Smart Shopping Carts to Increase Healthier Food Purchase:

A Conjoint Experiment

Niklas Eriksson

1,2 a

, Asle Fagerstrøm

, Valdimar Sigurðsson

, Nils-Magne Larsen

and Vishnu Menon

Arcada University of Applied Sciences, Jan-Magnus Janssonin aukio 1, 00420 Helsinki, Finland

Kristiania University College, Prinsens gate 7-9, 0152 Oslo, Norway

Reykjavik University, Menntavegur 1, 101 Reykjavík, Iceland

The Arctic University of Norway, Havnegata 5, 9404 Harstad, Norway

Massey University, PO Box 756, Wellington 6140, New Zealand

Keywords: Healthy Food Purchase, Self-Service Technology, Grocery Retailing, Decision Support, Conjoint Experiment.

Abstract: Shopping carts, in general, should be suitable for carrying smart technology in the retail store environment.

Also, a smart shopping cart can present verbal motivating stimuli to increase healthier food purchases. A

conjoint experiment was used to test with a hypothetical purchasing task for young consumers (n=91) the

potential of motivating stimulus on smart shopping carts to influence healthier purchases when buying frozen

pizza. The results show a positive impact for all stimuli stemming from the smart shopping cart, three of

which were health-based. This shows that stimuli revealing dynamic and personalized data through smart

technology in a physical grocery retail setting have the potential to outperform traditional brand statements.

Our conjoint experiment increased young consumers’ likelihood of choosing a healthier frozen pizza. This

result demonstrates that verbal stimuli on smart shopping carts can function as motivating augmentals on

young adult consumers’ healthier food purchases and are in line with the market positioning and customer-

service focus of many retailers and brands today, emphasizing a social marketing standing.

1 INTRODUCTION

Increased sedentary lifestyles and consumption of

unhealthy foods have caused an international

epidemic of potential health issues related to being

overweight or obese (Dobbs et al., 2014).

Cardiovascular disease, diabetes, osteoarthritis, sleep

apnea, and even some cancers are just a few of the

potential side effects of living an unhealthy life

(Wyatt et al., 2006). With the estimated annual cost

of overweight and obesity at two trillion USD, steps

must be taken to combat this downward spiral (Dobbs

et al., 2014). Recently, we have seen increased

prominence of studies promoting healthy foods,

particularly in the retail setting. For example, research

demonstrates that the majority of consumers’ food

https://orcid.org/0000-0001-9295-7114

https://orcid.org/0000-0002-8854-1658

https://orcid.org/0000-0002-2420-4863

https://orcid.org/0000-0001-7671-0250

https://orcid.org/0000-0002-0835-5338

purchases are unplanned and contingent upon stimuli

in the retail environment (Inman et al., 2009). The

physical retail environment is, therefore, an important

scene in which to study the effects of healthy food

promotion and how to influence healthier food

purchases most effectively. Retailers can and do play

a large part in influencing their customers’ purchasing

decisions (Larsen et al., 2017). Some of the main

factors that have been found to determine whether

healthy or unhealthy food is purchased include the

location of these foods within stores (Areni et al.,

1999; Sigurdsson & Engilbertsson, 2015), the

availability of healthy and unhealthy foods (Yeh et

al., 2008), access to accurate nutrition information

(Achabal et al., 1987; Dubbert et al., 1984), and price

(Sigurdsson et al., 2011; 2014).

Eriksson, N., Fagerstrøm, A., Sigurðsson, V., Larsen, N. and Menon, V.

Smart Shopping Carts to Increase Healthier Food Purchase: A Conjoint Experiment.

DOI: 10.5220/0011619100003476

In Proceedings of the 9th International Conference on Information and Communication Technologies for Ageing Well and e-Health (ICT4AWE 2023), pages 93-101

ISBN: 978-989-758-645-3; ISSN: 2184-4984

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

Despite the prominent role of carrying equipment

in retail, such as a basket and carts, there is, to the best

of our knowledge, limited literature involving such

equipment in grocery retailing and its potential to

promote healthier food purchases (Larsen &

Sigurdsson, 2019). Currently, there is also

surprisingly limited knowledge of consumer behavior

connected to the relationship between self-service

digital assistants and consumer food purchases. New

technology available in grocery stores may provide

useful mechanisms for promoting healthier food

purchases based on consumer activity at the point of

purchase (Nikolova & Inman, 2015).

Self-service technologies combined with existing

in-store equipment such as smart shopping carts (carts

with digital screens connected to the Internet) are

emerging as prominent options to influence buying

behavior in general and also to promote healthier food

purchases. Shopping carts have several strengths as

an object for smart self-service technology. They tend

to be in close vicinity to the shopper throughout the

whole shopping trip and thus can increase the reach,

frequency, and relevance of real-time personal

consumer-oriented stimuli (Inman & Nikolova,

2017). While smartphones have several potentials,

phones still have a small screen and can be left in the

pocket or purse to a large extent. Screens mounted on

carts can be larger than smartphone screens and

should, therefore, be much more in front of the

shopper. Furthermore, smart shelves and digital

signage have a weakness in their static positioning

within the store. Since a cart’s primary function is to

help consumers carry products, it provides

opportunities to identify consumers’ interests

instantly. Despite that some of the first use of smart

shopping cart at the beginning of the 1990s was not

that successful (Inman & Nikolova, 2017), we see

that new technology has been developed, and todays,

smart shopping cart has been significantly improved

(e.g., Bello-Salau, et al., 2021; Shahroz, et al., 2020).

A few studies have investigated the relationship

between technology-based solutions and healthier

food purchases. Reitberger et al. (2014) concluded

that the combination of Internet-of-Things (IoT) and

mobile devices is a promising approach toward better

(i.e., healthier) consumer food purchases inside

stores. Many consumers lack service in stores, and

consequently, self-service technologies can

contribute to the shopping experience (Kallweit et al.,

2014). Also, younger adult consumers expect smart

technologies to enable them to make more informed

purchases (Priporas et al., 2017). However, Kallweit

et al. (2014; see also Fagerstrøm et al., 2020)

highlight that the technology itself only barely

mediates users’ intention to use self-service

technology in retail; rather, it is about what kind of

service quality, such as information quality, the

technology can provide to the user that matters. Self-

service technologies such as mobile devices, smart

shopping carts, and information kiosks can contribute

to smart retail settings by creating additional value for

customers and retailers. For example, connecting

sensors such as location-based beacon technology to

self-service devices enables retailers to interact

directly with customers as they enter the store and to

push content such as product information, price, and

nutrition stimuli in real-time.

Based on the above discussion, understanding

consumer interaction with smart self-service

technologies in the retail grocery situation, especially

regarding healthier purchases, would be of great

interest to researchers and practitioners. The goal of

this study is, therefore, to investigate the relative

impact of three motivating stimuli on a smart

shopping cart for healthier purchases in grocery retail:

(1) nutritional stimulus based on a health index, (2)

personalized health score based on products in the

shopping cart, and (3) product popularity based on

popular healthy purchases of the week.

2 THEORETICAL FRAMEWORK

Consumer behavior is, to a large extent, regulated by

verbal stimuli in the form of speaking, writing,

signing, and other forms of verbal behavior (Pierce &

Cheney, 2013), such as advice, promises, laws, and

instructions (Foxall, 2010). Rule-governed behavior

is, according to Pierce and Cheney (2013), a term that

is used when the behavior is regulated by the

contingencies that the rule describes. For example, if

a consumer’s purchase is regulated by advice about

buying and consuming more fatty fish because it is

good for the heart and blood vessels, the behavior is

rule-governed.

The current study contributes to the literature on

rule-governed behavior by studying new stimuli

stemming from technology designed to have

motivating functions. The first is a verbal nutritional

stimulus stemming from a smart shopping cart where

we examine real-time stimuli stating a specific

product’s health ranking compared to other products

in the category. For example, a verbal stimulus on a

smart shopping cart screen can state that “A real-time

health comparison index identifies this product as one

of the most nutritious products related to calories and

salt.” A real-time health index stimulus ought to

increase consumers’ likelihood to purchase as it

ICT4AWE 2023 - 9th International Conference on Information and Communication Technologies for Ageing Well and e-Health

increases the reinforcing value in line with functional

consequences (Hayes et al., 2001).

The second augmental is a verbal personalized

health score stimuli on a smart shopping cart. The

verbal health score stimulus gives the consumer an

indication of the total nutrition of the products

purchased by the consumer. For example, when a new

product is added to the smart shopping cart, an

indication can be given that “Based on products in

your shopping cart, this frozen pizza is indicated as a

healthy purchase!” In this case, smart technology can

increase the functional reinforcing value already

attached to the product. Therefore, the personalized

stimulus on the shopping cart screen can be

categorized as a motivating augmental (Hayes et al.,

2001), and thus, it ought to have a positive effect on

the consumer’s likelihood to purchase.

The third augmental is a verbal stimulus

informing the consumer about the most popular

healthy product this week based on real-time

customer purchases. Other consumers’ actions, such

as popularity cues, can signal product quality

(Cheung et al., 2014) and can similarly be used to

signal healthier purchases. Research has found

popularity to make the customer more likely to buy

the product (Castro et al., 2013) and to increase

consumers’ willingness to pay more for the product

(Carare, 2012). For example, a verbal stimulus on a

smart shopping cart screen can state: “Real-time

product popularity: Based on real-time customer

purchase, this is the most popular healthy product this

week.” As for the previous two verbal stimuli, this

can increase the functional reinforcing value already

attached to the product. Therefore, a real-time product

popularity score stimulus on the shopping cart screen

can be categorized as a motivating augmental, and it

most probably has a positive effect on the consumer’s

likelihood to purchase.

3 METHOD

A survey based conjoint experiment was chosen for

the purpose of this study. Conjoint analysis is a hybrid

type of multivariate technique to understand

consumer preferences toward products and services,

and it is considered a realistic method to portray

consumer decisions (Hair et al., 2010).

3.1 Participants

A university student population was chosen as young

adults are interesting subjects to explore. They can be

reasonably considered heavy users of new technology

and market movers, paving the way for new types of

behaviors in retail. Also, overweight and obesity are

growing most rapidly in young adults (Katzmarzyk et

al., 2019), and they consume, to a great extent, ready

meals such as frozen pizza. A student sample was also

chosen due to limited resources to otherwise obtain a

fair number of participants. Students are less

demanding to recruit than external participants,

especially in an experimental setting taking place

physically at a university campus (as the present

study did). By conducting the study in a controlled

physical environment at the campus, we sought to

instead decrease possible disturbing effects of the

experimental setting. Further, students are quite

homogenous in Norway in regard to demographics

(age, socio-economic status etc.), which also ought to

reduce disturbing background effects.

The sample comprised 91 (34 men and 57

women) Norwegian undergraduate students from

Kristiania University College (Oslo) and the

University of Oslo. The sample is slightly skewed

toward females, and it profiles a relatively young

adult consumer group. The participants’ ages were

measured in three categories (18–22, 23-–30, and 31–

45), of which 56 were from the 18–22 category and

32 from the 23–30 category. Out of the 89

participants who answered the question about their

previous use of smart carts, only eight had used smart

carts previously. Participants’ limited use of smart

shopping carts was expected as these technologies are

still scant.

3.2 Design

The target product for the study was frozen pizza. The

food industry sees the health trend and focuses more

on healthier nutrition, such as fewer calories, less salt,

and more natural ingredients, when developing new

frozen pizzas. It is also reasonable to assume that a

frozen pizza is perceived as unhealthy or ‘junk food’

(Combet et al., 2014), and thus, the effect of healthier

options on the likelihood to purchase ought to be

high. This makes it an interesting product to study.

A conjoint experiment where all participants

received the same hypothetical shopping task and the

same varied intervention stimuli to evaluate on a

questionnaire was set up for the purpose of this study.

Data were collected in two separate physical sessions,

but the procedure was the same for all participants.

This type of conjoint experiment with a within-

subjects design and a survey helps to determine how

participants evaluate different predetermined

attributes related to the research object. Here, the

attributes were hypothetical verbal stimuli on a

Smart Shopping Carts to Increase Healthier Food Purchase: A Conjoint Experiment

simulated smart self-service shopping cart, and the

participants were asked to evaluate how likely they

were to purchase the frozen pizza presented to them.

Each attribute is specified by levels, representing

realistic features of each attribute.

Six attributes of verbal stimuli (of which three

were health-based) and their corresponding levels

were identified for the study (see Table 1).

“Nutritional stimulus,” “Healthy choice—shopping

cart,” “Healthy choice—popularity,” “Price levels,”

and “Taste” were operationalized at three levels.

“Price types” was operationalized at four levels. The

different levels of attributes are assumed to have a

varying impact on purchase behavior. The attributes

“Price types” and “Taste” were also pictured to

represent technology-based stimuli, where “Price

types” represented dynamic pricing, and “Taste”

represented statements on product taste, including

customer reviews of frozen pizza.

The “Nutritional stimulus” was pictured with a

real-time product health comparison index of calories

and salt as one level, a brand nutrition statement as a

second level, and no information regarding nutrition

as a third level. “Healthy choice—shopping cart” was

pictured with a personalized health score based on

products in the shopping cart as one level, a brand

statement of healthy choice as a second level, and no

information on healthy choice as a third level.

“Healthy choice—popularity” was pictured with real-

time popularity score as one level, a store statement

regarding product popularity as a second level, and no

information on popularity as a third level. The “Price

levels” were based on price searches from a

Norwegian online grocery retailer (www.kolonial.

no). The average price for frozen pizza was NOK

56.50, with the highest at NOK 83.80 and the lowest

at NOK 37.50. The four “Price types” with dynamic

pricing levels and the three levels of “Taste” were

based on studies conducted by Haws and Bearden

(2006) and Mudambi and Schuff (2010), respectively,

but adapted to fit the research context.

We used IBM SPSS™ for the design and analysis

of the study. Six attributes and their corresponding

levels would total 972 smart cart frozen pizza

configurations (3 x 3 x 3 x 3 x 4 x 3) based on the full

factorial design. Using the fractional factorial design

in SPSS, the number of configurations (i.e., cards)

was reduced to 29 (including four holdout cards). The

four holdout cards are used to validate the data and

expose possible errors in the model. A main-effects

model was chosen to design the study as it measures

the direct impact of each attribute. The main-effects

model assumes that the participant gets a total value

of the combination of stimuli by adding up the value

of each stimulus (Hair et al., 2010). Under this

method, the participants evaluated a set of

experimentally varied stimuli, the 29 configurations

(referred to here as stimulus cards). A full-profile

method was chosen to collect the data, where each

stimulus card was presented separately to the

participants. While analyzing the data, a discrete

measurement was used for all six stimuli.

Holbrook and Moore (1981) suggest using visual

stimulus cards to present the stimulus. The visual

stimulus cards were administered in a classroom

using a Microsoft PowerPoint™ presentation and a

pen-and-paper questionnaire for the participants. An

illustration of stimulus cards and questions is

presented in the Appendix.

3.3 Procedure

When the participants had voluntarily accepted to

participate in the study, they were presented with a

hypothetical shopping task. In the task, they were to

assume that they were going to purchase a frozen

pizza in the grocery store:

Assume that you are going to buy some groceries

for dinner, and you are now standing in the store. The

retail store you are regularly visiting has

implemented smart carts, as you can see in the

picture. The smart cart holds the shopping list that

you made and uploaded last night. The smart cart

also makes it possible to see the products you have

already picked in your shopping cart. You are now in

the selection process of frozen pizza, and the smart

cart screen gives you product information. Based on

previous experience, you know that the average price

of frozen pizza is about NOK 56.50. You will now be

presented 29 shopping situations. Evaluate the 29

shopping situations in relation to using the smart cart

when purchasing the frozen pizza.

Before the data collection started, an example

stimulus card was presented to the participants to

familiarize them with the procedure. Once all 29

stimulus cards were evaluated, the participants were

asked to provide some background information.

4 RESULTS

Based on the analysis, we evaluated the goodness-of-

fit of the conjoint model. We found that the

correlations between the actual and estimated

preferences are significant (Pearson’s R = 0.982, p <

0.001 and Kendall's tau = 0.873, p < 0.001). The

Kendall’s tau for the holdout cards is fair (0.667) but

ICT4AWE 2023 - 9th International Conference on Information and Communication Technologies for Ageing Well and e-Health

not significant (p = 0.087). Based on this, we can say

that the conjoint model has acceptable accuracy.

The results are presented in Table 1. The constant

is 5.087, and the impact estimate values of the levels

vary both negatively and positively with this value.

The importance values show notably that the stimuli

“Price levels” is evaluated as the most important

predictor of purchase for the frozen pizza.

“Nutritional stimulus” is evaluated second,

“Customer reviews taste” third, followed closely by

“Price types,” “Healthy choice—shopping cart,” and

“Healthy choice—popularity.” When taking a closer

look at the primary stimuli under investigation here,

“Nutritional stimulus,” “Healthy choice—shopping

cart,” and “Healthy choice—popularity,” we can see

that the impact estimates for Real-time nutrition

stimulus, Personalized health score, and Real-time

product popularity are positive and notably higher

than the alternative levels for each stimulus. Hence,

this type of stimulus increases the likelihood to

purchase. The No information level scores relatively

high negative impact estimates for all three stimuli. In

other words, providing no information regarding

“Nutritional stimulus,” “Healthy choice—shopping

cart,” and “Healthy choice—popularity” decreases

the likelihood to purchase. It is also worth noting that

“Price types” regarding fixed and dynamic pricing is

not clear-cut, as the impact estimates do not show a

clear positive or negative pattern. Then again,

customer reviews regarding taste score a relatively

high positive impact estimate; likewise, a below-

average market price positively impacts purchasing

behavior.

In further analysis, we conducted a simulation of

three scenarios regarding the “Nutritional stimulus,”

“Healthy choice—shopping cart,” and “Healthy

choice—popularity.” See Table 2. A conjoint

simulation strives to understand how the participants

would choose between different scenarios, including

a specific set of stimuli (Hair et al., 2010). Here we

wanted to better understand technology-based stimuli

in comparison to only traditional brand or store-based

stimuli and no information regarding healthy

purchases. Hence, in the first scenario (A), we set all

three health stimuli to simulate technology-related

stimuli; in the second scenario (B), we set all three

stimuli to simulate traditional brand or store

statements regarding the healthy purchase, and

finally, in the third scenario (C), we set all three

variables to simulate no information regarding

healthy purchase. We set “Price levels,” “Price

types”, and “Taste” to simulate a typical shopping

situation in which the price is fixed at an average

market level, and the brand provides a statement

regarding product taste. By conducting this

simulation, we gain insights into the predicted

preference proportions of the three scenarios. The

outcomes for each scenario case are shown according

to preference scores along with a preference

probability score, Logit (0-100%). Logit is an optimal

measurement for repetitive purchase situations (Hair

et al., 2010), which is typical for grocery shopping.

The outcome results for scenario A, 56.3% (Logit),

show that a very large proportion of the respondents

would base their purchase on real-time or

personalized health scores if they were provided with

such stimuli on a smart device like a smart shopping

cart. The Logit score drops to 25.6% for scenario B

and 18.1% for scenario C. In practice this means that

scenario A is clearly more preferred than scenario B

and C, while scenario B is slightly more preferred

than scenario C.

5 DISCUSSION

The conjoint analysis shows that, relative to the other

stimuli except for a below-average price, the three

technology-based health dimensions are important

stimuli when purchasing a frozen pizza; they notably

increase the likelihood of buying the product. No

information stimulus scored strong negative results,

indicating that leaving out healthy purchase stimuli

decreases the likelihood of buying. However, this is

not surprising, given that the product under

investigation is frozen pizza. As discussed, it is

reasonable to assume that a frozen pizza is likely to

be perceived as unhealthy, and thus motivating

stimuli indicating it is a healthy purchase have a

positive impact on the likelihood of buying.

Nevertheless, the simulation scores from the conjoint

analysis indicate a high preference for the three

technology-based health dimensions relative to

conventional brand statements and no healthy

purchase stimuli. This result is interesting as it

indicates that smart technology-based motivating

stimuli have the possibility to outperform

conventional brand or store statements regarding

healthier food purchases.

Overall, price level scored the highest

importance, showing that price is an important

attribute to increasing healthier food purchases. This

result is in line with previous findings (Sigurdsson et

al., 2011; 2014). However, it is reasonable to assume

that the participants in this study are quite price-

sensitive in general, as we studied undergraduate

students. Nutritional stimulus scored the second-

highest

importance after price level, which also

Smart Shopping Carts to Increase Healthier Food Purchase: A Conjoint Experiment

Table 1: Conjoint utilities and importance values.

Attributes Levels

Impact

estimate

Standard

erro

Importance

values

Nutritional

stimulus

1. Real-time nutrition: A real-time health comparison index identifies

this product as one of the most nutritious frozen pizzas related to

calories and salt—Find out more.

2. Brand statement: Fewer calories and less salt!

3. No health information

0.471

−0.146

−0.325

0.098

0.117

16.646

Healthy

choice—

shopping

cart

1. Personalized health score: Based on the nutrition content of the

products in your shopping cart, this frozen pizza is indicated as a

healthy choice—Find out more.

2. Brand statement: This is a healthier choice!

3. No health information

0.311

−0.089

−0.221

0.098

0.117

11.825

Healthy

choice—

popularity

1. Real-time product popularity: Based on real-time customer choice,

this is the most popular healthy product this week—Find out more.

2. Store statement: The store states that this is a popular healthy

product! —Find out more

3. No health information

0.223

0.056

−0.280

0.098

0.117

11.754

Price

levels

1. Below-average market price: Price NOK 37.50

2. Average market price: Price NOK 56.50

3. Above-avera

e market

rice: Price NOK 83.80

1.294

0.291

−1.585

0.098

0.117

33.376

Price

types

1. Fixed price: Price NOK xx

2. Dynamic price: Price NOK xx,

based on a national index updated

every month

3. Dynamic price: Price NOK xx,

based on a national index updated

every week

4. Dynamic price: Price NOK xx,

based on a national index updated

every hour EUR 100 per night

−0.168

0.058

−0.036

0.146

0.107

0.133

12.200

Taste

1. Customer review: Customer reviews on taste: 4.9 out of 5 stars

2. Brand statement: Supreme taste!

3. No information

0.503

−0.060

−0.443

0.098

0.117

14.199

(

Constant

)

5.087 0.090

Price was indicated b

rice levels in the con

oint

lan.

Table 2: Scenario simulation.

Stimuli and levels Outcomes

Scenarios Cases

Price

level

Price

types

Taste

Nutrition

stimulus

Healthy

choice—

shopping cart

Healthy

choice—

opularit

Pref.

scores

Logit

Health

scores

A Average Fixed

Brand

Statement

Real-time Personalized Real-time 6.156 56.3

Health

brand

statements

B Average Fixed

Brand

Statement

Brand

Statement

Brand

Statement

Brand

Statement

4.972 25.6

No health

info

C Average Fixed

Brand

Statement

Blank –

no info

Blank – no

info

Blank – no

info

4.324 18.1

a. 84 out of 88 sub

ects are used in the Lo

it method because these sub

ects have all non-ne

ative scores.

indicates that this type of health stimulus is perceived

as important. Accurate nutrition information has been

found to affect healthier food purchases positively

(Combet et al., 2014).

Grocery retailers are important shapers of stimuli

that influence consumers purchasing behavior

(Martinez et al., 2018). Based on the results from this

study, retailers ought to be able to impact healthier

food purchases positively by providing young adult

customers with self-service technological solutions

that include technology-based health motivating

stimuli like those used in this study. These types of

digital solutions may particularly benefit retailers and

brands who want to stand out as responsible actors in

healthy purchases. Policymakers and initiators of

healthy food consumption should notice these results

ICT4AWE 2023 - 9th International Conference on Information and Communication Technologies for Ageing Well and e-Health

as well. For example, subsidizing or providing

incentives to retailers regarding technological

solutions for healthier food purchases may push or

provide retailers with the capabilities to innovate and

adopt new digital solutions. Even small interventions

regarding consumer behavior could add up to

significant long-term health effects in society

(Wansink, 2016). Nevertheless, it should be noted

that the price level was the main influencer on the

likelihood to purchase; thus, retailers and brands need

also to consider price reductions if they want to

increase the purchase of healthier foods, at least

among younger and price-sensitive customers.

Self-service technological solutions such as a

smart shopping cart can be a vehicle for promoting

healthier food instead of being used to increase

buying behavior in general or worse, to promote

unhealthy options. Furthermore, Larsen et al. (2020)

have shown that most consumers go into a grocery

store today without a shopping cart, especially young

consumers often buying unhealthy food such as pizza.

Digital solutions can be one of the retailer’s solutions

to this problem by making the shopping cart more

attractive (see Arbore et al., 2014), especially for the

young consumer segment. Larsen et al. (2020) show

that younger consumers are underrepresented among

those using the traditional “non-technology”

shopping cart. Carts with attractive digital solutions

might help retailers increase the share of younger

consumers using a cart, which may increase store

experiences as well as store sales. Therefore, there is

a need for further research on how technology

solutions, for example, personal health advice, can

increase the likelihood of using a cart in grocery

shopping.

5.1 Limitations and Further Research

There are some limitations related to data collection

and interpreting the results in this study. Firstly, the

study’s reliance on a somewhat narrow undergraduate

student sample may influence the impact of verbal

health and price stimuli on purchases. Further studies

could, therefore, replicate this study with a wider

sample profile. A second limitation might be the order

effect, as the order of stimuli was presented

sequentially (Chrazan, 1994). Therefore, stimulus

cards could be randomized in similar future studies.

Finally, the studied hypothetical task was designed

for frozen pizza and six pre-defined stimuli variables.

The experimental design, in conjoint, calls for

execution assumptions and delimitations made by the

researcher (Hair et al., 2010). Hence, future studies

could be conducted in an in-store setting with a real

smart cart solution and use other types of products

and additional or different types of stimuli relevant in

a grocery shopping situation. Conjoint analysis

should be viewed as primarily explorative, although

it is regarded as a realistic way to capture consumers’

preferences (Hair et al., 2010).

REFERENCES

Achabal, D. D., McIntyre, S. H., Bell, C. H., & Tucker, N.

(1987). The effect of nutrition POP signs on consumer

attitudes and behavior. Journal of Retailing, 63(1), 9–

Arbore, A., Soscia, I., & Bagozzi, R. P. (2014). The role of

signaling identity in the adoption of personal

technologies. Journal of the Association for

Information Systems, 15(2), 1.

Areni, C. S., Duhan, D. F., & Kiecker, P. (1999). Point-of-

purchase displays, product organization, and brand

purchase likelihoods. Journal of the Academy of

Marketing Science, 27(4), 428-441.

Bello-Salau, H., Onumanyi, A. J., Michael, D., Isa, R.,

Alenoghena, C. O., & Ohize, H. (2021). A new

automated smart cart system for modern shopping

centres. Bulletin of Electrical Engineering and

Informatics, 10(4), 2028-2036.

Carare, O. (2012). The impact of bestseller rank on demand:

Evidence from the app market. International Economic

Review, 52, 717-742. doi: 10.1111/j.1468-

2354.2012.00698.x

Castro, I., Morales, A., & Nowlis, S. (2013). The influence

of disorganized shelf displays and limited product

quantity on consumer purchase. Journal of Marketing,

77, 118. doi:10.1509/jm.11.0495

Cheung, C. Xiao, B., & Liu, I. (2014). Do actions speak

louder than voices? The signaling role of social

information cues in influencing consumer purchase

decisions. Decision Support Systems, 65, 50-58. doi:

10.1016/j.dss.2014.05.002

Chrazan, K. (1994). Three kinds of order effects in choice-

based conjoint analysis. Marketing Letters, 5(2), 165–

172.

Combet, E., Jarlot, A., Aidoo, K. E., & Lean, M. E. (2014).

Development of a nutritionally balanced pizza as a

functional meal designed to meet published dietary

guidelines. Public Health Nutrition, 17(11), 2577-2586.

Dobbs, R., Sawers, C., Thompson, F., Manyika, J.,

Woetzel, J., Child, P., . . . Spatharou, A. (2014). How

the world could better fight obesity. McKinsey Global

Institute.

Dubbert, P. M., Johnson, W. G., Schlundt, D. G., &

Montague, N. W. (1984). The influence of caloric

information on cafeteria food choices. Journal of

Applied Behavior Analysis, 17(1), 85-92.

Fagerstrøm, A., Eriksson, N., & Sigurdsson, V. (2020).

Investigating the impact of Internet of Things services

from a smartphone app on grocery shopping. Journal of

Smart Shopping Carts to Increase Healthier Food Purchase: A Conjoint Experiment

Retailing and Consumer Services, 52, Article 101927.

doi: 10.1016/j.jretconser.2019.101927

Foxall, G. (2010). Interpreting consumer choice: The

behavioural perspective model (Vol. 10). New York:

Routledge.

Hair, J., Black, W., Babin, B., & Anderson, R. (2010).

Multivariate Data Analysis. Upper Saddle River, NJ:

Pearson Prentice Hall.

Hayes, S. C., Barnes-Holmes, D., & Roche, B. (2001).

Relational frame theory. A post-Skinnerian account of

human language and cognition. New York: Kluwer

Academic/Plenum Press.

Haws, K. L., & Bearden, W. O. (2006). Dynamic pricing

and consumer fairness perceptions. Journal of

Consumer Research, 33(3), 304-311. doi:10.1086/

508435

Holbrook, M. B., & Moore, W. L. (1981). Feature

interactions in consumer judgments of verbal versus

pictorial presentations. Journal of Consumer Research,

8(1), 103-113.

Inman, J. J., & Nikolova, H. (2017). Shopper-facing retail

technology: A retailer adoption decision framework

incorporating shopper attitudes and privacy concerns.

Journal of retailing, 93(1), 7-28. doi:

10.1016/j.jretai.2016.12.006

Inman, J.J., Winer, R.S., & Ferraro, R. (2009). The

interplay among category characteristics, customer

characteristics, and customer activities on in-store

decision making. Journal of Marketing, 73, 19-29. doi:

10.1509/jmkg.73.5.19

Kallweit, K. Spreer, P., & Toporowski, W. (2014). Why do

customers use self-service information technologies in

retail? The mediating effect of perceived service

quality, Journal of Retailing and Consumer Services,

21(3), 268-276. doi: 0.1016/j.jretconser.2014.02.002

Katzmarzyk, P. T., Chaput, J.-P., Fogelholm, M., Hu, G.,

Maher, C., Maia, J., . . . Tremblay, M. S. (2019).

International Study of Childhood Obesity, Lifestyle and

the Environment (ISCOLE): contributions to

understanding the global obesity epidemic. Nutrients,

11(4), 848.

Larsen, N. M., Sigurdsson, V., Breivik, J., & Orquin, J. L.

(2020). The heterogeneity of shoppers’ supermarket

behaviors based on the use of carrying equipment.

Journal of Business Research, 108, 390-400.

Larsen, N. M. & Sigurdsson, V. (2019). What affects

shopper’s choices of carrying devices in grocery

retailing and what difference does it make? A literature

review and conceptual model. The International Review

of Retail, Distribution Consumer Research, 1-33. doi:

10.1080/

09593969.2019.1581074

Larsen, N. M., Sigurdsson, V., & Breivik, J. (2017). The

use of observational technology to study in-store

behavior: consumer choice, video surveillance, and

retail analytics. The Behavior Analyst, 40(2), 343-

371. doi: 10.1007/s40614-017-0121-x

Martinez, O., Rodriguez, N., Mercurio, A., Bragg, M., &

Elbel, B. (2018). Supermarket retailers’ perspectives on

healthy food retail strategies: In-depth interviews, BMC

Public Health, 18(1019), 1-16.

Mudambi, S. M., & Schuff, D. (2010). Research Note:

What makes a helpful online review? A study of

customer reviews on Amazon.com. MIS Quarterly,

34(1), 185-200. doi:10.2307/20721420

Nikolova, H. D., & Inman, J. J. (2015). Healthy purchase:

the effect of simplified point-of-sale nutritional

information on consumer food purchase

behavior. Journal of Marketing Research, 52(6), 817-

835. doi: 10.1509/jmr.13.02

Pierce, W. D., & Cheney, C. D. (2013). Behavior Analysis

and Learning (5 Ed.). New York: Psychology Press.

Priporas, C-V., Stylos, N., & Fotiadis, A. K. (2017).

Generation Z consumers' expectations of interactions in

smart retailing: A future agenda. Computers in Human

Behavior, 77, 374-381.

Reitberger, W. Spreicer, W., & Fitzpatrick, G. (2014).

Situated and mobile displays for reflection on shopping

and nutritional choices. Personal and Ubiquitous

Computing, 18(7), 1721–1735.

Shahroz, M., Mushtaq, M. F., Ahmad, M., Ullah, S.,

Mehmood, A., & Choi, G. S. (2020). IoT-based smart

shopping cart using radio frequency identification.

IEEE Access, 8, 68426-68438.

Sigurdsson, V., & Engilbertsson, H. (2015). Vertical

allocation of brands in retail shelf-space and its effect

up on sales. In Marketing in Transition: Scarcity,

Globalism & Sustainability (pp. 73-73): Springer.

Sigurdsson, V., Larsen, N. M., & Gunnarsson, D. (2011).

An in-store experimental analysis of consumers'

selection of fruits and vegetables. The Service

Industries Journal, 31(15), 2587-2602.

Sigurdsson, V., Larsen, N. M., & Gunnarsson, D. (2014).

Healthy food products at the point of purchase: An in‐

store experimental analysis. Journal of Applied

Behavior Analysis, 47(1), 151-154.

Wansink, B. (2016). Slim by design: Mindless eating

solutions for everyday life. London: Hay House UK.

Wells, V. K. (2014). Behavioural psychology, marketing

and consumer behaviour: A literature review and future

research agenda. Journal of Marketing Management,

30(11-12), 1119–1158. doi: 10.1080/

0267257x.2014.929161

Wyatt, S. B., Winters, K. P., & Dubbert, P. M. (2006).

Overweight and obesity: prevalence, consequences, and

causes of a growing public health problem. The

American Journal of the Medical Sciences, 331(4), 166-

174.

Yeh, M.-C., Ickes, S. B., Lowenstein, L. M., Shuval, K.,

Ammerman, A. S., Farris, R., & Katz, D. L. (2008).

Understanding barriers and facilitators of fruit and

vegetable consumption among a diverse multi-ethnic

population in the USA. Health Promotion

International, 23

(1), 42-51.

ICT4AWE 2023 - 9th International Conference on Information and Communication Technologies for Ageing Well and e-Health

100

APPENDIX

Stimulus card 8.

Based on the information from the smart cart, how

likely is it that you would purchase the frozen pizza?

Not at all likely Certainly likely

0 1 2 3 4 5 6 7 8 9 10

Smart Shopping Carts to Increase Healthier Food Purchase: A Conjoint Experiment

101