Interactions of fast‐moving consumer goods in cooking: Insights from a quantitative ethnographic study

Fast‐moving consumer goods (FMCGs) are increasingly equipped with enhanced packaging that incorporates novel functionalities. Providing FMCGs with this technology is challenging due to their unique characteristics, such as their low cost and short lifespan. Knowledge derived from a comprehension of their interactions in practice can help develop FMCGs that better cater to consumer needs and are well‐integrated into real‐world contexts. To help develop a method for the formation of such practical insights, 20 households were visited where participants were then observed as they cooked a meal. The sessions were captured on video, and a detailed record of the interactions between individuals, FMCGs and other items was made. A quantitative ethnographic approach was applied to analyse and build an understanding of different aspects of these interactions including their frequential, sequential and correlational features. The findings are discussed through the lens of how an appreciation of the interactions of FMCGs can serve as a valuable guidance for the design and development of their enhanced counterparts. The discovery that FMCGs are linked to the use of other items, for instance, is proposed as an opportunity to make use of the unique properties of the other items that a given FMCG commonly interacts with as a resource to create functionalities. As an exploratory reflection of how FMCGs are utilised in practice, the methods and knowledge presented in this study can be valuable in creating enhanced FMCGs by advocating for a product development process in which decisions are firmly grounded in empirical insights.

Fast-moving consumer goods (FMCGs) are increasingly equipped with enhanced packaging that incorporates novel functionalities. Providing FMCGs with this technology is challenging due to their unique characteristics, such as their low cost and short lifespan. Knowledge derived from a comprehension of their interactions in practice can help develop FMCGs that better cater to consumer needs and are wellintegrated into real-world contexts. To help develop a method for the formation of such practical insights, 20 households were visited where participants were then observed as they cooked a meal. The sessions were captured on video, and a detailed record of the interactions between individuals, FMCGs and other items was made. A quantitative ethnographic approach was applied to analyse and build an understanding of different aspects of these interactions including their frequential, sequential and correlational features. The findings are discussed through the lens of how an appreciation of the interactions of FMCGs can serve as a valuable guidance for the design and development of their enhanced counterparts. The discovery that FMCGs are linked to the use of other items, for instance, is proposed as an opportunity to make use of the unique properties of the other items that a given FMCG commonly interacts with as a resource to create functionalities. As an exploratory reflection of how FMCGs are utilised in practice, the methods and knowledge presented in this study can be valuable in creating enhanced FMCGs by advocating for a product development process in which decisions are firmly grounded in empirical insights.

K E Y W O R D S
consumer packaged goods, data-driven design, enhanced packaging, packaging design, practice perspective

| INTRODUCTION
Fast-moving consumer goods (FMCG) are relatively low-cost, nondurable goods that are partially or entirely consumed upon each use. 1 FMCGs have a pervasive presence in people's lives, as they are an essential component of many of our daily activities. FMCGs not only constitute one of the largest areas of consumer spending 2 but they are also associated with some of the greatest concerns of society including healthy eating, 3 well-being, 4 and sustainability. 5 There is a growing effort within the industry to make innovations in FMCGs by incorporating technology into packaging, which imbue them with additional functionalities. 6 To this end, some FMCG packaging is being equipped with sensor technologies and is being integrated into communication protocols to transform them into enhanced products, which both improve the efficiency of their primary functions and incorporate novel functionalities that improve the consumer experience. 7 However, the quest for useful innovations for enhanced FMCGs has been challenging, and commercially-available implementations have thus far struggled to sway the market.
This research is based on the premise that, in order to develop enhanced versions of FMCGs, it is fundamental to first understand what is accomplished in these items and then to envision what might be accomplished with their enhanced versions. A deeper understanding of such context has proven useful in designing innovative packaging 8 and developing guidelines for assessing package performance. 9 Nevertheless, despite significant progress in the study of user-packaging interaction (for a review, see Mumani and Stone 10 ), there is a lack of understanding concerning how FMCGs are utilised by consumers beyond merely investigating pragmatic and primary functions of the packaging. 11 There is information about the production of FMCGs in factories, as well as their management and sale in the retail sector, 12 but little is known about how consumers utilise these products inside households. Building upon an understanding of the use of FMCGs in practice, this study seeks to apply that information to provide insights that can lead to creating designs providing features that benefit consumers and society, 13 developing novel digital dimensions for interaction and user experience 14 and making products which are more likely to be smoothly integrated into their practical context of use. 15 The purpose of this research is therefore to obtain a detailed understanding of interactions with and the actual usage of FMCGs in practice and to reflect on how such understanding might serve to later guide the design of enhanced FMCGs. A practice is a specific way of conducting a routine and is treated as the smallest unit of analysis and intervention by the practice perspective. 16 This perspective is adopted here to study the interactions of FMCGs, and other items involved in the practice of cooking. Among the different practices FMCGs are involved in, this study focuses on domestic cooking because of the unique characteristics it possesses, which make it an ideal candidate to reveal the common and familiar but oftenoverlooked patterns of use of FMCGs. Some relevant characteristics of cooking include the considerable amount of time most people spend on this practice daily, 17,18 its occurrence within the confines of a kitchen, and that people frequently express enjoyment, confidence and desire to improve their cooking skills, 3,19,20 and food-related FMCGs compose the largest group by sales of all FMCGs. 21 A quantitative ethnographic approach, 22 which integrates both quantitative and qualitative methods, is employed to analyse and give meaning to the vast amounts of data gathered about the interactions of FMCGs.
The approach was applied in analysing four different features of the items' usages and provided insights concerning their interactions.
Each of the findings can help guide the development of their enhanced versions by providing insights from a unique standpoint.
This study seeks to provide contributions in three different areas:

| Enhanced FMCGs
Currently, one of the most important areas of innovation for FMCG industries is focused on equipping packaging with technology.
FMCGs come in packaging, which conventionally serves the functions of containment, protection, communication and convenience. 23 Unlike durable goods, an FMCG's packaging is often part of the product itself, as it serves a functional purpose and therefore acts as a component of the product utilisation. 24,25 Improvement in the packaging helps FMCGs to better serve their primary purposes and to enhance consumers' experiences through added functionalities. 26 The FMCG sector is expected to reach more than 17 trillion USD by 2025, 21 and there is a growing competitive pressure from within industries for them to find replacements for their conventional packaging. 27 This pressure is driven primarily by two factors. One is the potential of enhanced FMCGs to meet the constantly shifting needs and demands of consumers, generally motivated by issues such as environmental concerns, the rise of digital consumers and changes in peoples' lifestyles. 28,29 The second is the increasing availability of technology necessary for the viability of enhanced FMCGs 30 such as the proliferation of smart devices and digital services, the standardisation of communication protocols 31 and declining prices of the required technologies.
Research into enhanced FMCGs has focused primarily on applications of active packaging and intelligent packaging. 26 Active packaging uses biomaterials to directly interact with its contents and improve their quality, safety and shelf life. 32,33 An example of active packaging is antimicrobial packaging, which reduces contamination and growth of microbial populations in meat products, thereby extending the shelf life and safety of the product. 34 Intelligent packaging contains a component, which monitors and provides reliable information about the condition of the product, packaging and the surrounding environment. 35 An example of intelligent packaging is the inclusion of freshness indicators that monitor the quality of perishables by reacting to the metabolites in the content, then relaying that information to consumers. 36 A third type of enhanced packaging is interactive packaging, which is experiencing a sharp growth in interest surrounding the development of new implementations. Interactive packaging refers to that which is based on reciprocal interactions between people and the product, thereby creating a two-way communication channel and provide a dynamic response from technology-based systems. 37 Interactive features include providing functionalities such as entertainment, helping with the collection of feedback and facilitating management. 38 Interactive FMCGs find their place on the Internet of things (IoT), a paradigm envisioned as a network of interconnected machines and devices capable of interacting with each other. 39 The core of the IoT is the collection of information by objects about their surroundings and the use of such information to make sense of their interactions and respond accordingly. 40 However, the research connecting FMCGs to the IoT is still in its early stages. This is in spite of the recent emergence of fields such as human-food interaction, 41 which are specifically aimed at enhancing one's experience with products and their packaging. 42 The limited interactive FMCG implementations have commonly focused on a specific aspect of a product's use. Some examples include (a) marketing-providing product information and functionalities to facilitate and encourage brand selection and decision-making at the point of purchase, 43 (b) product experienceincorporating entertainment such as playing music that makes interactions with and consumption of a product more enjoyable 44 and (c) security-assisting in and facilitating the authentication of products, as well as preventing counterfeiting. 45 However, interactive FMCG implementations and associated devices have still yet to be successfully extended to a large segment of products. The implementations, which have been put on the market, have experienced only a short-lasting interest from consumers and are almost always plagued by unforeseen issues, which only became apparent when finally deployed in real-life situations. The Amazon Dash Button, aimed at helping consumers to reorder FMCGs, was discontinued because, among other issues, its main function was found to be redundant given the availability of smart assistants. 46 TagItSmart, a platform to help with the lifecycle management of FMCGs, 45 has not been able to achieve widespread acceptance as the industry has still chosen to rely upon currently-existing and longstanding solutions such as barcodes. A smart bottle, which offered entertainment such as music to complement the drinking experience, 47 was discontinued after a brief period on the market.
One reason for the discontinuation of that and similar enhanced FMCGs may be that consumers simply never gained more than a passing interest and treated these products as little more than a novelty. 48 FMCGs, like any other object, lend themselves to a particular set of actions and interactions, which need to be observed from the dyadic human-object level to the practical and contextual level. However, while a considerable effort has been made towards understanding these isolated user-product interactions, there is a lack of more than a superficial understanding of interactions involving FMCGs in a more practical, contextual use. In a review of more than 100 studies on user-packaging interactions, of which approximately two thirds were related to FMCGs, it was found that the existing research was predominantly focused on the 'point of purchase' and 'checkout' stages of interaction. 10 Only four studies were found on 'handling', and even those focused solely on the ergonomic and mechanical properties of interactions rather than on their more practical use. The research on packaging has been conventionally dominated by engineering 49,50 and ergonomic aspects of packaging, 51 and even the more recent innovative methods 42 have not considered the practical use of FMCGs. A deeper understanding of the context of packaging use has proven effective in developing guidelines for assessing the packaging performance 9 and in providing a framework for design. 52 Practical knowledge about the use of objects can lead to broadening the set of elements considered in the design process and create products more likely to be adopted by consumers. 15

| Domestic cooking
There is a vast body of literature on cooking including large academic compendiums such as historical accounts and diverse non-academic sources such as cookbooks. This section makes no pretence to serve as an exhaustive review; rather, it aims to provide an account of some of the studies, which have more closely focused on interactions with FMCGs.
Conducting fieldwork studies, researchers attempt to get firsthand experience by observing people's practices as they occur in their natural settings and by immersing themselves as much as possible in the activity. A fieldwork study of cooking was used to estimate its environmental impact by measuring the energy consumption associated with the cooking process. These estimates were then employed to help people reflect on their environmental footprint and design ways to reduce such undesirable effects. 5 Based on videos of food consumption, researchers proposed a consumption life-cycle as a framework for identifying how and in which situations digital interventions could be useful in promoting changes in related behaviours. 54 Conducting a digital ethnographic study of people cooking together, researchers identified eight different formations in which people arranged themselves such as face-to-face, L-shaped and semicircular. 55 Through visits to and interviews in people's cooking spaces, it was found that the kitchen is a highly complex environment with a mix of fixed elements and flexibility. An example of a fixed element would be something like the physical structures of counters, while an example of flexibility might be the reorganisation of a spice shelf according to what people perceive as a useful classification scheme. 56 A number of studies exploring practices related to cooking provide valuable insights into FMCG interactions, though they do not focus specifically on the use of FMCGs themselves. After identifying the activities and places inside the households of two families in which FMCGs were used over a 24-hour period, researchers used the insights gathered to cite potential opportunities and challenges in the development of ubiquitous technologies. 57 Observing the various practices with food at home such as preserving, fermenting and pickling, researchers detailed the motivations, challenges and workarounds behind sustainability and associated practices. 58 Studying the routine of shopping for ingredients, researchers were able to identify the hidden methodologies employed by shoppers and consider how such methodologies may pose challenges for the design of proactive systems aimed at supporting the practice of grocery shopping. 59 Insights from observations were used to identify the requirements in developing kitchen utensils equipped with sensors and an infrastructure to monitor their use with the overall goal of measuring people's cooking competence. 60 Through an online survey on the frequency of use for 23 utensils, researchers measured their degree of usage and ascribed them a rank ranging from high to low. They did not find a strong relationship between the use of utensils and social and economic demographics. 61 Technological implementations in different stages of development, from prototypical to fully functional, allow researchers to explore their impact in a practical context. While a large number of technologies for the smart home have been designed to aid people in different aspects of cooking-social components, 62 counting calories 63 and cooking skills 64 -few have specifically focused on FMCGs. One implementation encourages people to scan their FMCGs upon their disposal and utilises computations to make predictions about their consumption. 65 Researchers demonstrated the challenges of making simple predictions and identified the contingencies that influenced them. An unpredictable vegetable box scheme, which delivered a randomised assortment of fruits and vegetables through a subscription service, was used by researchers to explore the consequences of delegating shopping to automated systems and how people accommodate their cooking habits to make use of food that was not purposely purchased. 66 Although prior work provides accounts of the cooking practice, it The proposed quantitative ethnographic approach allows for overcoming the challenges of analysing and giving a meaningful interpretation to vast quantities of data. Quantitative ethnography is a methodology that blends 'thin' descriptions-surface-level observations derived from numerical methods-and 'thick' descriptionsinterpretations that add contextual knowledge-to analyse data from fieldwork studies. 22 Quantitative ethnography seeks to bridge the gap between quantitative and qualitative methods. The context from ethnographic findings guides and provides a comprehensive interpretation of the statistical analysis, while the statistical analysis provides summaries and helps strengthen the relevance of ethnographic findings. The incorporation of quantitative methods allows one to manage and initially make sense of data and identify significant statistical patterns, such as distributions, outliers and correlations that emerge from data aggregates. This may not be identifiable from a purely qualitative analysis, which focuses on giving meaning to specific segments of the observed interactions. The resulting summary then grants us the opportunity to focus on specific details and incorporate knowledge gathered from the observed practice to achieve a qualitative and detailed description of the process. 68 The findings obtained from a combination of the ethnographic work are to be analysed using an analytical approach that is a method of breaking down a complex process into its parts. This approach allows one to attain an understanding of people's practices both through first-hand and immersive experiences, as well as through the application of an analytic perspective. The 'implications for design', 69 which can include concrete suggestions of features, design solutions and guidelines, can be incorporated into a wide range of design practices including product requirement specifications, use case modelling and the construction of mock-ups and prototypes. The proposed approach might seem to be unconventional from both an ethnographic and packaging design research standpoint. However, two considerations must be given. First, in the ethnographic approach, there is no set of rigid steps to follow; rather, the researcher has to formulate their approach to uncover what is seen but unnoticed 70 in everyday life. Second, methods for researching packaging design are often adaptations of others that have traditionally been used in a wider field. 71

| Study design
The study required participants to cook a meal from scratch in their household and permitted the researcher to record the participants' interactions. Recruitment was conducted through mailing lists, social media and referral. Participants were allowed to cook with a partner.
In these cases, the participant who contacted the researcher was considered as the primary participant, while any additional participants were considered as assistants. Participants were free to choose any meal so long as they had prior experience preparing it themselves.
There were no other restrictions placed on what time of day each session was to take place nor on its duration. The same researcher was designated to conduct all visits to the households. Participants gave verbal reassurance that the cameras were not disturbing them, as an example, P15 said 'No, you forget about them' in reference to the cameras. The recording started when participants began to retrieve ingredients and ended when they finished or served the meal. The researcher was present during the session, taking notes and engaging in conversation with the participants, so long as this was not considered a distraction for them. Lastly, a semi-structured interview was conducted to discuss the participants' cooking experience, after which they each received a £20 gift card as compensation.

| Data sample
The inclusion criteria for participation in the study were adult consumers of FMCGs between the ages of 18 and 80 and capable of preparing a meal by themselves. Twenty-three participants prepared meals in the study across 20 cooking sessions. Seventeen meals were prepared by a single participant, while only three sessions involved an assistant (p05, p09 and p18). The following demographic data correspond to the 20 primary participants as shown in Table 1

| Classification of data
The items were classified using a nested hierarchy consisting of three levels: type, category and item.

Type
Items were assigned to one of three types: FMCGs, utensils or environment items. The FMCG type consisted of all the products that meet the abovementioned characteristics for these goods (see Figure S1 for pictorial representations of FMCGs). Utensils were tools and devices that people were able to manipulate, and which were easily portable. Environment items were building structures, appliances and devices, which required an external energy source to function.

Category
Items of each type with similar characteristics and usage were grouped together. The categories of FMCGs were further informed by classifications of ingredients 72 and groceries. As an example, the category spices consisted of solid substances commonly packaged in a bottle and added to food to enhance its flavour.

Item
The objects were assigned a label under the name by which they are commonly known. For example, bottles of both fine and rock salt were labelled simply as 'salt'. Other characteristics such as properties of the product and its observed use were employed to differentiate items, which, albeit similar, are handled differently and cannot be easily considered as a replacement for one another. For example, a bottle of garlic granules and a bulb of garlic received distinct labels, dried garlic and garlic, respectively. If more than one item with the same label was used within the same session, each item received a second label to uniquely identify it. Either a numeric label or a descriptive one based on properties of the item was used for this second label.
The items each fell into categories, which could be defined by one of the 197 distinct labels. Out of those categories of items, 115 of them were FMCGs, 71 were utensils and 13 were environment items. The FMCGs were subdivided into 15 categories: baked goods, beverages, cleaning products, condiments, dairy products and eggs, dried goods, disposables and food storage, fruits, legumes, meats, oils and fats, rice and pasta, spices, stationery and vegetables. See Data S1 for a full list of items and Figure 3 for a visual representation of the hierarchy.
A note about the combination of FMCGs, when a given FMCG was mixed with another FMCG or group of FMCGs, the combination of FMCGs was labelled as food for practical reasons of data handling.
In this study, food is defined as the combination of two or more FMCGs such that the resulting amalgamation can no longer be said to exist as a group of distinct ingredients but rather as a combination of FMCGs as its components may no longer be separable.

| Data analysis
The basis of the analysis of this study consisted of manually capturing the interactions that the participants had with any item that they had contact with in their efforts to prepare a meal. An 'interaction' was considered as any instance in which an item was used either FMCGs that represented. An item was considered to be involved in a session if it had at least one interaction in that session.
• Interactions: counting the number of interactions participants had with FMCGs and other items while cooking. An item was considered to have an interaction each time that it was involved in an instance that met the aforementioned criteria for interaction.
• Phases: identifying the distinct periods within the cooking sessions in which interactions took place. Each session was divided into 10 periods of equal length, and interactions were assigned to their corresponding phases according to their start times.
• Conditionality: obtaining the conditional probabilities of one item being involved in a session given that another was involved. The probability of item A being used in a session given that item B was used in that same session is known as the 'conditional probability of A given B', denoted by P(AjB).

| FINDINGS
The main findings of each analysis method as applied to the three levels of the nested hierarchy-type, category and item-are described.
The analyses were not only focused on FMCGs but also integrated utensils and environment items to draw comparisons. The findings are divided into two sections. First, a quantitative section provides a summary of the statistical analyses. Second, a qualitative section provides an interpretation of the quantitative findings informed by the insights derived from the fieldwork and an analytical approach. Although the results presented in this section are already implicitly related to packaging, this relevance will be made explicit in Section 5.

| Involvement
Results showed that participants utilised a relatively small fraction of cheese, garlic and kitchen roll (see Figure 4).

F I G U R E 3 Hierarchical classification of selected fast-moving consumer goods (FMCGs)
The FMCGs that were most commonly involved were not necessarily those which were most available. The category with the greatest involvement was vegetables (30%), while the most commonly-available category found was spices (15%). Categories of FMCGs with a short life span had a greater rate of involvement in proportion to their availability; this was found in the categories vegetables (22% availability to 32% involvement), meats (4% availability to 7% involvement) and dairy products and eggs (6% availability to 9% involvement).

| Interactions
The results showed that

| Phases
The The distribution showed that the number of interactions across phases varies according to the category of FMCG (see Figure 6). The category vegetables was most commonly interacted with at the beginning with a peak of interactions before the second phase, spices was distributed along the session with the peak of interactions around the fifth phase and cleaning products had most of its interactions at the end with a peak of interaction around the eight phase.
The findings indicate that, in cooking, people had their first interactions with most of the FMCGs, which were needed for the meal at the beginning, then interacted with those products throughout the rest of the session. People kept retrieving FMCGs throughout the session, but at a diminished rate, and usually stopped retrieving FMCGs entirely by the last quarter. It might be that often more FMCGs were retrieved after serving the meal, but these were not recorded. For instance, people may have retrieved items from the condiments category such as balsamic vinegar and hot sauce when they sat down to eat. However, the methods in this study did not allow us to capture such items as they only recorded the interactions until the moment people served their meals. The findings also suggest that different categories of items interact at different phases in the cooking sessions.
For instance, items from the vegetables category were generally used at the beginning and required somewhat longer preparation times, as they often first had to be peeled and chopped for their eventual use in the recipe. This is in contrast to the categories of items, which were more commonly interacted with at the end of the sessions such as cleaning products, as they are commonly used to clean the mess produced by the previous categories of items.

| Conditionality
The concurrent analysis of FMCGs showed that FMCGs were utilised in recurring pairs with other FMCGs, utensils and environment items.  Of particular interest to this study is the use of ethnographic findings in the derivation of implications for design. 75 As stated by Crabtree et al., 76    The methods of this study are very time-consuming and labourintensive; the manual recording of interactions of a video took days of work. This can be made more efficient once the methods are automated, which seems plausible in the near future with the largescale introduction of RFID to FMCG packaging. For example, see the declaration of plan to introduce 100 billion electronic tags for products in convenience stores in Japan. 89 It should also be considered that although it is a the standard practice involves the use of video cameras to capture people's daily activities, and despite the fact most participants expressed that they were not disturbed by the presence of the cameras, still, it is a possibility that they could have been negatively influenced by their presence but did not express this to the researcher. Thus, the possibility that the participants modified their cooking process cannot be discarded. In future work, the use of less invasive methods should be explored. Furthermore, it should be considered that although the focus in this study was on drawing implications for designing products, which are either themselves or through their packaging embedded with technology, there are situations where the most appropriate solution is not to design any technological intervention at all. 90 The practice perspective argues that the introduction of new technologies is only one of many alternatives, which can be employed to bring change within a practice. 16 There is also need to reflect on the ways in which people already accomplish their goals as we already continually reconfigure spaces and technologies within them to meet particular demands. 91 The main contribution of this study is that it provides an understanding of the interactions of FMCGs in the practice of cooking.

| Implications for design
There are three specific contributions. First, it presents analysis methods to study the use of FMCGs in practice, which can be applied to a broader set of practices; for instance, by studying the use of Further research is needed to support the claim that a better understanding of interactions involving FMCGs can prove fruitful for the development of enhanced FMCGs. Undoubtedly, collaboration is essential for those personally invested and interested in FMCGs including designers, cooks and consumers. One future avenue discussed in this paper is the creation of design workshops in which actionable versions of the findings are provided to people so they can use them as a resource for design. A further step will be the creation of guidelines and prototyping tools for designing enhanced FMCGs.
The methods can also be applied to evaluate the potentiality and feasibility of proposed designs. The final step would be to develop enhanced FMCGs and deploy them in the field so that we might test their effectiveness and prepare them to be released onto the market.
Overall, given the crucial roles that FMCGs play in our everyday lives, it is necessary to create versions of them that best serve people's needs, and the proposed methods of this study and its findings can lead to such innovations.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in