Use of Consumer Neuroscience in the Choice of Aromatisation as Part of the Shopping Atmosphere and a Way to Increase Sales Volume

Abstract

The point of purchase is considered to be one of the few communication channels that is not yet saturated, and it has a relatively large potential for the future. A pleasant smell is also part of the shopping atmosphere. How smell affects customer behaviour and purchasing decisions is addressed by a relatively young scientific field, one of the existing kinds of sensory marketing—aroma marketing, otherwise called aromachology. Smell has mainly a subconscious influence; therefore, its examination is appropriate to be carried out using consumer neuroscience tools. This paper examined the perception of the shopping atmosphere in Slovak grocery shops and comprehensive interdisciplinary research on the impact of selected aromatic compounds on the cognitive and affective processes of the consumer, as well as the evaluation of the effectiveness of their implementation in food retail establishments. At the end of the paper, we recommend the possibilities of effective selection and the implementation of aromatisation of different premises, by which the retailer can achieve not only a more positive perception of the shopping atmosphere, but also an increase in retail turnover in individual sales sections.

1. Introduction

1.1. Retail Atmospherics

In the new era of retail, the importance of online shopping has grown, yet offline retail channels are irreplaceable in some features and are still justified in a way [1]. We are currently experiencing a kind of convergence of the online and offline worlds, but it will be crucial for the future of retail to know what in the two worlds are similar and vice versa and above all how the two worlds will affect new technologies [2]. The pressure of online stores forces brick-and-mortar retailers to innovate continuously and create a retail atmosphere to be an unforgettable shopping experience for customers, allowing them to compete with online shopping and keep their customers [3,4]. Retailers invest in the atmosphere of the store to create the most attractive environment for their customers that will drive shoppers to buy and support their business success [5,6]. The retail atmosphere can influence the consumer’s decision-making process in its various stages—evaluation, purchase, and after purchase [7]. Atmospheric retail variables are divided into external (external shop display, shop size, parking options, etc.) and internal, such as lighting, music, fragrance, or staff [8]. The rationale for solving this topic lies in the fact that the sales environment is one of the most striking aspects of sales, since most consumer decisions are made directly in the shop. In doing so, the decision is made not only on the basis of the price, quality, or visual communication of the shop, but it is also based on the subconscious influence of elements of the atmosphere of the purchasing environment [9]. One of the possibilities for creating a pleasant atmosphere in the shopping environment is the aromatisation of the premises, which was the object of our further exploration, and we are dealing with it more closely.

1.2. Olfactory Atmospherics

Studies show that different visual, auditory, olfactory, tactile, and taste atmospheric variables can influence customer perception and behaviour, with their combined effect likely to be even greater than the sum of their individual parts [10]. People perceive the world with all the senses at the same time, so the more sensual the experience, the more engaging it is [11]. Sense elements are an important aspect of retail outlets, as they can unknowingly influence consumer judgements and behaviour at the place of purchase [12,13,14]. One of the strongest senses is smell, so many companies use scent-based marketing techniques to create better and more memorable experiences for their customers or to add value to their products and services [15]. Understanding how and when specific fragrances affect the customer experience is key to their commercial success [10]. Fragrance has a huge impact on how people shop and how they relate to brands [16]. Fragrance-based marketing can create a comparative advantage for a particular business if implemented correctly [17]. It is not enough to use a fragrance that “smells well”; the smell must also be consistent with other factors (e.g., sensory environment) [18]. The consequence of marketing is that fragrances used in retail stores or other venues [19] can not only create pleasure, but they also support specific consumer responses, such as categorization, recall, or choice [20]. Although olfactory marketing has increased its presence in the commercial environment in recent years, the number of scientific studies on this issue is still limited [21]. Similarly, a small number of scientific studies on food-related aromachology are available, and methodological practices vary considerably between studies, making it difficult to compare their results [22]. Several studies have identified reactions to the smell of approach, such as stronger intentions to visit the store, to spend more time in it, and to seek diversity [23] and a willingness to pay a higher price and to spend a longer time viewing the goods [24,25,26,27]. Knowledge on the effectiveness of aromachology in influencing actual customer behaviour and ultimately increasing turnover is equally incomplete, as published results are often based on a highly controlled laboratory environment rather than a real business environment [28]. As mentioned above, customer reactions to odours can be the result of automatic and unconscious cognitive processes that occur without their knowledge [29]. Therefore, we considered the use of consumer neuroscience tools to be justified in the field of aromachology research. With this research, we intended to build on our previous studies, in which we also pointed to an innovative approach to assessing the impact of fragrances on consumer preferences, both in the laboratory and in real conditions, using objective neuromarketing measurements [30,31,32,33].

1.3. Consumer Neuroscience

Obtaining clear and objective information should be the basis for any marketing research. The emergence of consumer neuroscience as an extremely valuable part of marketing research has prompted restrictions on traditional research techniques aimed at self-assessment [34,35]. The main weakness of traditional research techniques is that they are only able to capture conscious reactions and customer decisions, while neuroimaging techniques offer objective measurements of nerve processes that appear in the process of decision-making itself [36]. The benefit of neuromarketing as a hybrid area, which encompasses the three main areas of neuroscience, psychology, and marketing, is that it makes it possible to capture the unspoken cognitive and emotional response of consumers to various marketing stimuli [37]. The aim of neuromarketing as a research technique is not to replace traditional research methods, but rather to provide unique and complementary information, and thus a different perspective on the issue under investigation [38,39]. Terminologically, it is necessary to distinguish between “consumer neuroscience” and “neuromarketing”. While the first term was used earlier in the academic sphere, the second is linked to the specific use of neurophysiological tools for the purposes of researching a particular society [40,41]. However, many companies currently use the term “consumer neuroscience” to describe their commercial solutions, thereby confusing the original intention with terminological resolution [42]. In connection with the subject of our research, the use of neuromarketing research is justified, as neuroscience and psychology not only make it possible to better understand the unconscious reactions of consumers, but they can provide a more comprehensive view of sensory perceptions [43]. Human perception is inherently multisensory; therefore, the knowledge from cognitive neuroscience about how it works is ultimately crucial to understanding and explaining customer experience [10,44]. Consumer neuroscience has a wide range of tools that Ref. [45] broke down into body response measurements (face expressions, mimic musculoskeletal movements, eye movements and fixations, blinking, electrodermal activity, pulse, blood pressure, breathing, etc.) and brain response measurements (blood oxygenation, electric fields, magnetic fields, etc.). Neurophysiological measurements are usually more time-consuming and costly, based on technical knowledge and, with the exception of mobile EEGs, are only usable in laboratory conditions. In contrast, physiological tools are more affordable, easy to integrate with other tools, and do not require much technical knowledge or preparation time. Moreover, their applicability outside the laboratory conditions contributes to the study of consumer behaviour directly in real-world conditions and, therefore, in commercial establishments [41,46]. One of the most widely used neuroscientific techniques for marketing studies is electroencephalography (EEG). The relatively low cost and portability of this device are crucial for its wide usability [36]. It is a non-invasive method that detects changes in electrical currents in the form of brain waves that arise from exposure to a tested person to a stimulus [47]. When selecting individual neuromarketing research techniques for the needs of a particular research problem, it is always necessary to take into account the required accuracy of the results and possible limitations of individual neuroscientific devices [48]. We decided for a version of the Emotiv EPOC+ device. The validity of the data obtained through the mobile electroencephalograph (EEG) from Emotiv EPOC has been verified by several researchers [49,50,51], who used this device to investigate the emotional state of respondents and who demonstrated that the device provides the same results as traditional stationary electroencephalographs (EEGs). Despite the huge contribution of neuromarketing research in academic or commercial research, it should be stressed that most neuroscientific consumer studies were conducted in laboratory conditions mainly focused on individuals’ brain responses to simple marketing stimuli. This shows that studies carried out so far have not fully exploited the potential of the neuromarketing approach involving more real and social consumption experiences [38].

2. Materials and Methods

The object of examining the present contribution is the perception of the shopping atmosphere in food retail stores and to what extent a pleasant smell is involved in its perception. At the same time, the impact of fragrance on purchasing behaviour by selecting and deploying fragrances in real retail conditions was also investigated. In this context, the aromas tested were suitable for use in the pastry and bakery products department, as according to the company’s internal data, the most impulsive purchases are made in this department, especially in the sub-category of unpacked pastries. The research process was divided into three separate parts/experiments.

2.1. Stage 1—Quantitative Survey of Perception of Shopping Atmosphere in Slovakia

Primary data on the perception of shopping atmosphere in Slovak grocery shops and the main buying motivators were obtained through an exclusive quantitative ad-hoc survey in the form of CAWI via an online managed access panel (full managed access panel). The survey was conducted through an online panel of the external company MNFORCE, Ltd., which provides comprehensive solutions from the areas of Knowledge Discovery in Databases (KDD) and marketing research. The survey was carried out in the period of 1 March 2020 to 5 March 2020 on a representative sample of 717 respondents of the economically active population responsible for purchasing food. The sample consisted of 53% men and 47% women, aged 18–29 years 16%, 30–39 years 17%, 40–49 years 21%, 50–59 years 20%, and 60 years and over 26%, with an even representation of all regions in Slovakia. A total of 54% of them live in cities with a population of up to 19,999 inhabitants.

The questionnaire consisted of 14 questions. We used both descriptive (numbers, quantiles, averages, and standard deviations) and inductive statistics for the evaluation.

In addition, we calculated an NPS (Net Promoter Score) from the data on the evaluation of the purchasing atmosphere in individual chains. Despite many critical views, the NPS is a globally recognised methodology for measuring customer loyalty. The NPS score is represented by a number ranging from −100 to 100 [52]. The result indicator is calculated using a relationship:

% Promoters − % Critics = Result of NPS

At the same time, through a statistical test, we verified differences in the perception of the purchasing atmosphere in individual chains. To identify differences, the data were transformed into categories—three categories—thus dividing the respondents into three categories according to their assessment of the shopping atmosphere (critical, neutral, and positive perception), while the Cochran Mantel-Haenszel test was used to frame whether there was a difference between the single outlets globally. When it was found that there was a difference, we used the Chi-square test as a post-hoc analysis to identify differences for the specific factors influencing perception.

It also included an association issue with a view revealing consumer preferences and a narrower pre-selection of aromas suitable for the department of pastries and bakery products. The test used GroupSolver, a branded tool that combines data mining, machine learning, and advanced statistics, allowing to dynamically process open responses of respondents through a self-calibrating tool and to quantify the quality findings (insights) [53]. The task of the respondents was to write the first word that struck them when they said pastries and bakery products. Wordle analysis was used to visualise the data. On the basis of the results, the aromas that were most portrayed by the associations of respondents associated with the sales department in question were profiled.

2.2. Stage 2—Testing in Laboratory Conditions Using EEG and Face Reading (FA)

Aromatic compounds, which were profiled on the basis of the initial associations and consultation with the management of the retail company, were subject to further testing under laboratory conditions. The added value of qualitative testing was the real testing of samples of the aromas in question (interaction with samples), as well as the recording of direct implicit and explicit feedback. A total of 53 participants participated in Experiment 2, with customers buying at the retail outlet in question (they buy food most often in the particular store). The sample size was determined by taking into account similar studies. In a previous study, we examined the perception of aromas for the candy department on a sample of 48 respondents [30]. Ref. [54] also examined the impact of aromas using electroencephalography (EEG) on a sample of 16 respondents in Croatia. The commercial survey agency 2muse, Ltd. (Bratislava, Slovakia) monitored the impact of Christmas fragrances on consumer emotions on a sample of 20 respondents [55]. The European Society for Opinion and Marketing Research (ESOMAR) claimed that most survey agencies that use consumer neuroscience use much smaller sample of respondents than in traditional market research, while 15–30 respondents are enough to achieve quality results [56].

The process of research under laboratory conditions itself consisted of five parts:

• Initial instruction of respondents, completion of consent to biometric and neuroimaging testing, as well as the processing of personal data in accordance with the GDPR and the Code of Ethics of Sociological Surveys.
• Olfactory sensitivity threshold test.
• Entry-driven interview in the form of CAPI.
• Tests of selected aromas using consumer neuroscience and aromatising box tools.
• Interview in the form of CAPI.

The olfactory sensitivity test was conducted to create segments of consumers with different sensitivity thresholds. The certified test from the German company Burghart contains three sets of samples (one N-Butanol; two phenylethanols) marked in red, blue, and green, arranged in descending order from 1 to 10.

The Emotiv EPOC headset was used to measure brain activity, consisting of 14 data and 2 reference electrodes distributed in accordance with the international 10–20 electrode distribution system, following international standards according to [57,58] that were distributed in the following positions AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8, and AF4 illustrated by Figure 1. We used a version of the Emotiv Epoc + device. The Emotiv EPOC wireless kit is capable of detecting the most important functions of brain activity that are commonly used in medical computer processing/brain-Computer Interface (BCI) simulations.

We monitored two basic emotional states, emotional engagement and frustration. Individual emotions were calculated on the basis of electrical activity recorded through a group of electrodes necessary for the calculation of the emotion. EEG channels that were required to calculate emotional engagement were allocated in positions O1, O2, F3, F4, F7, F8, FC5, and FC6. By means of gradual regression, the variables for this model were profiled from similar studies [59,60,61].

In the case of frustration, there were no specific studies to directly calculate it based on recorded brain activity through electroencephalography (EEG), but there are several studies [62,63,64] that clarified the role of the prefrontal cortex and the cortex lobe in recognising frustration, which provides some form of credibility of our experiment.

The data in question were obtained using the software tool Affective Suite, which records changes in emotions in real time. Each person tested had their own unique profile, in which the data obtained were recalculated according to certain personality characteristics. The sum of these differences was then used to normalise the data.

At the same time, emotional feedback was monitored through the somatic biometric method FaceReader 7 from the Dutch company Noldus, which identifies the emotional feedback (valence, excitement) of respondents with maximum accuracy based on observable changes in mimic muscles and recognises basic micro-emotions (happy, sad, angry, disgusted, surprised, and neutral) [65]. The validity of the recorded data mainly affected the sensing angle, the luminance of the environment, and the resolution of the recording equipment [66]. Facereader has been utilised in much research. We can mention, for example, [67], who focused on the examination of using the facereader method to detect emotional reaction to controversial advertising referring to sexuality and homosexuality. Similar investigations highlighting the importance of examining facial biometrics and microemotions have been carried out further, for example on smaller groups of respondents. One such survey focused on the selection of the most suitable aroma for a selected café in Slovakia, and it was carried out on eight respondents (see [31]).

The data collected from individual measurements were synchronised and correlated with each other in the Noldus Observer XT 10 programme environment. This programme allows to synchronise structured and unstructured data from individual devices while creating custom variables during the realisation of experiments [68,69]. The survey was carried out in the period of 01 May 2021 to 15 June 2021.

2.3. Stage 3—Experiment on the Influence of Aromas on Consumer Behaviour in Real Conditions

Based on the results of Experiment 1 and Experiment 2 and after consulting with the management of the retail company, we decided to put aroma into the department of pastry and bakery products, which provoked the respondents’ most positive feelings and a relatively higher levels of bias.

In view of the objectives set, two aromatisation units from REIMA AirConcept AS650 were placed in operation (one separately and the other as part of the feedback kiosk). The location (Figure 2) was also carried out with regard to the airflow in the space (taking into account the distribution of the intakes and evaporations of the treated air in operation).

The initial setting was made on the basis of the available knowledge and recommendations of Aroma Marketing, Ltd. (Nové Zámky, Slovakia), which deals with the aromatisation of the premises, with the lowest intensity set and controlled by the REIMA AirConcept mobile application (Figure 3). The interface also included the setting of operating modes depending on the opening hours.

The aroma intensity adjustment was calculated according to the size of the aromatised space and the exchange of air in the space. The volume of the sales section was approximately 500 m³. The air exchange in the space was at the level of 300 m³/h. The aromatisation unit allowed for adjusting the aromatisation strength in the range from 0 to 1.0 mL/h at a step of 0.1 mL/h. In our measurement, aromatisation was set at 0.6 mL/h. This, after stabilisation, ensured the concentration of aromas filling in the air at a level of approximately 1.1 mg/m³. At the same time, environmental factors that had a major influence on the spread of aromas in the space were continuously monitored [70].

All aromas used were safe in accordance with international standards and were manufactured under the supervision of the Research Institute for Fragrance Materials (RIFM) [71]. The use of aromas in research was governed by the ESOMAR Code of Ethics for Sociological surveys. The deployment of aromas in real conditions was governed by rules developed by the International Fragrance Association (IFRA) [72], which sets out precise procedures and recommendations while respecting consumer rights. Last but not least, aromatisation was also governed by the internal code of ethics of the participating institutions, in this case, the company Kaufland and the Slovak University of Agriculture in Nitra.

From the point of view of examining the impact of aromatisation on selected economic indicators in the selected commercial operation, data on the number of units sold and sales within the individual sub-category of pastry and bakery products, as well as the adjacent sales departments with a colonial (coffee/tea) in the Kaufland Bratislava—Petržalka store, were studied. Given the situation with the COVID-19 pandemic and the possible misrepresentation of the results, the data obtained were compared with the reference store Kaufland Nové Mesto nad Váhom, since, according to the company’s internal data, this was a comparable operation in terms of purchasing power, assortment of composition, layout of the store and size of the sales area, economic indicators, and location of the branch. The survey was carried out in the period of 10 November 2020 to 31 January 2021.

Within the framework of individual research activities, the aim was to verify the following hypotheses, which primarily concerned the demonstration of differences in the emotional perception of aromas. We tested the hypothesis at a materiality level of 0.05.

Hypothesis 0 (H0).

Respondents perceive aromas equally in terms of emotional bias.

Hypothesis 1 (H1).

Respondents perceive aromas differently in terms of emotional bias.

Hypothesis 2 (H2).

Respondents perceive aromas equally in terms of frustration.

Hypothesis 3 (H3).

Respondents perceive aromas differently in terms of frustration.

Hypothesis 4 (H4).

They are the same = there is no difference.

Hypothesis 5 (H5).

They are different = there is a difference.

The obtained processing of the data was processed through descriptive and inductive statistics in programme environments: MATLAB R2020a; RapidMiner 9.3; Mathematical statistical programme R version 3.6.3; CCA version 1.2.1 package; Microsoft Excel 2010.

3. Results

In order to obtain information on the perception of the shopping atmosphere in the conditions of Slovak retail, we conducted a quantitative survey on a sample of 717 respondents. In the first part, respondents were presented with certain characteristics of food sellers, and their task was to select the names of the shops to which the property would be suitable. A number of shops could also be selected for one property. The results in Figure 4 show that the “good shopping environment” feature was clearly associated with Kaufland with a total of 318 responses. Relatively often, this property was also associated with the Lidl store (294). The figures obtained for the designation of that characteristic within individual retailers indicated some differences in perception, which can be regarded as an important finding for making further decisions in terms of generating benefits for the USP (Unique Selling Proposition).

After the initial assignment of characteristics to those sellers with whom they were most associated, regardless of their preferred place of purchase, respondents continued by selecting the retailer where they buy food most often. The results in Figure 5 show that they most often shop in Kaufland (28%), Lidl (25%), COOP Jednota (16%), and Tesco (16%). Respondents who reported “other” most often reported Malina, CBA, Nitrazdroj, Koruna, Klas, and Milk-Agro.

In a separate question, respondents also assessed the shopping atmosphere of the store through a scale from 0 to 10 (where 10 means very pleasant). From the data, we calculated the NPS (Net Promoter Score) indicator to obtain a clearer picture of the perception of the shopping atmosphere in Slovak grocery shops. The indicator can acquire values from −100% to 100%, with negative values indicating the need for improvement. In Figure 5, you can see the NPS score of the shopping atmosphere rating, including the number of respondents who commented on the chain. The results show that Fresh scored the best score (4%), but in this case, only 23 respondents commented on the shopping atmosphere. Among the foreign chains, Billa (0%) and Kaufland (−3%) scored the best, which was also rated by the most (199) respondents. The worst was Lidl, which reached a score of −18%.

The above findings indicated that there are significant differences in perception of the shopping atmosphere in Slovak grocery shops. However, it is necessary to remember that it is not important to know the value of this indicator, but to monitor its development.

At the same time, we tried to find out if there was a difference in the perception of the atmosphere between shops at the declarative level of evaluation of selected properties. Using the Cochran Mantel-Haenszel test (p-value = 0.007287), we verified that there was a conclusive difference between the responses at a chain level; therefore, it was necessary to make comparisons in each category for each chain using the Chi-square test at a specified materiality level for each group. Only in the case of Kaufland (p < 0.001) in favour of one category of response confirmed a significant difference in the responses related to the “shopping atmosphere” rating.

Even when assessing the specific characteristic of the “pleasant aroma” related to the given food seller, it is possible to see again the differences in the rating in Figure 6. In this case, Tesco’s customers did not share the claim most (30%). On the contrary, the customers of Fresh (92%) and Billa (81%) were most associated with this claim.

Based on these findings, we decided to carry out research on the targeted impact of aromatisation on the perception of the atmosphere of the selected sales department and economic indicators. After consulting with the management of the retail store, we focused on the department of pastry and bakery products. This department is specific in that the amount of frozen assortment is baked directly at the point of sale, which causes the spread of the natural aroma that arises during the baking of these products.

The questionnaire survey also included an association question: what will the respondents first think about in connection with the department of pastry and bakery products? Using the Wordle analysis, it was possible, as shown in Figure 7, to see that people most often associated the pastry department with morning, pastry/bread, coffee, crunchiness, but also cake or chocolate.

Based on the acquired associations and available studies and after consulting with the management of the company, we specified five aromas that have been tested under laboratory conditions. Respondents evaluated the aromas of fresh baked bread, coffee and cake, chocolate, cookies, and coffee house at an explicit and implicit level.

In the first stage of laboratory testing, respondents were asked to select the one aroma they considered most suitable for placement in the department of pastry and bakery products without the possibility of smelling the sample. As can be seen in Figure 8, most respondents (24) labelled fresh baked bread (14) and the smell of coffee and cake and chocolate (12).

In the second stage, the respondent smelled individual samples and determined, on a scale from 0 to 10, the appropriateness of placing them in the pastry and bakery products department. During this process, respondents were also monitored for implicit feedback via face reading and electroencephalography.

Figure 9 shows the evaluation of the individual aromas in terms of suitability for use in the pastry and bakery products department. On the basis of average values, respondents identified the most with the aromas that contained coffee ingredients (coffee and cake (7.78) and coffee house (7.25)). On the contrary, they considered fresh baked bread (5.74) to be the least suitable for department of pastry and bakery products.

Based on the measurement of electrical brain activity, an emotional bias (engagement) score was calculated. It can be seen from Figure 10 that, at the implicit level, the coffee house aroma was attracted by most respondents (0.648). This can also be largely related to the increased efforts of respondents to recognise what type of aroma it is.

Using a single-factor variance analysis, we determined the statistically probable difference between the samples of aromatic compounds for the bias indicator. Compared to the observed aromatic compounds in the emotional engagement (interest) indicator, the value (p < 0.001) came out, so we concluded that there was a statistically conclusive difference in perception in terms of engagement between the aromas tested. In a post-hoc pair comparison of aroma pairs, the differences between the coffee house and fresh baked bread aromas were shown.

The highest rate of frustration based on average values was recorded for the cookie sample (0.501) (see Figure 11). It can be assumed that this result was due to the fact that respondents had a problem with identifying this aroma and that not everyone was able to determine immediately what could have caused frustration. The reason may also be the method of smelling itself (the distance of the nose from the grinding test vessel), resp. an aromatic compound that does not have a clear composition and can remind participants of several types of scents.

In general, it can be stated that these are lower values of frustration that are present in solving normal activities. At the same time, it should be emphasised that frustration does not interfere with emotional engagement. We also decided to statistically verify these differences of frustration (concerns) for individual samples.

The result of the test was that there is no evident difference in frustration depending on the change of aroma (p = 0.76).

In addition to electroencephalography, respondents received unconscious feedback by measuring microemotion based on facial expressions. By measuring emotions from facial expressions, we obtained information about valence (polar of emotion) (see Figure 12). In this case, based on the median average, respondents most positively perceived the aroma of coffee and cake (0.05) and the most negatively the aroma of fresh baked bread (−0.02). However, it is necessary to note the possible distortion of the results in that some respondents smiled during the initial smelling of the aroma in some cases, probably because of the initial surprise, which could subsequently artificially increase the positive mood rate for some samples, as the software evaluated the smile as positive feedback recognizing emotions based on facial expressions.

Based on a comparison of valence through the Kruskal-Wallis test, we found that there were statistically significant differences between aromas (p = 0.0122) in the emotions of the participants tested. The differences were confirmed between coffee house and fresh baked bread and also between coffee and cake and fresh baked bread (see

Table 1. Kruskal-Wallis test—comparison of the valence of individual aromas.

	Fresh Baked Bread	Coffee and Cake	Chocolate	Cookies	Coffee House
Coffee House	H5	H4	H4	H4
Cookies	H4	H4	H4
Chocolate	H4	H4
Coffee and Cake	H5
Fresh Baked Bread

Testing at α = 0.1.

).

The results of the conscious evaluation of the aroma samples in question show that respondents evaluated the most positively the aroma of coffee and cake and coffee house. At the subconscious level, the most positive values of valence (coffee and cake 0.05) and emotional engagement (coffee house 0.648) were recorded for coffee scents. On the basis of these findings and in consultation with the executives of Kaufland Slovenská republika, v. o. s., we decided to deploy the coffee and cake aroma under the real conditions of the bakery department. The choice of the aroma itself was also influenced by the fact that the bakery department contained a natural aroma of baked pastries, and the aim of the additional aromatisation was to complement the existing atmosphere in that sales department.

In the third part, in order to verify the impact of aromatisation on the selected economic indicators, we examined the impact of the aromatisation deployed on sales and the number of categories sold of two product groups in the pastry and bakery products department and in the sales outlets located in the immediate vicinity, where the side effect of the spread of the aroma may have been shown.

In view of the situation with the spreading COVID-19 pandemic and after consulting the Business Intelligence department of Kaufland Slovenská republika, v. o. s., we decided to compare the data recorded in the Petržalka store for the same period with the reference operation, in which no aromatisation was used in the pastry and bakery products department. The reason for not comparing the data with the same period of the previous year was mainly due to the significant distortion of results and findings. In this context, data analysts from the company also confirmed changes in consumer behaviour and a significant decrease in turnover in different product categories, which could significantly distort the results of the aromatisation impact. As a reference operation, after careful consideration by Kaufland, it was chosen to operate in Nové mesto nad Váhom. This branch has a similar location, sales area, traffic, assortment, redesign, and, based on Kaufland’s internal data, there is also approximately the same purchasing power in the city district of Bratislava in Petržalka as in the town of Nové Mesto nad Váhom.

In Figure 13, there is a comparison of subcategories in the pastry and bakery department, as well as the adjacent colonial department in the period of 10 November 2020 to 31 January 2021, when aromatisation with the coffee and cake aroma was deployed in the bakery department in the operation in Nitra. In both subcategories of the bakery department, it is possible to see a higher turnover in the Petržalka operation. In the case of colonial, the difference can be seen in the coffee subcategory (6%) in favour of an aromatised shop. The highest difference (18%) was presented by a subcategory of unpacked pastry, of which Petržalka sold EUR 4350 more. A significant difference (6%) can also be observed for the coffee subcategory, where the difference over the reporting period was EUR 1465. The total difference in pastry sales in the establishments covered was EUR 5011 in favour of the Petržalka operation (9%).

The subject of interest was also the sale of goods items in order to eliminate the distortion of data due to discounts and sales, which are also in the competence of store managers. Figure 14 shows a comparison of sales of the number of pieces in the bakery department and the adjacent colonial department in operations in Petržalka and Nové Mesto nad Váhom for the period of 10 November 2020 to 31 January 2021. As mentioned above, the bakery department in Petržalka was aromatised with the aroma of coffee and cake during this period. The biggest differences in terms of the number of items sold can be seen in the subcategory unpacked pastries (25%) and also packed (16%), but in this case, they were in favour of the Petržalka store. Although more packed pastries were sold in Petržalka, the total difference between the number of pastries sold in Petržalka and the reference shop is 51,232 (7%) in favour of the Petržalka store. Even in the case of subcategories of the neighbouring department of the colonial, it is possible to see higher sales in the aromatised operation in Petržalka by 5882 pieces (10%).

In view of the above, it must be concluded that the direct effect of the aromatisation was reflected in sales and the number of pieces sold. Even after considering all other factors that could have caused a higher turnover and the number of pieces sold in the aromatised operation, the costs of aromatisation of the sales department are incomparably lower than the effect obtained in the form of increased turnover resp. quantity of pieces sold.

4. Discussion

The aim of the research phases/experiments carried out was to determine whether there is a difference in perception of the shopping atmosphere of retail outlets and the extent to which a pleasant smell is involved in its perception. We verified the assumption of a different consumer perception of the purchasing atmosphere using the Net Promoter Score (NPS) indicator and the Cochran Mantel-Haenszel test, which confirmed that there is a demonstrable difference in the declarative evaluation of selected properties at chain level. Using the Chi-square test, we found a significant difference in the responses related to the assessment of the “pleasant smell’ property only in the case of Kaufland Slovenská republika, v. o. s.

Ref. [24] state that the presence of the surrounding fragrance in the store can increase natural excitement, thus increasing the attractiveness and pleasantness of the environment. Several studies have identified several reactions to the smell of approach, such as a stronger intention to visit the store, spend more time in it, seek diversity [23], willingness to pay a higher price and spend longer time viewing the goods [24,25,26,27]. In this context, it is proven that the longer time spent at the point of sale influences a more impulsive purchase [73]. At the same time, the object of research were aromas suitable for the use in the department of pastries and bakery products, while the tools of consumer neuroscience were also used, similarly to the study. Refs. [30,74] also highlight the choice of the right aroma for real-conditions deployment, arguing that marketers are increasingly using the surrounding scent as a strategic tool to differentiate themselves from the competition, attract customers, stimulate sales, influence moods and create overall pleasant and unforgettable shopping experiences. There are several aspects that affect the consumer when making decisions and choices, including mood or emotional state [75,76]. The role of emotions in the consumer decision-making process is explained by the principle of neurological and cognitive frameworks, such as somatic marker theory [77], which focuses on so-called attention to negative impacts in decision making. The presence of aromas in grocery shops is influenced by naturally released fragrances from exhibited goods in individual sales departments, the smell of baked pastries, the amount of people in the space and the adjustment of air conditioning [30]. Last but not least, the resulting impression of smell at the point of sale may be influenced by aromatisation in individual sales departments, e.g., the smell of coffee in the pastry and bakery products department. Based on the conscious feedback from our respondents, the aromas of coffee and cake (7.78) and coffee house (7.28) were mostly liked as well. The worst rated aroma was fresh baked bread (5.74), which could be due to the composition of the aroma itself (insufficient imitation of synthetic aroma). Fresh baked bread (24 respondents) and coffee and cake (14 respondents) were the most preferred aromas on the basis of the classic feedback. An interesting finding was that, while the respondents did not have the opportunity to smell the samples, they recommended the fresh baked bread fragrance as the best for the bakery department, but when testing the samples themselves, they evaluated the coffee aromas most positively, which can be a suitable addition for this department. For example, the smell of freshly baked bread is used in supermarkets to motivate customers to buy more. When the smell of freshly baked bread came into the grocery shop, sales in the bakery department tripled [78]. The fact that the fragrance can increase the sales volume is confirmed by the New York grocery chain, which admitted that it aromatises its premises with the smell of chocolate and fresh baked bread to make customers feel hungry and thus increase sales [79]. In this context, Kaufland’s data analysts also confirmed changes in consumer behaviour and a significant decrease in turnover across product categories, which could significantly distort the effects of aromatisation. When comparing the period of 10 November 2020 to 31 January 2021, when the aromatisation of the pastry sales department with the aroma coffee and cake was installed in the Petržalka operation, a difference of EUR 5011 was recorded in favour of the reference operation in Nové mesto nad Váhom compared to the operation in Petržalka. In the case of the neighbouring department of the colonial, there was also a difference in the coffee subcategory (6%) in favour of the aromatised store. Despite the fact that a larger amount of packed bread was sold in Petržalka, the total difference between the number of pieces of bread sold in the Petržalka store and the reference store was 51,232 pieces (7%) in favour of the aromatised store in Petržalka. Other studies talked about the ability of individual aromas to induce positive sensations, such as the smell of cookies has the power to induce positive feelings [80]. In our study, the first stated hypothesis was confirmed that there is a difference in emotional involvement between the tested aromas. We were unable to prove the second stated hypothesis, whether there are differences in frustration, as there was no statistically significant difference in the perception of respondents in this case. In terms of the perception of individual aromas, we tested the third stated hypothesis that there is a difference in their perception. Classical surveys are usually used to understand the perception of aromas. However, these approaches require larger samples of respondents to obtain reliable results and are based on the assumption that participants are able to express their preferences, as argued in [81]. However, it should not be forgotten that food stimuli can influence preferences and eating habits even unknowingly, as [82] claimed; therefore, the use of neuroscientific methods is justified in this area. Their main objective is to measure the critical aspects of consumer perception not only in the area of unconsciousness (attention, emotional response, and memory) but also in the area of declarative (position and preference) [83,84]; thus, findings can be applied in the management decision-making in the creation of effective sales strategies [85].

In this context, it should be noted that the development of consumer neuroscience discipline in the food sector and retail also faces some concerns about the interpretation and findings of some commercial studies carried out so far [86], given the fact that some companies make even contentious claims without evidence-based citations [87,88]. In this context, it should be borne in mind that academic studies are based on strict protocols and adherence to methodological procedures [89], which allow for a new perspective on the unconscious perception of the consumer. Nevertheless, even in this context, there are some critical views on the existence of strong unconscious influences on decision-making and related behaviour [90].

5. Conclusions

In our paper, we examined the perception of the shopping atmosphere in Slovak grocery shops and the impact of selected aromas on the cognitive and affective processes of consumers using consumer neuroscience tools. We also focused on assessing the cost-effectiveness of their implementation in retail operations.

As part of the first stage/experiment survey on perception of the shopping atmosphere in the conditions of Slovak retail, which we carried out on a sample of 717 respondents, it was clear that the characteristic of “good shopping environment” is associated with Kaufland, with a total of 318 responses. The above findings indicate that there were significant differences in the perception of the shopping atmosphere in Slovak grocery shops, and there is room for improvement.

From the initial words (associations) given by respondents, GroupSolver identified the related fragrant preferences for the sales department of pastry and bakery products. The results show that the respondents associated the bakery department mostly with the morning, bread/rolls, coffee, crunchiness, and cake or chocolate. Based on the results and consultation with the company dealing with space aromatisation, available studies, and the existing portfolio of fragrant compounds, five aromas meeting the specified attributes (universality, suitability for the department of pastry and bakery products) were profiled. On the basis of the above, the fragrances specified were fresh baked bread, coffee and cake, chocolate, cookie, and coffee house, which formed the basis for further testing under laboratory conditions.

The second part/experiment consisted of research into the impact of selected aromas on human emotions using electroencephalography (EEG) and facial expression tracking (FA) in laboratory conditions. Aromas, which were specified on the basis of the previous test, were tested in a special aromatising box in the Laboratory of Consumer Studies FEM SUA in Nitra.

The test was conducted with 53 respondents who were subjected to an olfactory sensitivity test. Based on measurements of brain activity, the highest rate of emotional engagement based on the median of this emotion was recorded with coffee house (0.648). In general, the two coffee aromas tested achieved a higher degree of bias. In this context, using a single-factor variance analysis, we identified a statistically probable difference between samples of aromatic compounds in terms of emotional engagement, which showed that there was a statistically significant difference in perception in terms of engagement between the aromas tested. The highest rate of frustration based on average values was recorded for the cookie sample (0.501). It can be assumed that this result was due to the fact that respondents had a problem identifying this aroma, and not everyone was able to recognise it, which could have caused frustration. We decided to statistically verify these differences in frustration (concerns) for individual samples as well. However, there was no statistically significant difference in the perception of respondents in this case. The most positive emotions were recorded with coffee aromas (coffee and cake 0.05; coffee house 0.04). Based on the mutual comparison of valence through the Kruskal-Wallis test, we found that there were statistically significant differences in the emotions of the participants tested between aromas. In this context, post-hoc tests of the pairs were performed, and it was found that respondents reacted statistically differently, in particular, for coffee house and fresh baked bread, as well as between coffee and cake and fresh baked bread.

Based on these findings and after consulting with the managers of the retail company, we decided to use the aroma of coffee and cake in the real conditions of the pastry and bakery department.

The last stage (3) consisted of research of the impact of the selected aroma on economic indicators in the selected commercial operation. In this case, in view of the situation with the spreading COVID-19 pandemic and after consulting the Business Intelligence department of Kaufland Slovenská republika, v. o. s., we decided to compare the data recorded in the Petržalka store for the same period with the reference operation in which aromatisation was not deployed in the bakery department. The reason for not comparing the data with the same period of the previous year was mainly due to the significant distortion of results and findings.

Our assumption of the impact of aromatisation of the sales department on consumer decisions in the form of a 7% increase in turnover and sales of pastries (increase 9%) and colonial (increase 10%) compared to the reference operation was confirmed. At the same time, taking into account all factors that may have influenced and distorted the increase in sales and turnover in the bakery and colonial department, it can be stated that aromatisation was effective for retail. These factors included price concessions, promotional products, availability of assortments, seasonality, and weather.

The three main factors that influenced the cost of aromatisation included the size of the aromatised space (in m²), the time of aromatisation, and the exchange of air in the space. In addition, the aromatisation equipment itself can be bought or leased by the seller. Most companies dealing with the aromatisation of premises provide complex packages of services, including regular service and filling, which is an advantage from the management point of view in a larger number of operations and related costs, especially for larger business entities.

Based on empirical knowledge and restrictions related to the pandemic, we plan to carry out a similar research project with an even larger sample of tested respondents, taking into account the weather, season, olfactory sensitivity (anosmia, hyposmia, normosmia), and participants’ fatigue (beginning and end of the week). Future research will be carried out under different air quality conditions (CO₂, VOC, temperature, humidity) to identify possible changes in smell perception. Given the need to cover the upper respiratory tract, smell perception will also be simulated in order to quantify the impact of pandemic restrictions (mandatory upper respiratory tract protection in buildings) in real-life conditions. In terms of the technologies used in this study, we would like to carry out similar research with a 32-channel electroencephalograph (EEG) and its extension to the biometric method of measuring skin resistance (GSR). From the point of view of economic indicators, we plan to deploy aromatisation to reference plants and to monitor longer periods of time, as well as seasonal effects.