Changes in food purchasing behaviour and the impact on diet and nutrition: 2021 to 2023
1 Introduction
This report provides information on trends in the purchasing of food and drink. It examines changes in purchasing behaviour and presents the impact on overall sales, and changes in type of food purchased and its nutrient content.
It analyses purchasing of food, drink and alcohol to be taken into the home and food and drink to be consumed out of the home. It compares the 12-week period to the beginning of September 2023 with the corresponding periods in 2022 and 2021. This coincided with price increases and a continuing recovery from the COVID-19 pandemic which impacted consumer behaviour.
It is important to note that it is difficult to isolate the impact of any one potential causal factor on changes in purchasing. The COVID-19 pandemic had a large impact on people’s purchasing habits, with people eating out of the home much less. The data in this report covers the COVID-19 recovery period where people were starting to eat out of the home a lot more. Comparable data prior to the COVID-19 pandemic are not available.
This report is the second part of a series of reports. The first, Consumer food purchasing and consumption: Latest trends in Food Price inflation, Part 1 - FA0504 was published by the Department for Environment, Food & Rural Affairs (DEFRA) and is concerned mainly with trends in buyer behaviour and not its nutritional impact. The living costs and food survey published by the Office for National Statistics (ONS) also looks at spending on food and drink but as part of the entirety of a household’s total spending and also does not directly consider the nutritional impact. The main focus of this new report is to look at the impact on nutrition and specifically how the purchasing of different nutrients has changed.
2 Main messages
Take home Purchasing
People’s behaviour changed in relation to buying food and drink to take into the home in 2023 compared with both 2021 and 2022. Partly in response to food and drink price inflation (25.9% increase compared with 2021, 13.1% increase compared with 2022) shoppers decreased their “category repertoire” (bought less variety of things), decreased their “repeat volume” (bought less of what they used to buy) and changed their “product mix” (changed from more expensive tiers of products to cheaper tiers). At the same time, total volume sales decreased by 7.3% on 2021. Volume sales of Drinks (including alcoholic drinks) and Proteins (including beans, pulses, fish, eggs, meat) had the biggest declines (down 9.9% and 10.2% respectively since 2021). Although volume sales of Starchy Carbohydrates (such as bread, pasta, rice) declined, there was a small increase in the proportion of sales made up by this category (0.6 percentage points since 2021). Some groups were impacted by food and drink price inflation more than others. Those who considered themselves struggling, in less affluent households (C2DE) and those households with young families saw the largest impact on their expenditure. This may be because relatively more of their expenditure is on categories which have seen higher price increases compared to more affluent households.
Out of Home Purchasing
The prevalence of eating out of the home also increased in 2023 (volume up 10.6% on 2022). Price inflation was evident in the out of home market, although “category repertoire” and “repeat volume” increased indicating people were eating out of the home more. This may be due to recovery from the COVID-19 pandemic. There were increases in the market share for snacks (up 1.3 percentage points on both 2021 and 2022) but decreases for main meals (down 1.7 percentage points on 2021, 0.6 percentage points on 2022). This may have been driven by a return of workers to the office following the pandemic, as the biggest absolute change in volume was seen in those working over 30 hours a week.
Impact on Nutrition
When looking at take home and out of home purchasing combined there was an increase in the amount of calories purchased (up 1.4% on 2022), saturated fat (up 3.1% on 2022) and salt (up 1.4% on 2022), while there was a decrease in the amount of sugar purchased (down 0.7%). The increases were driven by the out of home sector as they reduced in the take home sector. Comparisons with 2021 are not available for the out of home sector.
For the eating out of home nutrition analysis, it is possible to look at the “coping strategies” to determine the reasoning behind any increase or decrease in the specific nutrient. For all the nutrients looked at, the main driver for the increases was “repeat volume”, signifying that people were eating out of home more. When looking at take home and out of home combined, the increase in calories purchased (up 1.4% on 2022) was bigger than the increase in volume purchased (0.6%) indicating that people’s diets were more calorific. For the take home market, Discretionary Foods made up the biggest share of calories (24.8%), though Starchy Carbohydrates increased their share the most (up 0.4 percentage points on 2022).
For sugar, when looking at take home and out of home combined there was an overall reduction in sugar purchased (down 0.7% on 2022), although out of home increased by 16.4% and take home reduced by 2.7%. For take home, Discretionary Foods had the biggest share (43.3% in 2023), though the biggest reduction in share was seen in Drinks (down 1 percentage point on 2022). For saturated fat, across both sectors combined, the increase (up 3.1% on 2022) was greater than the increase in volume purchased (0.6%) indicating people’s diets contained proportionally more saturated fat. For the take home market, Dairy and Alternatives had the biggest share of saturated fat (34.9%), and this category increased its market share (up 1.1 percentage points on 2022). There was also an increase in salt for take home and out of home combined (1.4%) that was greater than the increase in volume (0.6%), indicating that people’s diets contained proportionally more salt. For the take home market, the Discretionary Foods category had the biggest share of salt (29.2%), though it’s share decreased (down 0.7 percentage points on 2022) while the share of the Miscellaneous category (including pizzas and ready meals) increased (up 0.5 percentage points on 2022).
High fat, sugar or salt (HFSS)
For food taken into the home, the proportion of sales in legislative categories considered HFSS fell between both 2021 and 2022 (down from 48.5% to 47.6%), and between 2022 and 2023 (down from 47.6% to 44.1%). This pattern was seen among most of the individual legislative categories. This is consistent with findings elsewhere in the report, with an overall reduction in sugar, and increases in fat and salt being driven by the eating out of home market.
3 Data
This report uses data from several different Kantar data sets. The analysis of take home data comes from the Grocery Purchase Take-Home Household Panel, while the eating out of home analysis comes from Kantar’s Out of Home Panel, which is a subset of the Grocery Purchase Take-Home Household Panel. Detailed methodology and more information on data sources is available in the methodology annex. Please note that the scope of this analysis is restricted to Great Britain (England, Scotland and Wales).
4 Results: Food, drink and alcohol taken into the home
There are several different factors which help explain how shoppers have changed their spending behaviours during this period. Table 1 details these different factors. The impact of these factors on take home food, drink and alcohol purchased is quantified in the figures within this report. Please note that the data used for this analysis is split into “Sectors” (for example “ambient bakery products”), “Categories” (for example “total bread” which is part of the sector “ambient bakery products”) and “Sub-categories” (for example “white bread” which is part of the category “total bread”). This provides an increasingly granular description of how different food and drink products are grouped together.
| Factor | Definition |
|---|---|
| Category repertoire | Changes in the amount of food, drink and alcohol (“volume”) purchased by sub-categories in the time period (from a list of 1,741 sub-categories within food, drink and alcohol) |
| Repeat volume | Changes in volume within sectors bought by each household in both time periods |
| Store mix | Changes in the amount of money spent (“spend”) by stores with different average prices (for example, switching from M&S to Aldi). |
| Brand/private label (PL) mix | Changes in spend by brands and supermarket own label (referred to as “private label” - given that private label is on average cheaper in a given category, such as switching from McVities digestives to Tesco own label digestives) |
| Promotion mix | The impact on spend from shoppers engaging more or less with promotions on food, drink and alcohol. A negative result means shoppers are saving money by using promotions more. Some shoppers may take advantage of buying more if an item is on promotion to save money, whereas others may wish to do this but can’t afford the additional cost of buying more on that shopping trip |
| Product mix | The impact on spend from other changes to product choice (such as moving spend from more expensive brands or supermarket own labels to cheaper ones, for example switching from Tesco “Finest” range to Tesco economy brands) |
| Inflation exposure | The amount of extra money shoppers spend on groceries due to price changes on identical items |
| Population size | The impact on spend from the changing number of overall households |
| Universe penetration | The impact on spend from shoppers entering or dropping out of the take home food, drink and alcohol market |
To help with the interpretation of the take home results, categories from the original data have been grouped into broader “food groups” in the context of a healthy balanced diet (see Table 2).
| Food group name | Types of sub-group |
|---|---|
| Bread, rice, pasta and other starchy carbohydrates | Bread, rice, pasta, noodles, cous-cous, breakfast cereals, potato products, morning goods |
| Fruit and vegetables | Fresh, frozen, tinned |
| Beans, pulses, fish, eggs, meat and other proteins | Fresh, frozen and canned meat and fish, canned and dried beans, peas and pulses, eggs and nuts |
| Dairy and alternatives | Milk, butter, cheese, yogurt, cream |
| Oils and spreads | Cooking oils, lard, suet, margarine |
| Discretionary foods | Cakes, snacks (sweet and savoury), biscuits, sweets, chocolate, ice cream, preserves and condiments |
| Drinks | Alcohol, soft drinks, tea, coffee, hot chocolate |
| Miscellaneous | Pizzas, ready meals, instant hot snacks, soups |
Figure 4.1 and Figure 4.2 quantify these factors. Figure 4.2 shows that in response to shoppers paying over 25% more for the same products than 2 years previously, they cut back on the amount of food, drink and alcohol purchased for take home and moved to cheaper items. This can be seen with the negative values for category repertoire (the variety of different sub-categories shoppers purchased from), repeat volumes (the amount of product a shopper purchased) and product mix (moving from premium or mainstream branded or supermarket own products to mainstream or value branded or supermarket own products). Shoppers also paid 13% more for the same products compared to 1 year previously but Figure 4.1 shows the amount they saved from changing the types and brands of products they bought was much smaller.
Figure 4.3, Figure 4.4 and Figure 4.5 shows the volume purchased in different food groups (defined as Kantar food and drink categories mapped to Eatwell categories) in the 12 weeks to September 2023, compared to the same period both 1 and 2 years previously. These charts show that the absolute volume for both the drinks group and the beans, pulses, fish, eggs and meat group have declined faster than the other food groups. This has resulted in them taking a smaller share of the total food and drink market, while other food groups such as bread, rice, pasta and other starchy carbohydrates, dairy and alternatives and fruit and vegetables have increased their share.
5 Results: Demographics
The data underpinning this analysis can be split into 3 broad areas of demographics:
- The social grade of a household (based on the job of the main wage earner in the household)
- The life stage of a household (based on the age and number of children within a household)
- How a household says they are coping with the increased cost of living
Figure 5.1 to Figure 5.6 show the impact of different factors for these demographic groups on the total spend on food, drink and alcohol taken into the home. Figure 5.1, Figure 5.2 and Figure 5.3 compares the 12 weeks to September 2023 with the same period a year previously. This shows that shoppers who said they were struggling due to the increase in food and drink prices, those from less affluent households and households with younger children were more affected by food and drink inflation than other households. This may be because relatively more of their purchases come from categories which have seen higher price increases compared to more affluent households. The main way in which shoppers from these groups compensated for these increased prices was to reduce the volume and variety of products purchased by more than other groups.
Figure 5.4, Figure 5.5 and Figure 5.6 presents the same comparison for 2 years previously. It shows that inflation exposure was a lot higher compared to 1 year previously, and that the same groups (those struggling, those from less affluent households and those households with younger children) were most impacted though the differences between most groups are small. The main way shoppers from those groups tried to compensate for this was by a change in the product mix, so switching to purchase standard or cheaper brands or supermarket own labels as opposed to more expensive brands and supermarket own labels (often referred to as “trading down”).
6 Results: Eating out of home
As with food taken into the home, there are a number of factors in the eating out of home sector that have affected the amount of money consumers have spent. These are broadly similar to those seen for take-home, though there are some differences. The factors for the eating out of home market are presented in Table 3.
| Factor | Description |
|---|---|
| Category repertoire | Measures changes in the amount of food and drink purchased (“volume”) driven by changes in the number of separate eating out of home categories of food and drink “sub-markets” (such as pizza or hot drinks) each individual purchased from in the time period (from a list of 223 sub-markets within the eating out of home sector) |
| Repeat volume | Measures changes in volume within the sub-markets bought by each individual in both analysed time periods |
| Channel mix | The impact on the amount of money spent (“spend”) from shoppers switching between different types of establishment “channel”, (such as coffee shops or quick service restaurants) with different average prices. Note: a movement needs to be for the same type of item, for example purchasing a burger from McDonald’s as opposed to a full-service restaurant |
| Store mix | The impact on spend from shoppers switching between businesses with different average prices within the same channel, for example from Pizza Express to Pizza Hut |
| Vouchers and promo mix | The impact on spend from shoppers engaging more or less with promotions or vouchers. In this instance, a negative result means shoppers are saving money by using promotions or vouchers more |
| Inflation and product mix | The net effect of price inflation on the same items, and the impact on spend from other tweaks to product choice (such as moving spend to a cheaper sandwich option). It is not possible to create an inflation figure for the total of the eating out of home sector due to the way the survey works (it doesn’t track specific products between years), although inflation for barcoded out of home items such as a can of coke purchased in a supermarket as part of a meal deal can be calculated |
| Population size | The impact on spend from the changing number of overall individuals |
| Universe penetration | The impact on spend from shoppers entering or dropping out of the eating out of home market altogether |
Figure 6.1 and Figure 6.2 show the impact of these different factors on consumer behaviour in relation to eating out of home in the 12 weeks to September 2023 compared to both 1 and 2 years previously. While both ‘category repertoire’ and ‘repeat volume’ have contributed to the increased spend on eating out of home (most likely due to recovery from COVID-19 pandemic), the effect of inflation and product mix is also very noticeable. While it is not possible to calculate a figure for the total out of home market (due to the way the survey works), for pre-packaged items with a barcode (e.g. bottles of soft drink or chocolate bars) the price inflation was 19.4% in the 12 weeks to September 2023 compared to the same period 2 years previously, and 11.5% compared to one year previously. To counteract this, shoppers made cheaper choices in terms of where they purchased their food.
Figure 6.3 to Figure 6.6 detail how different population groups are affecting the growth in eating out of home. Most of the absolute change is driven by those in full-time work, which may be due to workers returning to the office post COVID-19. However, there are some smaller groups whose purchasing within the out of home market has also increased by more than other groups, such as children/students and those not working.
Figure 6.7 to Figure 6.9 shows how the mix of products eaten out of home has changed. In the 12 weeks to September 2023, the volume of main meals increased compared to the same period both 1 and 2 years previously, but not as fast as other categories, so the volume share for main meals decreased. In comparison, volume sales for cheaper alternatives such as snacks and cold drinks increased more quickly which may be due to an increase in purchasing of “meal deals” in supermarket food to go sections as workers returned to offices.
7 Results: Nutrient analysis
The following section looks at how nutrients purchased has changed comparing the 12 weeks to September 2023 with the same time period both 1 year and 2 years previously. The nutrients considered in this section are calories, sugar (including free sugars for food, drink and alcohol taken into the home), salt, and saturated fat for both take home and out of home.
To provide some context, Figure 7.1 looks at the change in total volume of food, drink and alcohol purchased as well as purchases of different nutrients compared to both 1 and 2 years ago for food, drink and alcohol to be taken into the home. It shows that the volume of products purchased has decreased by 3.5% and 7.3% respectively compared to 1 and 2 years ago. Purchases of 3 of the 7 nutrients (salt, saturated fat, and fibre) have increased compared to a year ago. Protein shows almost no change, although there is a decrease compared to 2 years ago. The largest decrease can be seen in sugar, which is down 7.4% compared to 2 years previously, and down 2.8% compared to a year previously. Free sugars are also down 8.1% and 3.6% respectively.
Figure 7.2 looks at the percentage change in the amount of different nutrients sold for eating out of the home in the 12 weeks to September 2023 compared to the same period 1 year previously (2 year comparison is unavailable). It shows that the increase in purchases of sugar, calories and saturated fat is greater than the growth in sales overall. This may be due to the changes in the category mix as explained previously, with meals taking a smaller percentage of total out of home spend and snacks and drinks a larger share.
Figure 7.3 and Figure 7.4 show the change in overall spend and volume for the take home and out of home markets in 2023 compared with 2022.
The following sections look at each nutrient separately. They follow a similar format for each nutrient and start by looking at the overall impact on the take home and out of home combined before then looking at each sector separately.
As with the overall take home and eating out of home purchasing, there are a number of different factors which affect the amount consumers spent. Generally they are similar to those seen before (see Table 4).
| Factor | Description |
|---|---|
| Category repertoire | Measures changes in nutrients driven by changes in the number of separate sub-categories of food, drink and alcohol each household purchases during the time period (from a list of 1,741 sub-categories within the food, drink and alcohol dataset) |
| Repeat volume | Measures changes in nutrient within sectors bought by each household in both analysed time periods |
| Channel mix (out of home only) | The impact on sales of the nutrients from shoppers switching between the different type of establishments (channel) they purchase from. A movement needs to be for the same type of item such as hot drinks, cold drinks, snacks, quick meals, main meals |
| Store mix | The impact on the nutrients purchased from shoppers switching the store they purchase items from within the same category |
| Brand / private label mix (take home only) | The impact on nutrient purchased from shoppers moving their purchasing between brands and supermarket own label (private label, given that brands will have a different average nutrient content per 100gm compared to private label) |
| Vouchers and promo mix | The impact on nutrients purchased from shoppers engaging with promotions (recognising that items purchased on promotion will have a different average nutrient per 100g content to items purchased off promotion) |
| Nutrient profile and product mix | The impact on nutrients purchased from other changes to product choice (e.g. moving from mainstream to value private label, or from premium to mainstream brands, which will have different nutrient content to each other), plus the impact of changes to nutrient content per 100gm of like-for-like items |
| Population size | The impact on nutrients purchased from the changing number of overall households |
| Universe penetration | The impact on nutrients purchased from shoppers entering or dropping out of the analysed market altogether |
Purchasing of calories
Total calories purchased (take home plus out of home calories) were up 1.4% in the 12 weeks to September 2023 compared with the same period a year previously . This increase was greater than the increase in combined volume as a whole (+0.6%). This was mostly driven by an increase in out of home purchasing of calories; take home calories were relatively unchanged over this period (-0.1%, see Figure 7.5).
While the take home figure for calories showed little change, this masked a number of changes underneath the headline figure. For example, a reduction in volume (both in terms of repeat volume and also a reduction in the category repertoire) removed a large amount of calories from diets, however the product mix acted in the opposite direction. Calories from drinks declined the most, while the proportion of calories purchased from bread, rice, pasta and other starchy carbohydrates proportionally had the biggest increase in share of volume. In absolute terms, the discretionary foods group accounted for around a quarter of the reduction in the amount of calories purchased. These changes can be seen in Figure 7.8.
Figure 7.11, Figure 7.13 and Figure 7.15 shows the difference in the purchasing of calories to take into the home between different demographic groups. In the 12 weeks to September 2023, nearly all groups analysed saw a reduction in calories purchased compared to both 1 and 2 years previously. The only exception was households with young families (households where the oldest child is between 0 and 4 years of age), which saw an increase in calories purchased compared to the same period 1 year previously. Households with older dependents saw the biggest decrease in calories purchased in the 12 weeks to September 2023 compared to the same period both 1 and 2 years previously.
Demographic groups
Out of home
In the eating out of home market, purchasing of calories grew mainly because shoppers were buying more volume. Changing product mix (shoppers buying proportionally more snacks and drinks) and an increasing category repertoire (shoppers buying a wider variety of products) also contributed positively to this change (see Figure 7.17).
Purchasing of sugar
The total amount of sugar purchased for both being taken into the home and eating out of home was down 0.7% in the 12 weeks to September 2023 compared to the same period a year previously. In the same period, overall volumes were up 0.6%. There was a reduction in the volume of sugar purchased and taken into the home during this period, which was almost balanced in absolute terms by an increase in sugar purchased in the out of home market sector (to give a total reduction of 0.7%, see Figure 7.18). There was also a reduction in the amount of free sugar purchased and taken into the home in the same period (down 3.6%).
The volume of sugar purchased in the take home sector was down mainly as a result of a reduction in the total volume of food, drink and alcohol purchased. This is evident both in terms of shoppers buying from fewer categories, and less volume from those they did buy from (see Figure 7.19 and Figure 7.20). As with calories, Figure 7.22 shows that sugar purchased via drinks declined the fastest, with sugar coming from carbohydrates such as bread, rice, pasta and other starchy foods increasing as a proportion of all sugar purchased. Figure 7.25 shows the reduction in free sugars, with the biggest reductions seen in the discretionary foods category and the drinks category.
Figure 7.27 to Figure 7.32 shows the difference in the purchasing of sugar to take into the home between different demographic groups. In the 12 weeks to September 2023, all groups saw a reduction in sugar purchased compared to both 1 and 2 years previously. Households with older dependents saw the biggest decrease in sugar purchased in the 12 weeks to September 2023 compared with the same period both 1 and 2 years previously, a pattern that was replicated for free sugars (Figure 7.33 - Figure 7.36).
Total sugar
Free sugar
Total sugar by demographic groups
Free sugar by demographic groups
Out of home
In the out of home sector, the purchasing of sugar grew mainly because shoppers were buying more volume. The changing product mix (for example shoppers buying more snacks and drinks) created a 3% increase in sugar purchased, while an increased category repertoire (shoppers buying a wider range of food and drink) contributed a similar amount (2.9%). Channel and store mix also contributed as shoppers bought different items because of their channel and store choices (Figure 7.37).
Purchasing of saturated fat
The sales of saturated fat in both the take home and eating out of home sector increased by 3.1% in the 12 weeks to September 2023 compared to the same period a year previously. This increase was greater than the combined increase in total volume (0.6%) and was mostly driven by an increase in out of home purchasing of saturated fat (+16.8%). Take home saturated fat was also up (+1.3%) despite overall take home volumes declining (Figure 7.38), primarily driven by an increase in sales in the dairy and alternatives category.
The volume of take home saturated fat increased in the 12 weeks to September 2023 compared to the same period a year previously. Although shoppers generally cut back on volume (with volume reductions reducing the amount of saturated fat purchased by 4.3%), other factors acted to push the volume of take home saturated fat upwards. One example being changes in the take home product profiles and product mix adding 5% to the amount of saturated fat purchased (Figure 7.39).
Figure 7.41 shows that the discretionary foods and dairy and alternatives food groups accounted for the majority of saturated fat sold in the take home market. However, compared to the same period a year previously, the proportion of saturated fat sold from dairy and alternatives has increased.
Figure 7.44 to Figure 7.49 below shows the difference in the purchasing of saturated fat in the take home market between different demographic groups. In the 12 weeks to September 2023, all groups analysed saw a reduction in saturated fat purchased compared to 2 years previously. The majority of groups saw an increase in saturated fat purchased compared to the same period 1 year previously; the only exception was households in social grade D (households where the main wage earner is in a semi-skilled or unskilled occupation), which saw a decrease. Households with older dependents and households pre-family saw the biggest decrease in saturated fat purchased in the 12 weeks to September 2023 compared to the same period 2 years previously, while households with young families (where the oldest child is aged between 0 and 4 years) and those in social grade AB (where the main wage earner is in a higher and intermediate managerial, administrative and professional occupations) saw the biggest increase compared to the same period one year previously.
Demographic groups
Out of home
Out of home purchasing of saturated fat grew mainly because shoppers were buying more volume, although there was also a significant contribution from the changing product mix (for example, sweet and savoury snacks were among the fastest growing out of home items over this period, Figure 7.50).
Purchasing of salt
The total volume of salt purchased in the take home and out of home sectors increased 1.4% in the 12 weeks to September 2023 compared to the same period a year previously. This increase was greater than the combined overall increase in total volume (0.6%). Out of home salt purchasing increased by 7.0% while take home salt purchasing increased proportionately by a smaller amount (0.5%, Figure 7.51).
Though the take home volume for salt increased by 0.5%, this masked a number of changes underneath this headline figure. For example, a reduction in volume purchased removed significant amounts of salt from diets (the discretionary foods group contributing significantly to this). In contrast, the product mix acted to increase the amount of salt purchased (Figure 7.52 and Figure 7.53). Although the discretionary foods group are still important to take home salt (accounting for 29% of the nutrient), this group declined in importance compared to the same period both one and 2 years previously (Figure 7.55). The 3 food groups with the most salt (“discretionary foods”, “beans, pulses, fish, eggs and meat” and “bread, rice, pasta and other starchy foods”) all showed decreases in the 12 weeks to September 2023 compared to 2 years previously, and little change compared to 1 year previously.
All the demographic groups analysed showed a decrease in the amount of salt purchased in the 12 weeks to September 2023 compared to the same period 2 years previously. However, those households that were classified as “pre-family” (working households that do not have any children), “young family” (households where the oldest child is between 0 and 4 years of age) and “middle family” (households where the oldest child is between 5 and 9 years of age) showed an increase in salt purchased in the 12 weeks to September 2023 (Figure 7.57 - Figure 7.62).
Demographic groups
Out of home
Out of home purchasing of salt grew because shoppers were simply buying more volume. However, a change in product mix (as shoppers bought proportionally more sweet and savoury snacks instead of meals), and store mix (shoppers visited stores with lower average salt content for similar items) led to a reduction in the amount of salt purchased (Figure 7.63).
Food high in fat, salt and sugar (HFSS)
This section looks at food and drink that is considered to be high in fat, salt and sugar (HFSS), specifically in categories that are covered by legislation around promotions. A food product is considered to be HFSS if it has a nutrient profile model (NPM) score of 4 or greater (1 for drinks). Details of how NPM scores are calculated can be found in the methodology annex. Please note that this section will only consider food, drink and alcohol purchased to be taken into the home and not the out of home sector.
Figure 7.64 shows the total volume purchased in these categories as a percentage of total volume purchased overall, as well as how this differs for calories, sugar, saturated fat and salt. In the 12 weeks to September 2023, all the legislative categories combined accounted for 22.2% of the total of food, drink and alcohol purchased to be taken into the home. However, they also accounted for 34.2% of calories and 34.1% of saturated fat purchased to be taken into the home and 41.2% of sugar taken into the home. Although they each only account for less than 1.5% of the total volume of food and drink taken into the home, confectionery accounts for 10.1% and sweet biscuits 6.7% of the sugar taken into the home. Each of these categories also account for around 5% of the total calories taken into the home and a relatively high percentage of saturated fat (6.3% in confectionery, 7.3% in sweet biscuits). Ready meals account for 2.5% of food and drink purchased to be taken into the home, but 4.5% of salt, while crisps and savoury snacks account for less than 1% of volume taken into the home but 3.3% of salt.
Figure 7.65 shows the proportion of volume sold in the legislative categories that is considered to be HFSS in the 12 weeks to September 2023, and how this changed compared to the same period both 1 and 2 years previously. The proportion of volume sales in these categories that is HFSS has reduced from 48.5% in 2021 to 44.1% in 2023. Numerous categories have shown reduction across the 2 years, but the amount of food and drink purchased that is HFSS is falling most rapidly in product categories that contain pizzas (57.3% to 43.1%), soft drinks (40.1% down to 30.5%), crisps and savoury snacks (93.5% down to 88.4%) and breakfast cereals (31.6% down to 25.0%).
8 Conclusion
The findings of this report are consistent with other related reports and surveys (such as those released by DEFRA and ONS) by concluding that consumer behaviour has changed to counteract the impact of the increase in price of food and drink by, for example, buying less, moving from branded to supermarket own brand products and shopping at cheaper stores. However, this report was the first to look at the impact of this on the population’s diet and nutrition.
Purchasing data analysis shows that calories purchased increased by 1.4% in the 12 weeks to September 2023, compared with the same period one year previously. This was coupled with an increase in salt and saturated fat purchasing, and a small decrease in sugar purchasing when in home and out of home sales are considered together. However, the change was driven by the out of home sector.
The increase in calories, salt and saturated fat purchased are greater than the overall increase in volume purchased, suggesting that, in 2023, people’s diets contained more calories, and proportionally more fat and salt than one year previously.
It is not possible to quantify the specific contribution of food and drink price inflation to these dietary changes as other factors were operating at the same time such as behaviour changes due to recovery from the COVID-19 pandemic. In addition, comparable data for the periods prior to the pandemic are not currently available, so it is not possible to determine the extent to which this is a return to pre-pandemic behaviour.
9 Methodology annex
The report uses several different Kantar datasets which are listed below. These datasets cover Great Britain (GB i.e. England, Scotland and Wales). The sources used for specific pieces of analysis are highlighted.
Grocery Purchase Take-Home Household Panels (buyer behaviour)
The Grocery Purchase Take-Home Household panel for GB is made up of 30,000 households. The panel is designed to be demographically representative of the Great British population, such as by region, household size, presence of children, and age of main shopper. Data from the panel is demographically weighted to reflect the respective populations.
Panellist population targets are obtained from the Office for National Statistics (ONS). The Broadcasters Audience Research Board Establishment Survey (also known as BARB) is used to inform some of the more specific demographic splits such as life stage. Data from the panels is then used to make estimations about the GB populations’ shopping behaviours.
Recruitment to the panels occur through paid social channels, affiliate partners, parenthood clubs and direct volunteering. Panel members are paid a flat fee each week. Incentives are used to reward participation, typically as vouchers for high street retailers.
The panel is a continuous reporting panel, i.e. once someone joins, they will remain part of the panel for as long as they want to take part and meet Kantar panellist compliance requirements. There is a high level of panel continuity, with around three-quarters being a member for at least a year, and around two-thirds being a member for at least two years.
The panel reports on a continuous basis on all Fast Moving Consumer Goods (FMCG) purchases that are brought back into the home. FMCG is an industry term for products that sell quickly at relatively low cost, also called consumer packaged goods. The panel reports where items were purchased, what was purchased, how much was paid and whether the product was on promotion (e.g. multibuys or extra (volume) free). Once shopping is brought into the home, all FMCG product barcodes are scanned and the till receipt is submitted. This covers online and in-store purchases.
Nutrition data is collected directly from food labels by Kantar fieldworkers visiting stores on a rolling basis throughout the year. The frequency of data collection is every 4 months. The collection of nutrition data has also been supplemented with web scraping information from Tesco, Sainsburys, Asda and Iceland websites on a weekly basis, with the aim of improving the timeliness of nutrition information for manufacturer products sold in those stores, and own label products for those retailers.
Out-of-Home Panel (buyer behaviour)
The Out-of-Home Panel is a sub-set of the GB Grocery Purchase Take-Home Household Panel. It is made up of 7,500 individuals from GB who use a purpose-built app to input all the food and non-alcoholic drinks they have purchased for consumption out of the home (the products which never make it into the home), as well as takeaways and deliveries which are brought back into the home. This excludes alcohol.
Data from the Out-of-Home Panel is demographically weighted to reflect all individuals in GB. The Out-of-Home Panel is designed to be demographically representative of the Great British population, such as by region, age, gender and whether there are children in the household.
Pressure Groups
The Pressure Group survey has been completed by Kantar panellists across the GB Grocery Purchase Take-Home Household Panel and results in panellists being classified into one of three different segments: Comfortable, Managing or Struggling.
Panellists were asked “Which of the following best describes your household financial situation?”:
The Comfortable are those who answered “I don’t have to restrict my spending in any way” or “I’m comfortable & can afford most things”.
The Managing are those who answered “On the whole I can manage ok”.
The Struggling are those who answered “I’m just making ends meet” or “It’s not sufficient to cover expenses”.
Panellists may change their answer to this question each time that they are asked causing the number of people in each pressure group to change over time. The data in this report is based on answers that were given in January 2023. In the January 2023 wave, the survey was answered by 12,154 respondents in GB. Figure 9.1 shows how the proportions of GB buyers in each group has changed between the first survey in May 2020 and January 2023. Over that time, it broadly shows a decrease in the proportion of shoppers identifying as Comfortable and an increase in customers identifying as Struggling.
Inflation methodology
The Kantar inflation measure included in this report includes food, non-alcoholic drinks and alcoholic drinks.
Kantar’s measure of inflation is based on over 75,000 Food, Drink and Alcohol products compared year-on-year. Price data is observed via the till receipts shared with Kantar by panellists (which is a mandatory part of the panel task).
Individual products are weighted in the inflation measure according to their importance in each store (based on the percentage spend in £s). Weights are determined at the individual product level and vary every period.
Kantar takes an average of each product’s importance across the two time periods being analysed. For example, when looking at the inflation figure for the 12 week period to September 2023, the price change of each product is calculated using the 12 week period to September 2023 as year 2 and the 12 week period to September 2022 as year 1, and the weight of the product is the average sales of that product across the two 12 week periods.
For a product to be included in the calculation, it must be present in both of the analysed time periods in the retailer. Such products typically account for around 80% of market spend. The inflation measure does not include products which were new or removed and these products may have different characteristics to the products that are present in both time periods.
Products are also excluded if they show an extreme price change year-on-year, (e.g. more than a 400% increase or more than 80% decrease), as such changes are outside what is deemed ‘normal’ and may be incorrect movements that would distort the reported figure. The calculations are based on final average price which will include the impact of any price promotions in place at the time.
This calculation is used. Y represents the year.
Methodological limitations for Kantar’s data collection
All forms of data collection are subject to certain limitations. The following are worth noting in the context of this report:
There is a minimum home technology ownership and internet access requirement to be part of the GB panels. This means the purchasing and consumption of groups who do not meet these minimum requirements are less well represented within the findings.
Buyer behaviour panels reflect what the shopper has both chosen and been able to buy. If a product is not available, the panellist is not able to buy it.
The buyer behaviour panels do not include food and drink that has been obtained at no cost to the panellist (e.g. from food banks or gifted). Data relating to food and drink acquired through these sources is therefore not included in this analysis.
There is a degree of uncertainty present in the results calculated but there are no confidence intervals associated with the estimates which means that the statistical significance of the changes over time or differences between categories cannot be assessed. Kantar Worldpanel do not calculate confidence intervals or carry out statistical significance tests as the data are collected from a panel and not from a simple random sample. Please also note that Kantar target specific socio-demographic groups and then weight the data so it is representative of the population. Therefore, statistical significance testing on these results would not be valid, however large changes over time or large differences between groups are more likely to be statistically significant than smaller changes and differences.
For the collection of take home nutrition data, Kantar FMCG’s fieldworkers enter stores to collect nutrition information on a rolling 4-month basis but this does not update all products in the dataset each time. If no nutrition information is found for a product during a reporting period then the most recently collected information is used. The webscraping from four retailers mentioned earlier helps to reduce the use of older data.
Various factors influence representation differences across socio-demographic groups. For example, as is the case in any form of panels, incentivisation levels to perform the data collection tasks are set to maximise compliance and minimise the influence on purchasing behaviour; these may be less appealing to those on higher incomes, therefore we could have a relative over-representation in lower-income households and an under-representation in high-income households.
Markets are measured in different volume measures. Therefore, for comparability, packs are used as the basic unit unless specified. The packs measure does not take pack size into account and therefore a limitation is that a change in pack volume may not be fully reflective of the change in volume within packs (in Kgs/Litres) purchased.
Additional points to note
Population targets are updated annually in January. This means changes in the population are accounted for throughout the analysis.
Rounding of numbers will mean that percentage figures shown in graphs may add up to more or less than 100%.
Time periods
Kantar data reporting periods cover four weeks, and these four week periods are normally grouped to form longer time periods such as 12 or 52 weeks. Some findings cover 52 weeks and this is referred to for ease of reading as years. The 12 week analysis does not equate to calendar or financial quarters.
The duration of the time period is as stated for each analysis.
Data included in the calculations covers January 2021 to March 2023.
The findings are reported in 12 week periods. This is because inflation and buyer behaviour is a fast-moving topic and using data at this 12 week level allows observation of these fast moving trends. Some commentary refers to change in comparison with September 2021, as this is the period from which Food, Drink and Alcohol price inflation began to rise. Comparisons to September 2022 are also included.