t want to turn the store into NASCAR …
The question then becomes, do you want the advertisements to be
And so, a part of the evolution of a retailer’s understanding of their
relevant or not? The universal answer is yes. Tackling the tough part,
media business is to understand the balance between maximizing
how do you make ads more appropriate? We believe in more trans­
media revenue and optimizing the customer experience. And it’s a
parency. It should start with that. Whenever someone engages in a
delicate balance. —Wade
platform, the consumer should know what is being used and what is
being recorded during the usage of that platform. Secondly, people Because large retailers can place ads on their owned and operated
do not want to be identified. Personal information should not be media it is important to note from this comment that large retailers are

5
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

doing more in terms of ad placement on their websites but still keeping

with a clean uncluttered look even with sponsored content banner ads,
skyscraper ads, and sponsored search results on retail landing pages.

4.4. Treat Me like you know me

During the media targeting stage, it is important for the retailer to

find the right balance by providing value in the exchange relationship
Fig. 3. Highest priority types of help for personalization
with the consumer. The retailer wants to avoid reaching the tipping
Number of Respondents = 2585 consumers. Source: Gartner (2018).11
point where the consumer feels that the retailer knows too much about
them or is being creepy:
it also carries higher sales uplift for both the retailer and their CPGs
There’s a balance between relevance and creepiness, and so we talk suppliers who are advertising with the retailer.
about that all the time. Because when we ask customers what they
want, they say, "We want you to act like you know me." … What we 5. Discussion and conclusion
find is that as long as they feel like we’re treating them like an in­
dividual and it’s relevant to them or it delivers offers, then they’re 5.1. Theoretical implications
fine with us using the data to talk to them. —Wade
In this step, the retailer is working with CPGs to provide value to This research provides new insights from the perspective of large
customers, with increasingly relevant personalized offers and pro­ retailers surrounding the personalization privacy exchange (i.e. data
motions based on their past purchase history. monetization) by introducing concepts used by retailers including retail
media networks, their use of walled gardens for how they manage and to
4.5. Opting-out of personalization keep customer data private while also allowing customers to benefit
from personalization. There are several target strategies when it comes
Because retailers want to maintain their relationship with Sarah, to customer data including utilizing demographic data, contextual tar­
they have also built a “do not contact” list into this functionality. If Sarah geting, inferred behavioral targeting, and purchased-based targeting.
no longer wants to be retargeted with personalized ads, she can be Facebook and Google platforms were mentioned previously as doing a
excluded from this type of targeting by opting-out: good job with these first three types of targeting but not with purchase-
based targeting because they don’t control or have access to retail point-
I think it’s really for retailers to understand what their privacy policy of-sale (POS) data. In our model we illustrate how customer data is both
is to be clear to customers on how we use their data and how we kept private and enhanced by the retailer within the walled garden
don’t use their data, and really the approach that all of us use is an through data onboarding as illustrated in Figs. 1 and 2a. This is similar to
"opt-out." And it’s fairly standard practice in the US with advertising, previous research by Aguirre et al., (2016) summarizes how data is
you know, like, “do-not-call” lists and that kind of thing for online enhanced by social advertising sites to provide inferred behavioral data.
advertising. There’s a whole opt-out procedure. Basically, if you opt- Our model then shows that this data is anonymized so customers are not
out it prohibits us from being able to use data to retarget. If you read targeted using 1-to-1 marketing but rather are aggregated into segment
any privacy policy for many retailers they use the exact same lan­ groups with similar product purchasing behaviors using purchase-based
guage and it’s clear on how you can opt-out of a personalized ask. targeting (i.e. targeted advertising). This is similar to what Amazon does
We’ll still send you advertising, it just won’t be data-based or for its personalized ads using past purchase behaviors. Retailers are now
purchase-based (i.e., personalized). —Wade monetizing their customer data by following the Amazon model (i.e.
As mentioned above, customers already have a relationship with the “Amazon effect”) consumer data are being collected, managed, and
retailer and in many cases have opted-in to this type of personalization exchanged in new ways through the use of walled gardens and retail
of ads through loyalty programs. A recent study by Gartner surveyed media networks. Enterprising retailers understand the commercial value
over 2500 consumers and found that over 52% of consumers want of this data and have grown their capabilities to target their customers
personalized communications from brands they have given permission through personalized advertising and growing their retail media net­
to, and only 12% of consumers did not want personalized messages works. Through these efforts, retailers are generating ancillary revenue
(Marder, 2018). The type of personalized help given to consumers also from CPGs through unconventional means, the proceeds of which can
mattered. According to the Gartner survey, the five most preferred types hedge against margin pressure, be invested in e-commerce and omni­
of personalization help are shown as follows: channel capabilities, and fund other business priorities like acquisitions
Through this partnership between the retailer and their suppliers, or wage increases (Nguyen et al., 2020; Yurova et al., 2017) (see Fig. 3).
retail media networks are used to place targeted ads for their customers
and engage in right touching with the customers providing them special 5.2. Personalization and privacy control
offers, discounts, rewards, and promotions for the products that they
desire and are more likely to purchase. At the same time retailers are In building these retail media networks we found that retailers desire
increasing customer choices and are providing their customers the op­ to retain their customers and should follow their customers’ desires for
tion of opting-out of personalized targeting at any time. data protection and privacy. Based on our observations we now sum­
marize and recommend the following three pillars that should be used
This is a win-win-win. The customer only sees ads that are relevant to
by retailers in building retail media networks as shown in Fig. 4.
them. This means that the retailer looks better in the sight of the
First, we recommend that customers be given the choice (essentially
customer because we’re talking to them the way that they want to be
and on or off switch) of either opting-in to loyalty programs as the means
talked to, and the yogurt company wins as well because we’re
for permission to engage in purchased-based targeting of advertising
helping them really narrow down who they’re talking to so they
messages, i.e. personalization, or electing to be added to a do-not-
don’t waste all this money and all these [ad] impressions on people
contact list that should be run prior to any ad execution on the retail
who are never going to buy their product. —Wade
media networks. This will allow for customers to either be rewarded for
According to the retailer, this improves the exchanges between the and consent to permission for the exchange of information for person­
retailer and the consumer due to greater personalization/treatment, and alization and its subsequent discounts/rewards or to opt-out of the

6
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

Fig. 4. Three pillars for retailers building personalization and privacy control.

exchange. If the customer does not wish to participate in targeted There are important practitioner implications to keep in mind for
advertising then they would still see seasonal ads or ads that are CPGs advertisers when choosing to use retail media networks. Retail
contextual, but the ads themselves would no longer be personalized media networks have additional advantages over other advertising
based on customer purchases and insights. Second retailers should stop networks since customers use large retailers as their preferred shopping
sharing customer data other than what is essential to transact business outlet for most CPGs brands. US digital retail media ad spending has
and build a walled garden around their customer data. Keeping this data grown from $6.2 billion in 2016 and is expected to grow to $52.21
private enhances security and builds trust with the retailers’ customers. billion by 2023 (eMarketer, 2018b; eMarketer, 2022). Additionally,
This should follow the established guidelines mentioned previously as unlike other ads that tend to be disruptive to customers, this is not the
set forth by the DAA. Part of this is informing customers what data is case for retail ads, since Customers desire purchase goods on retail
collected and how it will be used by the retailer and their partners. This websites and are seeking product information from the retail media
allows the retailer to offer incentives, customized shopping experiences, sites, so the purchase intent is congruent with the desired ad information
and the ability to reward their best customers for the value that they for the brands that they are shopping for. One recent study by the
bring to the retailer. Third, by acquiring ad agencies and or partnering marketing platform Skai reported that of the 1000 customers in their
with third-party ad technology platforms, large retailers now control survey from Amazon, over 50% of respondents didn’t remember seeing
their own retail media networks enabling them to enhance and any ads in Amazon search results over the last 30 days. This is interesting
personalize customers shopping experiences throughout the customers’ because all of the Amazon search results are personalized targeted ads
shopping journeys. (Barba, 2022). Personalization through purchased based targeting is a
strategic advantage for retailers. Because retailers have first-party data
on what shoppers purchase, retailers can prove efficacy or return on
5.3. Managerial implications investment (ROI) to advertising practitioners through closed-loop ad
reporting because they have access to customers’ online and offline
This research illustrates how the new retail media networks have purchase behavior:
changed the funding model for omnichannel retailing through new
forms of data monetization. It describes what retail media networks are, That’s where retailers come in. They have purchase intent and actual
provides a case illustration for how they work, and estimates the size and purchase data. This is something both Google and Facebook are not
growth of these retail media networks. Large retailers now look more the best at, but which is very useful for targeting and segmentation.
like media agencies through their efforts to use their owned and oper­ These retailers are starting to truly understand the goldmine that
ated media to generate additional revenue from their core business and data is, and are starting to capitalize via their media networks.
audiences through personalized targeted advertising using data mone­ —Nate Shurilla (Lipsman, 2019, p. 6).
tization. Through the process of right touching, retail media networks CPGs no longer must wonder if a consumer saw their ad since re­
are valuable tools for advertisers and retailers. To create these retail tailers can now prove efficacy by providing that information about
media networks, large retailers have both acquired advertising agencies advertising campaign effectiveness in terms of ad exposure and purchase
as well as have partnered with technology ad platform providers such as (i.e. sales lift).
Microsoft, Neilson, iRI, Quotient, CitrusAd, following Amazon’s There are however some disadvantages for retailer media networks
playbook. since they are having to catch up with other media platforms by either
purchasing advertising agencies or by partnering with third-parties
1
because they lack their own in-house data management platforms
Portions ©2018 Gartner, Inc. and/or its affiliates. All rights reserved. No (DMPs) to manage their own first-party data (Lipsman, 2019, p. 8).
further reproduction or distribution permitted without the express written
The authors conclude that much of the content, lower prices, and
consent of Gartner, Inc. Content excerpted from Consumers and Personaliza­
enhanced services of omnichannel retailers are advertising-funded and
tion, written by Andrew Marder, Gartner, Inc.

7
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

Table 2
Size estimate of large retailers’ retail media networks.
Retailer Entity Monetization Scope Est. Size of Media Business 2017–2018 Food Sales US Store
($billion)* Count*

Ahold Peapod Digital Labs Media In-store print ads, data monetization $35 Million 42.3 1960
Partnerships
Albertsons Albertsons Performance Display, Data Targeting, In-Store Ads $50–75 Million 60.2 2330
Media
Amazon Amazon Marketing Services Display, Sponsored Search, Data Targeting $10 Billion Online 14 472
Kroger Kroger Precision Marketing Display, Data Targeting, In-Store Ads $125–150 Million 117.9 3651
Target Roundel/Index Exchange Display, Native, Sponsored Search, Email, Data $200 Million 32.7 1834
Targeting
Walmart Walmart Media Group Display, Native, Sponsored Search, Data Targeting, $750 Million Online, $200 266.2 5834
In-Store Digital & Print Ads Million in-store

Source: Pea Pod Digital Labs,2 Mintel Group Ltd.*

that customers are exchanging their information in return for enhanced the US and EU markets, these are important areas that will continue to
products and services. Today, large retailers are not only fighting for a be a focus by advertising and marketing practitioners for years to come
larger share of consumers’ wallets and sales growth, but they are also (US Senate, 2019).
positioning themselves as media companies generating billions of dol­
lars in new advertising revenue. This market represents about 15.2% or Data availability
$11.35 billion in US Digital Ad spending in 2019 (eMarketer, 2019).
According to estimates in our discussions with large retailers and sec­ Data will be made available on request.
ondary data from Mintel suggest that the current revenues for retailers’
media networks in the US are around $12 billion (see Table 2). References
This growth in retail media networks is expected to continue. In 2021
intentwise reported the following estimated revenues for retail media Ad Age, 2013. What exactly is first-party data? Retrieved from. https://adage.com/artic
le/glossary-data-defined/party-data-defined/245054/.
networks as follows: Kroger between $0.30 - $0.75 billion, Target $0.38 Ad Age, 2014. What Is Data On-Boarding? the Transfer of Offline Data to the Digital
- $0.95 billion, and Walmart $1.34–3.35 billion (Reddy, 2021). A recent Realm. AdAge, p. 1. Retrieved from. https://adage.com/article/glossary-data-defin
report by Goldman Sachs estimates that in the US retailers are positioned ed/data-boarding/293651/.
Ad Age, 2021. AdAge Leading National Advertisers Report, 2021, July. Retrieved from.
to generate $15-$20 bn in retail media revenue from CPGs manufac­ https://adage.com/article/datacenter/ad-age-leading-national-advertisers-2019-i
turers by 2025 (English et al., 2021). With the recent pandemic, this may ndex/2178026.
have been accelerated as CPGs sales generated through e-commerce are Aguirre, E., Roggeveen, A.L., Grewal, D., Wetzels, M., 2016. The personalization-privacy
paradox: implications for new media. J. Consum. Market. 33 (2), 98–110.
expected to rise from 15% to 20% by 2024 (English et al., 2021, p. 3). Allan, D., 2018. California’s New Data Privacy Law Could Begin a Regulatory Disaster.
Online ed. Fortune. Retrieved from. http://fortune.com/2018/10/23/california-dat
5.4. Limitations and future research a-privacy-law-gdpr/.
Ansari, A., Mela, C.F., 2003. E-customization. J. Market. Res. 40 (2), 131–145.
Barba, J., 2022. Survey: How Consumers Feel about Amazon Ads. Path to Purchase IQ.
The authors of this work lay out a research agenda involving factors Retrieved from. https://pathtopurchaseiq.com/survey-how-consumers-feel-about
that have yet to be examined in data-privacy literature. This research is -amazon-ads.
Becker, J., 2021. Quotient Partners with Peapod Digital Labs to Launch Industry-First
limited by being conceptual and exploratory in nature. Although experts Promotion Amplification Tool to Maximize Temporary Price Reductions through
were used to better understand the phenomena of retail media networks, Automated Digital Media ([Press release]).
these findings were not statistically tested, leaving room for interpre­ Belch, G.E., Belch, M.A., 2018. Programmatic: Advertising’s Newer, Better Moustrap - Is
Buying Better with Robots?, Advertising and Promotion: an Integrated Marketing
tation of the results. At the same time, the purpose of this research was to Communications Perspective, eleventh ed. McGraw-Hill Education, New York, NY,
introduce what retail media networks are and discuss how they provide pp. 338–339.
personalized advertising and maintain data privacy. Blom, A., Lange, F., Hess, R.L., 2017. Omnichannel-based promotions’ effects on
purchase behavior and brand image. J. Retailing Consum. Serv. 39, 286–295.
As privacy regulations and networking models continue to change,
https://doi.org/10.1016/j.jretconser.2017.08.008.
future research should investigate these new marketing models of retail Bunker, M.P., Bartholomew, D.E., Rawwas, M.Y.A., 2016. Listening to the consumer’s
media networks and their impact on retailing and relevant policy and voice in website analytics. Int. J. Bus. Inf. Syst. 23 (1), 116–130. https://doi.org/
privacy implications for consumers. Adopting GDPR essentially takes 10.1504/IJBIS.2016.078025.
Chaffey, D., 2007. Customer Engagement Interview with Richard Sedley of cScape. Smart
away the customers’ right to choose personalization. Turning this off Insights.
means you will still see ads, but they may no longer be relevant ads. If Egelman, S., 2019. Ad IDs Behaving Badly. The Appcensus Blog. Retrieved from. https
there is legislation in this area the authors recommend that it be at the ://blog.appcensus.mobi/2019/02/14/ad-ids-behaving-badly/.
eMarketer, 2018a. Programmatic Ad Spending US, 2016-2020, pp. 1–21. Retrieved from.
national level and for fewer calls for states to adopt guidelines similar to www.eMarketer.com.
the EU’s GDPR. One example of this is the California Consumer Privacy eMarketer, 2018b. U.S. Digital Ad Spending by Industry, 2016-2019. Retrieved from.
Act, as it would be hard to create separate standards and operational www.eMarketer.com.
eMarketer, 2019. Digital Ad Spending, by Industry, US. 1. Retrieved from. www.eMar
procedures for each state (Allan, 2018). Additionally, future research on keter.com.
data privacy should come up with more specific terms, as in the case of eMarketer, 2022. U.S. Digital Ad Spending by Industry, 2019-2023. Retrieved from.
the EU’s GDPR, to guide companies and consumers in developing www.eMarketer.com.
English, J., et al., 2021. The Merchant-Media Model: A New Era for Retailers as Ad
personalization and data-privacy strategies, and to guide regulators and Platforms, vol. 39. Goldman Sachs & Co. LLC.
legislators in establishing and enforcing penalties for not adhering to Fromm, J., 2016. The Amazon Effect” and the Future of Retail Competition. Retrieved
regulations. With several data-privacy policies already in place across from. http://www.millennialmarketing.com/2016/11/the-amazon-effect-the-
future-of-retail-competition/.
Gruber, V., Holweg, C., Teller, C., 2016. What a waste! Exploring the human reality of
food waste from the store manager’s perspective. J. Publ. Pol. Market. 35 (1), 3–25.
2
Portions ©2019 Pea Pod Digital Labs. All rights reserved. No further https://doi.org/10.1509/jppm.14.095.
reproduction or distribution permitted without the express written consent of Karray, S., Martín-Herrán, G., Sigué, S.P., 2022. Managing advertising investments in
marketing channels. J. Retailing Consum. Serv. 65, 102852 https://doi.org/
Pea Pod Digital Labs. Content excerpted from “View of the Retail Monetization
10.1016/j.jretconser.2021.102852.
Marketplace,” written by Mark Williamson, Pea Pod Digital Labs.

8
D.E. Bartholomew and M. Williamson Journal of Retailing and Consumer Services 69 (2022) 103119

Katz, J., 2019. Retailers Innovating to Strip Market Share from Google, Facebook, Peapod, 2019b. Peapod Spotlights Retailer Digital Media Platforms ([Press release]).
Amazon. Retrieved from. https://insights.digitalmediasolutions.com/articles/reta Phelps, J., Nowak, G., Ferrell, E., 2000. Privacy concerns and consumer willingness to
iler-ad-networks. provide personal information. J. Publ. Pol. Market. 19 (1), 27–41.
Kelly, Sean U., Cheng, Lili, Ryszard, K., Kott, Richard, Hughes, L., Portnoy, William L., Reddy, S., 2021. Retail Media Estimates. Intentwise. Retrieved from. https://www.linkedi
2005. In: Walled Garden,” United States Patent Office. Microsoft Technology Licensing n.com/posts/sreenathkreddy_retailmedia-amazonadvertising-amazonppc-activi
LLC, US, pp. 1–22. ty-6898647604415082496-_iQm/.
Klecker, A., 2020. Supermarkets Adding ‘Media Company’ to the Résumé. Retrieved Riordan-Butterworth, B., 2012. Data Management Platform. IABWiki, p. 636. Retrieved
from. https://progressivegrocer.com/supermarkets-adding-media-company-resume. from. https://wiki.iab.com/index.php/Data_Management_Platform.
Learmonth, M., 2012. Advertising Becomes Amazon’s Newest Low-Price Weapon like Schumann, J.H., von Wangenheim, F., Groene, N., 2014. Targeted online advertising:
Everything Amazon Does, Advertising Will Make Products Cheaper, vol. 5. AdAge. using reciprocity appeals to increase acceptance among users of free web services.
Leppäniemi, M., Karjaluoto, H., Saarijärvi, H., 2017. Customer perceived value, J. Market. 78 (1), 59–75.
satisfaction, and loyalty: the role of willingness to share information. Int. Rev. Retail Shen, A., Ball, A.D., 2009. Is personalization of services always a good thing? Exploring
Distrib. Consum. Res. 27 (2), 164–188. the role of technology-mediated personalization (TMP) in service relationships.
Lipsman, A., 2019. How Retailers Are Building Digital Ad Businesses, vol. 15. eMarketer. J. Serv. Market. 23 (2), 80–92.
Retrieved from. https://content-na1.emarketer.com/retail-media-networks-2019. Thompson, C.J., Locander, W.B., Pollio, H.R., 1989. Putting consumer experience back
Marder, A., 2018. Consumers and Personalization. Gartner, PowerPoint, vol. 12. CEB into consumer research: the philosophy and method of existential-phenomenology.
Iconoculture. J. Consum. Res. 16 (2), 133–146.
Markos, E., Milne, G.R., Peltier, J.W., 2017. Information sensitivity and willingness to Tyrväinen, O., Karjaluoto, H., Saarijärvi, H., 2020. Personalization and hedonic
provide continua: a comparative privacy study of the United States and Brazil. motivation in creating customer experiences and loyalty in omnichannel retail.
J. Publ. Pol. Market. 36 (1), 79–96. J. Retailing Consum. Serv. 57, 102233 https://doi.org/10.1016/j.
Montgomery, A., Smith, M.D., 2009. Prospects for personalization on the internet. jretconser.2020.102233.
J. Interact. Market. 23 (2), 130–137. US Senate, 2019. How Equifax Neglected Cybersecurity and Suffered a Devastating Data
Nguyen, C., Romaniuk, J., Cohen, J., Faulkner, M., 2020. When retailers and Breach. HSGAC. Retrieved from Washington: http://www.hsgac.senate.gov/
manufacturers advertise together; examining the effect of co-operative advertising download/majority-and-minority-staff-report-how-equifax-neglected-cybersecurity
on ad reach and memorability. J. Retailing Consum. Serv. 55, 102080 https://doi. -and-suffered-a-devestating-data-breach.
org/10.1016/j.jretconser.2020.102080. Vesanen, J., 2007. What is personalization? A conceptual framework. Eur. J. Market. 41
Pallant, J.I., Pallant, J.L., Sands, S.J., Ferraro, C.R., Afifi, E., 2022. When and how (5/6), 409–418. https://doi.org/10.1108/03090560710737534.
consumers are willing to exchange data with retailers: an exploratory segmentation. Watkins, S., 2015. Kroger Buys Big Chunk of DunnhumbyUSA, Changes Name. Cincinnati
J. Retailing Consum. Serv. 64, 102774 https://doi.org/10.1016/j. Business Courier, Electronic.
jretconser.2021.102774. Yurova, Y., Rippé, C.B., Weisfeld-Spolter, S., Sussan, F., Arndt, A., 2017. Not all adaptive
Paulson, C., Luo, L., James, G.M., 2018. Efficient large-scale internet media selection selling to omni-consumers is influential: the moderating effect of product type.
optimization for online display advertising. J. Market. Res. 55 (4), 489–506. J. Retailing Consum. Serv. 34, 271–277. https://doi.org/10.1016/j.
Peapod, 2019a. Peapod’s New Innovation. Lab & Digital Media Platform ([Press jretconser.2016.01.009.
release]).