Online grocery shopping (OGS) as scripted behavior: insights on shopping routines of households with children

The Covid-19 pandemic impelled many consumers to shift from shopping for groceries in-store to shopping for them online (Eikelmann et al., 2022). For some, online grocery shopping (OGS) became a way to cope with the crisis (Fuentes et al., 2022), whereas many have found OGS to be convenient (Driediger and Bhatiasevi, 2019; Verhoef and Langerak, 2001), time saving (Anesbury et al., 2016; Frank and Peschel, 2020; Stenius and Eriksson, 2023) and helpful for meal planning (Berg and Henriksson, 2020; Brand et al., 2020). In addition, many other digital food services are increasing in popularity (McKinsey, 2023), suggesting that the overall digitalization of the food access landscape is slowly but fundamentally changing the way consumers acquire food (Granheim et al., 2022). Digital platforms enable consumers to compile their grocery baskets to meet various aims, such as optimizing food supplies and personalizing meal planning, including health- and sustainability related goals (Eikelmann et al., 2022), which are important overarching priorities in the food industry today (Demartini et al., 2018). The needed investments, however, pose considerable risks to retailers (Thomas-Francois et al., 2023), and therefore, understanding how consumers use these services is crucial.

The transition from in-store to online grocery shopping is demanding for many consumers (Wang and Somogyi, 2018), requiring considerable readjustment (Fuentes et al., 2022; Hansson et al., 2022), which may explain why OGS, despite important benefits, still accounts for only a narrow slice of the total grocery sales, even in the more advanced markets, such as the UK (Eikelmann et al., 2022). Grocery shopping is deeply habitual (Stenius and Eriksson, 2024) and embedded in other household practices (Dyen et al., 2018), and therefore resistant to change. Adopting new food-related routines requires work, and any new routines must be fitted into the existing ones (Dyen et al., 2018; Fuentes and Samsioe, 2021). How new routines emerge and reconfigure old ones, is not well understood (Samsioe and Fuentes, 2022). Therefore, to understand OGS and fuel its development and market growth, it is important to investigate those who have made this transition successfully and adopted OGS as their regular way to acquire groceries (Eriksson and Stenius, 2024; Stenius and Eriksson, 2023).

To-date we know very little about how consumers conduct their OGS. Few studies (if any) have specifically focused on the actual routines, the sequences of actions that typically constitute routinized OGS. Using the lens of practice theory, Samsioe and Fuentes (2022) investigated how the use of certain digital food platforms (excl. OGS) become routinized, without, however, exploring the actual routines. In an ethnographic interview study, Heidenstrøm and Hebrok (2021), investigated how online shoppers navigate between the platform and their kitchen. Their aim was, however, to understand food waste, not shopping. We are not aware of any studies that map the actions that households take when shopping for groceries online as a regular practice. This study fills this gap. We describe grocery shopping as a necessary, routinized activity embedded in other household routines, including steps that are invisible but potentially consequential for retailers. By understanding how households conduct OGS in their everyday, we can generate important new knowledge for retailers on how to develop their services to support the routines of these households.

Specifically, this study focuses on households with children, a key customer segment for grocery retailers, and aims to (1) map and organize all actions involved when shopping for groceries online, and develop a comprehensive general script for OGS. Furthermore, the study seeks to (2) identify typical OGS routines. The research questions are as follows:

The article is structured as follows. The literature review summarizes prior relevant research and introduces script theory. Methods for data collection and analysis are described next, whereafter the results are presented and explained. The findings and their implications are then discussed.

As part of the broad digitalization underway in the food industry, encompassing everything from production to distribution (Demartini et al., 2018), digital services for consumers are transforming how we buy groceries but also how meals are planned, food supplies optimized and other food-related priorities addressed, such as health-conscious and sustainable consumption (Eikelmann et al., 2022). Prior research into OGS has established that perceived benefits, such as time saving (Frank and Peschel, 2020) and convenience (Driediger and Bhatiasevi, 2019) predict adoption, whereas perceived risks and costs deter from it (Brand et al., 2020). Studies have further pinpointed certain customer groups as more likely adopters of OGS (Brand et al., 2020; Eriksson and Stenius, 2022, 2024; Van Droogenbroeck and Van Hove, 2017).

An overlooked aspect of grocery shopping is that it is repetitive, regular routinized behavior (Stenius and Eriksson, 2024; Thomas and Garland, 2004). We suggest that a closer look at how households routinely plan and execute their food-related chores can relevantly extend present understanding of OGS.

A routine refers to “a type of higher-order habit that involves sequencing and combining processes, procedures, steps, or occupations” (Clark, 2000), to achieve a goal in the everyday life (Ong, 2006), sometimes by multiple actors in cooperation (Cohen and Bacdayan, 1994). Characterizing grocery shopping quite accurately. It typically involves a chain of familiar steps (e.g. meal planning, item listing, shopping, paying), some more automatic than others, sometimes involving several individuals, with the aim of feeding a household.

Some empirical studies have investigated the transition from in-store to online grocery shopping (Fuentes et al., 2022; Hansson et al., 2022), or to another digital food/meal service (Berg and Henriksson, 2020; Fuentes and Samsioe, 2021; Hertz and Halkier, 2017; Samsioe and Fuentes, 2022) using the lens of practice theory (Reckwitz, 2002), and concluded that it is often demanding. Prior research suggests further that routinizing new shopping practices requires the right combination of meanings, competencies, materialities and temporal patterns (e.g. Hansson et al., 2022; Samsioe and Fuentes, 2022). (see Table 1).

Few studies have, however, shed light on what the new routines look like when OGS has been successfully adopted, and how they interconnect with other household routines. An exception is that by Heidenstrøm and Hebrok (2021) who explored how shoppers combine ordering online and taking stock of their pantry/fridge supplies, but with focus on food waste, not shopping routines. OGS as a chain of actions has not, to our knowledge, been investigated. An alternative theoretical lens suitable for this aim is presented next.

Shopping for groceries usually consists of a sequential pattern where familiar steps follow each other in a relatively stable order (Dyen et al., 2018) suggesting that we follow a script, a mental representation of “an event sequence, such as paying at the checkout of a supermarket, incorporating assumptions about the characteristic actors, objects, and actions, in a prototypical instance of the event” (Colman, 2015). Pieters et al. (1995) suggest that scripts contain the action rules and serve as blueprint to a behavior in a specific situation, whereby routines manifest scripted behavior, overlearned through repetition (Schank and Abelson, 1975).

A script always contains the same actors, the props involved and the principal actions needed to accomplish an activity in a specific temporal or causal order (Schank and Abelson, 1975). The action sequences are further grouped into larger segments and organized hierarchically (e.g. Bozinoff and Roth, 1983; Smith and Houston, 1986 in Erasmus et al., 2002). For instance, the script of OGS includes the household members involved, the props (e.g. pantry, fridge, computer) and the script segments, such as meal planning, order placement, goods receipt. Each segment then includes action steps, representing the lowest level of the hierarchy (Erasmus et al., 2002).

Those advocating the use of script theory in consumer research argue that individuals can reliably recall a familiar routine and thus: eliciting scripts can shed light on how an activity is typically enacted (Erasmus et al., 2002; Ong, 2006). The improved understanding can help retailers offer more relevant services, benefiting thereby both consumers and retailers (Erasmus et al., 2002). Consequently, in this study scripts are elicited from regular online shoppers to identify and organize all actions involved in OGS, and to shed light on the hierarchical and sequential patterns on how households with children typically shop for groceries online.

A qualitative design was deemed appropriate for the study of underinvestigated household activities. For the data analysis we adopted an essentialist approach (Braun and Clarke, 2006; Vaismoradi et al., 2013) that is we assume the respondents to convey reliable accounts of their OGS routines. This concurs with Erasmus et al. (2002), who recommend script elicitation for mapping routinized consumer activities.

The data was collected through 14 semi-structured deep interviews in Finland, in 2023 (Jan–Mar). Purposive sampling using pre-defined selection criteria was applied: participants had to represent households with children under 18 (an important shopper group); be from urban areas (access to full-basket OGS); minimum one year of OGS (habits take 18–254 days to develop; Lally et al., 2010); shops at least twice a month (regularity).

The interviewees were recruited through social media, the networks of the research group and referrals from selected interviewees. For demographics and other characteristics of the participants, see Table 2. Most households had two adults and two children. Ten households had shopped for groceries online regularly for at least three years and nine households shop at least weekly. All households acquired their groceries mainly from one of the two dominant grocery chains in Finland (S-/K-chains). The sample size was deemed sufficient as data saturation became evident.

The semi-structured interviews were conducted online in Zoom^TM or Microsoft Teams^TM in early 2023 (Jan–Mar), with one adult from each household, with duration from 36 to 73 min, and supported by an interview guide (see  Appendix 1), developed with the help of the one used by Ong (2006).

Before the interview, the participants were informed about the purpose of the study, and their rights, in accordance with the prevailing ethical guidelines for research with human participants (https://tenk.fi/en/tenk). The participants were then asked about the last time they shopped for groceries online, as asking about a specific incident helps generate a reliable account of what actually happened (Schwarz and Oyserman, 2001). Furthermore, the respondents were at their homes, which is for many the context where they conduct OGS, and therefore a condition that can help recall the event accurately (Neisser and Hyman, 2000). Probes and additional questions were used to elicit detailed information on the routines (Magaldi and Berler, 2020).

The interviews were recorded, transcribed and coded by the researcher using MaxQDA^TM, a software designed for analyzing qualitative data (no AI features used). The data analysis was performed in two parts. To respond to RQ1, all actions, and their sequences and hierarchies, were first identified from each individual interview, using thematic analysis both deductively and inductively (Braun and Clarke, 2006; Cassell and Symon, 2004). This was followed by grouping and organizing these actions into higher level script segments, which was used to design a comprehensive general script for OGS.

To respond to RQ2 (identifying typical OGS routines), an illustrative script visualizing the actions, the hierarchy and action sequences, was drawn for each household, using the comprehensive general script as a guide. When outlining these scripts, similarities and differences in the actions, sequences of actions and their order within the script segments, were identified. The scripts were then compared with each other. Similarities and differences between scripts were observed and a more systematic comparison was performed to see if the scripts could be reasonably grouped to suggest typical routines.

The study was conducted in Finland, where the grocery market is characterized by two dominant grocery chains, the S and K chains, with 82.6% of the market (NielsenIQ, 2024). They are also the main providers of full-basket OGS, even if many other digital food services are available, such as quick commerce, mealbox and surplus food redistribution services. The market is quite small compared to leading European markets (Eikelmann et al., 2022).

The analysis established that the main steps taken by all respondents were: planning, item picking, order placing and goods reception either through pick-up or delivery. These are labelled as script segments (Erasmus et al., 2002), each consisting of a set of actions.

Planning consists mainly of meal planning including actions such as thinking of what groceries are needed, checking the week’s schedule for household members, looking at recipes, checking ingredients at home, listing meals and items needed and/or copying and updating a previous grocery shopping list, or meal plan in an external app, albeit not necessarily in this order. Booking the time slot of delivery or pick-up is included in planning, even if some may book it in connection with item picking, or when they are about to place the order.

Item picking consists of opening the online store and selecting items, which can include several actions such as copying a previous order or list, deleting or adding items from/to that list, picking other needed items, choosing items based on written lists, choosing frequently used or recommended items, searching for items and checking offers.

Order placement consists of checking the order, sending the order and entering/confirming the payment details. In OGS planning is not necessarily completed before item picking starts, or it may even start after some initial picking has taken place. Even if planning and picking items are two different activities, buying groceries online is flexible and allows for quick changes. The shopper might have added a long list of items to basket but decides then to look at a recipe and redo some or all planning, and subsequent item picking. Even if the order has been placed, it can be complemented, or the shopper might even return to the planning phase.

Goods delivery or pick-up constitutes the last segment in the OGS script. It consists of a notification from the retailer with the receipt, and a notification of delivery or pick-up, depending on the chosen method. Pick-up consists of driving to the store, picking up the goods, loading them in the car and driving home. Delivery consists of receiving the goods at the door and of payment (if not paid in advance). After the goods receipt, the items are unpacked and stored, and the cardboard boxes stored/taken for recycling.

For the complete general script, aggregating all actions and highlighting the sequential and hierarchical structure of OGS, see  Appendix 2.

A comparison of the individual scripts suggests that all scripts fall into two main groups, or align with one of two typical routines: impromptu shopper and planner routines, each representing a group of households with some within-group variation. The main differences were in the amount and style of planning, the timing of planning and picking items and order frequency. For details of the analysis see  Appendix 3.

Impromptu shoppers typically conduct their OGS in a single session meaning that planning, item picking and order placement are done concurrently. These households use little time for planning, and they typically wish to have the order completed quickly.

The routine of household R8 exemplifies how impromptu shoppers conduct their OGS (Figure 1). This household usually shops once a week, for 3–4 days per order, making sure there is enough food for weekdays. On weekends they shop more in-store. The order is often placed on Sundays with goods delivered on Mondays.

In this household one adult is mainly responsible for OGS. The routine starts with booking a time slot for delivery by copying/placing the previous order. On the same day, usually in the evening, the household starts planning and picking/deleting items, typically in the kitchen. They plan on the go, check existing ingredients and pick the needed ones concurrently. When all items are picked, and any additional wishes added, the order is placed, and the delivery received the following day.

Most households with an impromptu routine shop once a week for food needs ranging from 3 to 4 days to one week. The order is placed one day before the delivery or pick-up. Curiously, several households ordered on Sundays with delivery or pick-up day on Mondays.

Even if these households only do light planning, about half of them use a shopping list to support their OGS. Some prepare by booking a time slot for the pick-up or delivery in advance, or by listing needed items during the week. One adult was typically more responsible for OGS.

Two households differed from the others. One used the online store mainly for bulk shopping, whereas the other booked a weekly time slot and added items in an improvised fashion during the week, and only planned, if necessary, just before placing the order.

Compared to impromptu shoppers there was more within-group variation among planners but structured planning was an essential part of the OGS routine for all. Meal planning is often an element of the overall weekly planning, where the households consider upcoming activities and their impact on the meal plan. The activity consists of anticipating and preparing for the food needs for the upcoming week, including checking the weekly schedule. For example, plans to eat out, hobbies and time constraints for cooking affect the OGS routine, along with unexpected changes, such as visiting friends or eating out.

Most of these households conduct planning and item picking in several sessions before the final order is placed, but with some differences in how they plan. Some households plan nearly everything first and then pick the items, whereas others plan and pick simultaneously. To exemplify this, the routines of two households are described. The first is a household (R12) that plans extensively but finalizes planning before item picking (Figure 2). This household usually shops once a week for the whole week, and starts by booking a time slot for the pick-up. Planning is ongoing, involving creating a meal plan and a shopping list, usually in several sessions. The item picking, in turn, is often done in a single session after the meal plan and shopping list have been finalized. Typically, the order is checked once more the day before delivery to ensure all needed items are included.

Four other households in this group also typically plan first and then pick the items. Two households conduct all steps in one session: they make a meal plan and a shopping list, and immediately thereafter pick the items and place the order.

The other planners, albeit only two households in the sample, plan and pick simultaneously but complete the order gradually over time. To exemplify, one of them will be described (R7) (Figure 3). This household usually shops for a two-week period, which involves extensive planning. The order cycle is between 13 and 15 days. The grocery shopping begins by creating a meal plan a few days after the previous delivery. The meal plan consists of a list of days and meals for the upcoming two-week period. Meal planning and item picking happen simultaneously, usually so that a meal is planned, and then the items for that meal are picked. Another meal and the items for that meal might be picked in the next session. Thus, planning is completed gradually over several days until the final order is placed a couple of days before pick-up. The order is checked once more the evening before pick-up and necessary changes made.

This study investigated OGS routines of an important customer group, households with children (Stenius et al., 2023). To answer RQ1, the data analysis generated a comprehensive general OGS script, consisting of four script segments: planning, picking items, order placement and goods reception, and mapped all actions included in each segment (see  Appendix 2). To our knowledge this type of research has not been conducted before.

For retailers, the general script highlights all possible steps of OGS, including actions related to planning, which are not visible through online store data, but may nevertheless be consequential for them. It also provided a tool for analyzing the individual scripts of each household, making it possible to detect differences and similarities in household routines, which helped respond to RQ2, that is identify typical OGS routines.

Two such OGS routines for households with children were found: that of impromptu shoppers and that of planners. The impromptu shoppers did at best light planning, usually concurrently while picking items, in a single session. Planners, in turn, had a structured planning routine involving the use of meal and/or shopping lists, and they invested a considerable effort in planning in several sessions before the order was placed (see Figures 2 and 3). We are not aware of prior studies that have attempted to detect typical OGS routines.

The study shows that OGS suits many different household styles and routines. The actions performed are quite similar across households, but their order and timing differ as every household must be able to embed their shopping in their other household routines, in line with prior research (Dyen et al., 2018; Fuentes and Samsioe, 2021). The ability to shop for groceries anytime from anywhere, and over an extended time period with pausing and continuing later, is an important feature and affordance of OGS, and one that helps adjust the shopping routine to the schedules, life situation and preferences of many different households. This feature also differentiates full-basket OGS from in-store shopping.

The study highlights the key role of planning for many households with children. Some plan little; others plan extensively, but nonetheless they all plan. Some have a routine that resembles that of in-store shopping: they plan offline and only log in to select the items and place the order. Others have learned to integrate their shopping and meal planning, in line with Stenius and Eriksson (2023), and they complement the unfinished shopping list over several days. An important observation is that meal planning is often a component in the overall weekly planning of a household, and the OGS routine reflects this.

We argue that research and theory development would benefit from accurately describing grocery shopping as a necessary, regular and routinized activity embedded in other household activities. Using script theory as a lens we mapped the sequential and hierarchical patterns of OGS routines, and contribute with a comprehensive general script of OGS, containing all steps including ones that are invisible, but potentially consequential, for retailers. In addition, we showed that typical OGS routines could be detected, which further expands our understanding of OGS. Future research and theory development can utilize our work and develop the ideas further. One avenue could be to explore all food-related household routines further, or focus more on specific drivers of food-related activities (e.g. health-related/sustainability concerns), and investigate how these are reflected in shopping channel choices and routines.

For practitioners the study suggests that OGS is an opportunity to add flexibility to the lives of many different types of time-constrained households with children, an attractive target group for grocery retailers (Stenius et al., 2023). How this group is served online can have long term implications as these, oftentimes young families, are only establishing their grocery shopping habits, routines and loyalties. By tracing the steps contained in the OGS routines of these households, the study revealed that the routine often begins before entering the online store, highlighting the role of planning and differences in the way households plan. Especially for those engaging in highly structured meal planning, OGS is not an isolated activity but a component in the overall weekly plans. Retailers should support this phase as it seems to be central for sustaining the routine, and presently many shoppers do their planning in another application. If retailers’ script of grocery shopping focuses on the steps taken in the store only, the digital affordances might go underutilized.

This study, in line with Stenius and Eriksson (2023), suggests that customers could greatly benefit from proactive help, especially in meal planning. Retailers have abundant data on the preferences and habits of households, and thus making suggestions, even full-scale meal plans, should be possible. Alternatively, easy AI facilitated planning tools could be an integrated feature of the online store. Such features could be attractive for planners but also impromptu shoppers, who want to shop fast. Prior evidence indicates that shoppers welcome easy choices and features that speed up shopping (Anesbury et al., 2016). Households that do extensive weekly planning, starting in the weekly activities, could appreciate an integrated meal planner that would help combine meal planning and shopping activities. With an automated move-to-basket feature it would further integrate the OGS process, and speed it up.

Furthermore, groceries are bought in regular cycles, which can be detected and monitored by retailers to time assistance and marketing efforts. Any major deviations from the usual routines can be spotted and reacted to, for instance by sending personalized reminders or ready meal plan suggestions to make OGS an easy routine to fall back to.

In addition to a comprehensive general script for OGS for households with children, this study identified two typical routines for OGS, and highlighted the significance of planning in the routines. A number of ideas for how retailers could develop the service supporting the integration of planning and shopping were proposed. The digital shopping environment, plausibly even more so with the help of AI, provides ample opportunities to design attractive, value-adding services to support the OGS routines of households with children.

Because of the small sample, conclusions about generalizability should be drawn with caution. Nevertheless, two distinct routines were identified, and relevant information on how households with children conduct their OGS was generated. A larger sample could have facilitated the identification of typical routines, which now required several iterations. A more refined set of typical routines, especially among planners, can plausibly be identified in future research. To what extent the findings reflect the market studied (Finland) is difficult to determine.

The data collection and the main analysis were performed by one person but kept transparent to the other researchers throughout the process. The results were perused, discussed and modified jointly by two researchers, in many iterations. We are confident that the final presentation is appropriately supported by the evidence.

Coincidentally, most respondents in this study were women. The respondents were chosen based on who in the household had the responsibility for OGS. Especially among planners, the adults typically shared the responsibility, and the choice of the respondent was a happenstance. Considering the aims of the study, respondent’s gender is unlikely to play any role.

To deepen the understanding of routinized OGS, more studies that look into what people actually do, are needed. Researchers in other countries are encouraged to utilize and expand upon our contributions, and investigate OGS routines in their markets to understand them better. The planning phase is particularly relevant, as it is so important for households with children.

1. Age

2. Gender

3. Family size (amount of children and age)

4. Occupation (working full-time, part-time, studying, unemployed)

   • Work from home/office/hybrid?

5. Urban or rural living?

6. How long been shopping groceries online, and how often?

1. Tell me about the last time you shopped for groceries?

   • Is that typical for you? Why do you think that?

2. How did you go about shopping the groceries?

3. How did you plan the shopping?

4. How did you do the meal planning?

5. How did you go about choosing where to shop?

6. What type of platforms do you use for shopping food online and how?

7. When did you do the online grocery shopping? In what situation?

8. How did you go about choosing the products?

9. How does your family go about shopping groceries in a typical week? Other forms of food provisioning?

   • What is typical from week to week? Why do you think that is?

   • Does this change?

10. How do you think the online shopping might have affected your family? Can you think of other ways? Would you explain more?

11. Do you see any other implications of shopping groceries online?

    • What positive implications?

    • What negative implications?

12. What are your future expectations for online grocery shopping services?

Source(s): Authors’ own work