Benchmarking for continuous improvement in food retail environments: a systematic scoping review

Abstract

Background

Benchmarking, closely related to audit and feedback processes, is an implementation strategy that continuously measures performance against defined functions and metrics and implements change accordingly. Benchmarking is increasingly being applied in public health nutrition to improve food environments for health. However, the process and mechanisms by which benchmarking interventions can drive continuous improvement in food retail settings is not well understood. We performed a systematic scoping review to identify existing food retail benchmarking initiatives, compare against an evidence-informed six-step benchmarking approach, and explore qualitative insights including underlying theories and enablers of benchmarking methodologies from existing research.

Methods

We searched five databases for studies published up to 2024, using search terms related to food retail settings and benchmarking.

Results

Analysis of 46 papers found that all studies identified functions and metrics to benchmark against, while only one paper incorporated benchmarking as part of a continuous improvement cycle with repeated assessments. Approximately one-third of studies developed recommendations for retailers, with only eight studies describing feedback of results to retailers. Included studies originated from the fields of retail management (67%), public health nutrition (24%), environmental sustainability (7%) and one from the field of consumer affairs. Very few studies articulated a theory of change or logic model for their interventions. Enablers of effective benchmarking articulated within studies were found to be engagement from food retailers, access to food retail data, flexible and adaptable benchmarking tools and strategic communication of findings to retailers.

Conclusions

Our findings suggest significant opportunities to strengthen benchmarking in the food retail setting as part of continuous improvement with greater emphasis on integrating theory as well as effective communication of feedback to retailers. This evidence can inform public health approaches to transform food retail environments for improved population nutrition.

Trial registration

Not available.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26725-6.

Background

Benchmarking has been applied in healthcare service delivery using indicators or targets to drive performance change through quality improvement of outcome measures [1–4]. However, there is increasing attention on the use of benchmarking within food retail environments to improve healthy food retail policy and practice and support public health [5, 6].

Food retail environments, including supermarkets, convenience stores and specialty stores, play an integral role in determining food purchasing behaviours as they influence the product availability, pricing, placement and promotion of foods and drinks [7–9]. Food retailers therefore have the opportunity to promote healthier options to their customers, thus contributing to the healthiness of purchases [10]. Globally, there is an imperative to shift towards health-promoting food retail environments to meet global targets to halt the rise in nutrition-related chronic disease such as diabetes and reduce premature mortality from major non-communicable diseases by 25% [11, 12]. A fast growing body of evidence indicates effective health-promoting initiatives in food retail for supporting healthier purchasing, but evidence to support the uptake and adoption of these is needed [10, 13–16].

Benchmarking, as part of a continuous improvement cycle, is a process used by organisations to continuously measure performance and implement change accordingly. Benchmarking processes involve a number of steps across the literature [17–20], and is described further below in the Methods. It is closely related to the implementation science strategy of audit and feedback, defined in the context of healthcare delivery as the collection and summarising of performance data over a time period to give to decision makers to monitor, evaluate and modify behaviour [21]. Within the updated Consolidated Framework for Implementation Research (CFIR), benchmarking is considered a tool for driving implementation of interventions via applying performance measurement pressure [22], which is highly relevant in the food retail sector. Benchmarking has a longstanding use by food retail businesses to improve commercial outcomes, including competitiveness and customer satisfaction [17], and an emerging application by public health nutrition researchers and organisations [5].

In the public health sector, an international network of researchers (International Network for Food and Obesity/NCDs Research, Monitoring and Action Support (INFORMAS)) has developed tools to benchmark food retail environments against best practice local policies, guidelines or voluntary codes [23]. Whilst indicators for assessing and comparing retail stores against best practice for healthy food retail have been developed and reported across a range of contexts [6, 23–30], the mechanisms by which benchmarking as a continuous improvement process can potentially drive change in the food retail sector is not well understood.

This review therefore aimed to address the overall research question: What can be learned from food retail benchmarking to support health enabling food retail environments? To address this question, the aims of the review were to: (1) examine the application of benchmarking approaches to drive performance change within food retail settings, (2) compare the application of benchmarking approaches against the evidence-based six steps of benchmarking, and (3) to explore qualitative insights from published research, including definitions of benchmarking, underlying theories of change/logic, performance outcomes and reported enablers and barriers to benchmarking in food retail settings. This review intentionally has a broad scope in order to learn from benchmarking initiatives in the wider food retail sector which may not foreground health as an objective. As such, the novel, interdisciplinary approach to this review focuses on identifying learnings from the process of benchmarking and its application within the food retail sector, rather than an in-depth analysis of the specific functions or metrics that have been benchmarked. The optimisation of benchmarking methodologies, within this sector, will contribute to the body of evidence needed to implement and sustain health-promoting food retail benchmarking interventions.

Methods

Benchmarking steps

The benchmarking process involves a number of steps (ranging from 5 to 16 steps). After listing the steps from benchmarking approaches defined in published literature [17–20], authors identified that the paper by Elmuti et al. [18] effectively synthesised the common steps across other approaches, and outlined clear definitions of discrete and actionable benchmarking steps relative to the other approaches. This, in turn, would support accurate and consistent application of the benchmarking approach across the study. As such, the following six steps, closely adapted from Elmuti et al., were applied for the purposes of this review: (1) Determine which organisational or business functions or processes to benchmark (e.g., environmental performance or company nutrition commitments); (2) Identify performance variables and measures; (3) Evaluate and compare performance; (4) Develop recommendations and actions to ‘meet and surpass’; (5) Implement and monitor; (6) Recalibrate/continuous improvement [18].

Methodological approach

A systematic scoping review methodology was used to examine the extent and nature of benchmarking applied in food retail settings and identify research gaps [31]. The study’s design followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses-Extension for Scoping Reviews (PRISMA-ScR) checklist and Guidelines (see checklist in Supplementary file 1) [32]. A protocol was drafted by two authors (MFa and JB) and a third researcher and circulated with members of the research team, with all feedback discussed and integrated to guide the research questions and methodology.

Sources of evidence

The review incorporated peer-reviewed literature available in English (Table 1). No date restriction was applied.

Table 1.

Final Inclusion/exclusion criteria

	Inclusion criteria	Exclusion criteria
Date	No restrictions
Language	English
Publication type	• Full text, peer-reviewed articles	• Commentary/opinion pieces • Abstract only • Systematic/scoping/narrative reviews • Conference presentations/articles from conference proceedings
Population (Setting)	• Food retail settings with primary focus on supply of groceries (supermarkets, corner stores, grocery stores) • Study includes multiple settings (e.g. food service + supermarkets) but stratifies results by store, chain, or company	• Food service facilities (restaurants, cafes, diners, schools, workplace canteens) or other non-grocery settings • Food retail settings without direct customer interface (e.g. warehouses/distribution centres) • Online food retail settings
Intervention	• Performance variables are modifiable by the store, chain or company • Relates to a program, policy or intervention or evaluation that includes at least one of the six steps of benchmarking • Stated intention of data collection is to drive performance change via benchmarking and/or to increase accountability	• Benchmarking initiative relates to food safetya • Benchmarking initiative relates to alcohol or tobacco retailingb
Comparator	• Study compares against a standard/guideline or uses a set of indicators to analyse food environments or interpret findings • Store-level performance comparisons are made by store, chain or company

^aMany of the food safety studies involved external laboratory testing, and it was therefore determined that the learnings from food safety benchmarking may be too specific to have relevance for benchmarking for healthy food environments

^bSales of alcohol and tobacco within food retail settings are often bound by a different set of regulations and retailing practices. It was determined that the findings from benchmarking initiatives related to alcohol and tobacco would have limited application to inform benchmarking for healthy food retail environments

Database search

Five electronic databases were searched in September 2022, and again on 4 July 2024 to update the review: PubMed, Embase, Scopus, CINAHL and MEDLINE. The search strategy was developed and refined in PubMed, with the final search strategy including terms covering two main concepts: food retail settings [“food retail*” OR grocer* OR supermarket* OR “food store*” OR “food outlet*” OR bodega* OR “corner store*” OR “convenience store*” OR “grocery store*” OR “food environment*” OR “food shop*”]; and benchmarking [monitor* OR “quality improvement” OR audit* OR “continuous improvement” OR benchmark*]. The developed search strategy was then applied across all other databases. Search results from database searches were uploaded into Endnote (version 20.4.1) and Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia), and duplicate entries removed.

Eligibility and Screening

An initial set of eligibility criteria was drafted (Table 1, Supplementary file 2) and reviewed by the research team. An iterative, team approach was taken to refine the eligibility criteria during the abstract/title screening process, as is recommended within a scoping review [33].

Study selection

Title and abstract screening

Two researchers independently screened the titles and abstracts of the first 50 results to assess the suitability of the criteria and identify any inconsistencies in approach, where the criteria could be clarified. Authors determined that some of the initial eligibility criteria required a deep understanding of the study which was often beyond the detail available in an abstract. From this, an initial set of refinements were made upon discussion with the senior author (JB), and the titles and abstracts of the remaining studies were screened by MFa and a research assistant using Covidence. Further clarification of the inclusion and exclusion criteria was identified upon discussion with the research group (see Supplementary file 2), with feedback from all authors integrated to produce a final set of eligibility criteria shown in Table 1 above.

Full-text screening

Eligible records from the abstract screening underwent full-text screening, documented on a spreadsheet (Google Sheets, Google, California, USA) with the final assessment for inclusion/exclusion updated in Covidence. Two authors (MFa and EvB) independently screened the full texts of a sample of records (n= 30, 8% of records included at full-text screening) to determine the degree of consistency between individual screening assessments. An agreement rate of 87% was found (26/30 records), which is similar to previous scoping reviews [34], and was therefore deemed acceptable by the research team. The remaining studies for full-text screening were allocated to one of two researchers (MFa and EvB). Studies for which the eligibility was deemed unclear were screened by both researchers. Conflicts were resolved by discussion between the two researchers, and where agreement was not reached, conflicts were resolved by the senior author (JB). A flowchart describing how these eligibility criteria were applied is provided in Supplementary file 3.

Data extraction and analysis

Data were extracted using Google Sheets, with separate columns in a single worksheet for charting of each variable: year of publication, journal name, journal scientific field, geographic location, food retail setting, recruitment/sampling strategy and study aim/s. Study aims were ordered into four categories according to their intent: (1) describe development of benchmarking indicators, comparisons or tools; (2) demonstrate, test or validate the utility of a benchmarking approach; (3) collect data to establish benchmarks of optimal performance; (4) apply benchmarking approach to assess retail performance. For studies articulating more than one of these aims, the category that reflected the furthest step in the six-step benchmarking process was selected.

Data specific to benchmarking were extracted as follows: (1) benchmarking definition (if provided), (2) benchmarking frequency, (3) which of the six benchmarking steps were included with a brief description of each, (4) mechanism through which benchmarking achieves performance change interpreted via a theory of change or factors reported narratively, (5) benefits/advantages of benchmarking processes, (6) impact/change in food retail outcomes, (7) enablers and barriers to successful benchmarking approaches, (8) transferability of benchmarking to other settings and (9) recommendations for improvement. Data extraction was completed by two authors (MFa and EvB). During this process, if a researcher was uncertain about the extraction of a variable, notes were made on the Google Sheets spreadsheet. Where the researcher was uncertain about their extraction for a study, extraction was completed by a second researcher, and discrepancies resolved via discussion. Google Sheets was used to analyse data on benchmarking study characteristics. For the qualitative variables 1, 4–9 listed above, data were quantified to identify the proportion of studies including information against that variable, while qualitative comments for each variable were inductively grouped into themes. This was done by one author (MFa or EvB) making a short summary of the extracted qualitative data for each variable from each study, and then inductively grouping these summaries into themes using colour coding. These were then checked by a second author (MFa or EvB), with any uncertainties resolved via discussion with the senior author (JB).

Results

Forty-six studies were included in the final review (see Fig. 1) [23, 24, 30, 35–77].

Fig. 1.

PRISMA diagram for study inclusion and exclusion

Study characteristics

Benchmarking was largely applied in the field of retail management, business and operations (n = 31, 67%), with other applications in public health nutrition (n = 11, 24%). A further three studies applied benchmarking in the field of environmental sustainability within food retail (n = 3, 7%) and one in consumer affairs. Most studies were conducted in Europe (n = 27), North America (n = 6) and Oceania (Australia/New Zealand) (n = 5). Fewer studies were published in Asia (n = 3), Africa (n = 1), and South America (n = 1). Supermarkets were the most common food retail setting described (n = 27, 59%), followed by grocery stores (n = 9, 20%), and convenience stores (n = 2, 4%) using the store setting term as reported in the papers. Eight studies (17%) included multiple food retail settings (e.g., supermarkets and convenience stores). Sample sizes ranged from 1 to 100 food retail companies to 1–2500 stores.

Study aim

The aims for each of the included studies is summarised in Table 2. The terms ‘benchmark’ or ‘benchmarking’ were included in the study aim of 19 studies (41%). Thirteen studies (28%) described the development of benchmarking indicators, comparisons or tools, six studies (13%) aimed to demonstrate, test or validate the utility of benchmarking, three studies (7%) collected data to establish performance benchmarks, and 24 (52%) applied a benchmarking approach to assess retail performance.

Table 2.

Study characteristics

First author, year	Country, continent	Food retail setting (as reported)	Sample size	Field of study	Sampling, recruitment	Study aims (as reported)	Definition of benchmarking (as reported)
Aggelopoulos, 2023 [36]	Greece, Europe	Supermarkets	1 chain, 83 stores	Retail management, business and operations	Convenience sampling based on interest of retail firm in the project and data availability	Evaluate the relative efficiency of 83 food retail stores of a major Greek Supermarket network	Not described
Aggelopoulos, 2024 [35]	Greece, Europe	Supermarkets	1 chain, 23 stores	Retail management, business and operations	Convenience sampling based on interest of retail firms in the project and data availability	Investigate the impact of acquisition and organic growth on the operating efficiency and total factor productivity change of retailing networks.	Not described
Álvarez-Rodríguez, 2019 (b) [38]	Spain, Europe	Grocery stores	1 company, 30 stores	Environmental sustainability	Not described	Evaluate the relative operational efficiency of 30 groceries and establish quantified operational and environmental benchmarks associated with optimised store performance.	Not described
Álvarez-Rodríguez, 2019 (a) [37]	Spain, Europe	Grocery stores	1 company, 30 stores	Environmental sustainability	Not described	Demonstrate the applicability of a methodology combining life-cycle assessment with data envelopment analysis for sustainable retail management via benchmarking similar commercial entities.	Not described
Athanassopoulos, 1995 [41]	United Kingdom, Europe	Grocery chains	21 chains	Retail management, business and operations	Not described	Seek to provide a framework for assessing corporate performance by using ratio analysis and data envelopment analysis as comparators	Not described
Athanassopoulos, 2003 [40]	United Kingdom, Europe	Grocery chains	28–35 chains	Retail management, business and operations	Not described	Advance the theory and practice of strategic group theory by means of powerful benchmarking tools that allow the composition of strategic groups on the basis of empirically derived production function frontiers.	Not described
Barros, 2006 [42]	Portugal, Europe	Multiple settings - supermarkets and hypermarkets	22 chains, each with 1–47 stores	Retail management, business and operations	Data from publicly available sources	Analyse a representative sample of hypermarkets and supermarkets working in the Portuguese market, using a benchmark procedure to compare companies that compete in the same market and thereby deriving managerial and policy implications.	Not described
Baviera-Puig, 2020 [43]	Spain, Europe	Supermarkets	1 chain, 61 stores	Retail management, business and operations	Not described	Create a classification system which incorporates local market data (using GIS) into evaluation of store efficiency (using DEA), which can be applied as an internal benchmarking tool.	Not described
Bolton, 2003 [44]	United States of America, North America	Supermarkets	17 chains, 212 stores	Retail management, business and operations	Chosen to be representative of small and large markets	Empirically investigate grocery retailer pricing and promotion strategies by analysing pricing and promotion decisions for an assortment of brands and categories at different stores and markets.	Not described
Bradley, 2016 [45]	United Kingdom, Europe	Multiple settings - non-specialised stores, specialised stores,other retail	Not described	Environmental Sustainability	Retailers selected from food business database	Develop a framework to measure environmental impact (emissions and water use) of food retailers and their product portfolios in a given geographical area, with a case study of application in Southampton, UK.	Not described
Chatzipetrou, 2016 [46]	Greece, Europe	Supermarkets	159 stores	Retail management, business and operations	Not described	Investigate the implementation of quality costing approaches in Greek supermarkets, given the operating environment for businesses and the demographics of supermarkets in Greece.	Not described
Cortés Rodríguez, 2023 [47]	Spain, Portugal, Europe	Small front supermarkets	1 chain, 2876 stores	Retail management, business and operations	Existing relationship between researchers and company managers	Analyse how Hoshin Kanri (HK) provides mechanisms that enhance the key success factors in lean management implementation.	Process of developing plans targets, monitoring and areas of improvement based on the previous level’s policy and assessment of the previous year’s performance
de Melo, 2022 [39]	Brazil, South America	Supermarket	1 company	Retail management, business and operations	Not described	Optimize the performance of the cashier’s service macroprocess, by reducing the average time taken to carry out the activity.	Not described
de Waal, 2017 [48]	Netherlands, Europe	Supermarkets	122 stores	Retail management, business and operations	Contacts of an external consultancy organisation in the supermarket industry. Sample stores dictated by whoever responded to the survey.	Test the application of the high-performance organisation (HPO) framework to measure internal quality and performance of franchise-owned supermarkets, and validate its ability to generate practical recommendations for improving performance of supermarkets.	Not described
Fenyves, 2020 [49]	Hungary, Europe	Grocery stores	563 stores	Retail management, business and operations	Stores listed on an existing database with annual reports.	Show how DEA can be utilised in measuring and comparing the performance of companies.	A method of measuring and improving our organisation using which we compare ourselves with the best.
Froulik, 2023 [50]	Czech Republic, Europe	Multiple settings- discount stores, supermarkets, hypermarkets	9 chains	Retail management, business and operations	Not described	Analyse the economy of the largest retail chains operating in the highly competitive FMCG market in Czechia, an analysis of the relationships between selected economic indicators and key national economic variables.	Not described
Gauri, 2013 [51]	United States of America, North America	Grocery stores	50 chains 2500 stores	Retail management, business and operations	Store data from databases	Measure and compare the inefficiencies of major grocery retailers across various formats and pricing strategies using stochastic frontier methodology.	Measuring the productivity of the stores of their own chain relative to other stores in a similar space.
Greene, 1984 [52]	United States of America, North America	Supermarkets	7 stores	Consumer affairs	Not described	Provide a protracted analysis of price levels and trends as a function of a comparative food price information program sponsored by a Public Interest Research Group	Not described
Gupta, 2010 [53]	India, Asia	Grocery stores	43 stores	Retail management, business and operations	Not described	Measure the productivity of various grocery retail outlets with specific reference to Delhi and National Capital Region (NCR) so as to identify the right potential across the formats, and to understand the areas which need to be reconsidered for future growth.	Comparing a retail outlet’s performance with that of the best performing outlets.
Hendrix, 2023 [54]	Netherlands, Europe	Supermarkets	1 chain	Retail management, business and operations	Not described	Study an inventory control problem of a perishable product with a fixed short shelf life in Dutch retail practice and derive practical heuristics in case the age distribution of the stock is unknown to approximate the situation in practice more accurately..	Not described
Horacek, 2018 [55]	United States of America, North America	Convenience stores	124 stores	Public Health Nutrition	Chosen based on proximity to a college campus.	Describe the development, reliability and validity of a novel, easy-to-use tool - the Convenience Store Supportive Healthy Environment for Life-promoting Food (SHELF) Audit.	Not described
Jensen, 2019 [56]	Denmark, Europe	Supermarkets	1 chain	Public Health Nutrition	Data provided by the chain.	Develop and pilot a methodology to benchmark a retailer’s ‘calorie efficiency’ (defined as generated profit per calorie sold) relative to other retailers	Not described
Kasture, 2019 [57]	New Zealand, Oceania	Supermarkets	2 chains	Public Health Nutrition	Market share data	Benchmark food companies (including supermarkets) in New Zealand on the comprehensiveness, specificity and transparency of their nutrition-related commitments	Not described
Min, 2010 [58]	United States of America, North America	Supermarkets	7 chains	Retail management, business and operations	Chosen for similar scales, location, product offerings, target consumer bases, and service amenities	Develop a set of benchmarks that helps supermarkets monitor their service delivery process, identify relative weaknesses, and take corrective actions for continuous service improvements.	The direct measurement of competitor performance and information on what customers really want and what competitors were doing to meet customer needs.
Niemann, 2018 [59]	South Africa, Africa	Grocery stores	1 chain	Retail management, business and operations	Purposive sampling to identify major retailers. Case study retailer chosen based on accessibility and relevance.	Identify the key factors affecting collaborative planning, forecasting and replenishment in one major grocery retailer and identify what framework is applied to implement these processes.	Not described
Pulker, 2019 [60]	Australia, Oceania	Supermarkets	3 chains	Public Health Nutrition	Purposeful selection based on location, and ‘optimised’ stores	Identify Australian supermarkets’ public health nutrition-related CSR policies, assess their quality and identify evidence of supermarkets putting them into practice supermarket own brand foods.	Not described
Ramos, 2023 [61]	Portugal, Europe	Supermarket	1 store	Retail management, business and operations	Store chosen based on product availability in terms of generating dataset	To design novel approaches to forecast retailer product sales taking into account the main drivers which affect SKU demand at store level.	Not described
Reiner, 2013 [62]	Central European countries, Europe	Multiple settings - supermarkets, hypermarkets	1 chain, 202 stores	Retail management, business and operations	Not described	(1) Develop an in-store logistics conceptual framework to form the basis for what processes to assess; (2) evaluate and quantify these processes to assess efficiency; (3) compare performance across different store formats and identify what processes lead to different efficiency outcomes.	Not described
Roig-Tierno, 2018 [63]	Spain, Europe	Supermarkets	11 chains, 45 stores	Retail management, business and operations	Sales floor area, owned by a firm	Develop the competition index (CI) to evaluate the competition faced by different outlets (based on attractiveness and distance), which allows for evaluation of strategies to improve their performance.	Not described
Sacks, 2019 [30]	Not applicable - description of a tool	Supermarkets	Not applicable - description of a tool	Public Health Nutrition	Chosen based on influence on food environments (most often decided by market share)	Describe the development of the Business Impact Assessment- Obesity and population-level nutrition (BIA-Obesity) tool and process for benchmarking food and beverage companies (including supermarkets) on their nutrition-related policies, commitments and practices.	Not described
Sacks, 2020 [24]	Australia, Oceania	Supermarkets	4 chains	Public Health Nutrition	Market share data	Quantitatively assess and benchmark the comprehensiveness, specificity and transparency of nutrition-related policies and commitments of major food companies in Australia across three sectors: food and beverage manufacturers, supermarkets and quick service restaurants	Not described
Sellers-Rubio, 2006 [64]	Spain, Europe	Supermarket chain	100 chains	Retail management, business and operations	Included based on sales floor area from database	Estimate the economic efficiency of supermarket chains in the Spanish retailing industry.	Measurement of the current performance level of an organisation and the best practically possible level, in order to subsequently identify the underlying causes of each difference.
Sellers-Rubio, 2009 [65]	Spain, Europe	Supermarkets	42 chains	Retail management, business and operations	Stores chosen based on inclusion in a database	Estimate the influence of inventory investment, wage levels, and age of the firm on retailer technical efficiency.	Measuring the difference between the current performance level of an organisation and the best possible practice, in order to later identify the underlying causes of this difference.
Shakoor, 2017 [66]	United Arab Emirates, Asia	Supermarkets	1 chain, 5 stores	Retail management, business and operations	Not described	Measure and benchmark the efficiency of retail services.	The continuous process of measuring the company’s products, services, costs and practices against those of competitors or firms that display the “best in class” achievements.
Smith, 2000 [67]	United Kingdom, Europe	Supermarkets	Not applicable - data collected from customers/consumers	Retail management, business and operations	Not described	Examine the issues involved in the collection of consumer perceived competitor comparisons and how this might differ across service industries.	A comparative process and performance measures for management decision making and control.
Sorensen, 2017 [68]	Multiple countries, Europe, Oceania, North America, Asia	Multiple- supermarkets, hypermarkets, convenience stores	42 stores	Retail management, business and operations	Settings chosen to permit certain replications or comparisons to assess generalisability	Establish fundamental patterns of in-store shopper behaviour in food retail settings to develop evidence-based benchmarks for evaluating in store activities.	Not described
Stasova, 2022 [69]	Slovakia, Europe	Multiple settings- food retail, other food retail, non-specialised retail	50 chains	Retail management, business and operations	Convenience sampling- from list of entrepreneurs in food retail; also had to meet characteristics related to business type, legal form, size, financial statements	Analyse the financial health of food retail stores using appropriate statistical methods, to analyse the market position of businesses, to identify weaknesses in businesses that may contribute to a poor financial situation, to compare the results obtained and to submit proposals to improve the current financial health of businesses.	Not described
Swinburn, 2013 [23]	Not applicable	Food retailers	Not applicable - description of a tool	Public Health Nutrition	Not applicable- description of a tool	Describe the INFORMAS Network’s monitoring approach to benchmark food environments over time and between countries.	Benchmark; a standard or point of reference against which aspects of food environments may be assessed and compared.
Trivedi, 2017 [70]	United States of America, North America	Grocery stores	1 chain, data from 24 stores	Retail management, business and operations	Not described	Offer a means of benchmarking stores and/or categories in responding to promotions, taking into account the varying contextual characteristics that may inherently aid or hinder such a response.	Not described
Van Dam, 2022 (a) [71]	Belgium, Europe	Supermarkets	5 chains	Public Health Nutrition	Market share data	Benchmark and quantitatively assess the transparency, specificity and comprehensiveness of nutrition-related commitments, as well as related practices of the largest Belgian food companies.	Not described
Van Dam, 2022 (b) [72]	France, Europe	Supermarkets	6 chains	Public Health Nutrition	Market share data	1) Benchmark and quantitatively assess the commitments and practices related to obesity prevention and population nutrition of the largest French food companies 2) Assess whether stronger nutrition-related commitments translated into stronger practices	Not described
Vandevijvere, 2017 [73]	New Zealand, Oceania	Supermarkets	3 chains, 15 stores	Public Health Nutrition	Convenience sample of stores from dominant supermarket chains Number of cash registers as proxy for store size	Validate a set of simple indicators for measuring the relative availability of healthy versus unhealthy foods in-store.	Not described
Vandevijvere, 2019 [74]	New Zealand, Oceania	Supermarkets	Not applicable - description of a tool	Public Health Nutrition	Not applicable - description of a tool	Design a food environments feedback system (FoodBack) for empowering citizens and local change agents to create healthier community food places in New Zealand.	Not described
Vaz, 2012 [75]	Portugal, Europe	Multiple settings - supermarkets, hypermarkets	1 chain, 36 stores	Retail management, business and operations	Selected by company managers to ensure homogenous sample and relevance of results.	Design a new method to compare the performance of different types of stores, that can contribute to the enhancement of the operation of retailing organisations, guiding them towards continuous improvement.	Not described
Vyt, 2008 [76]	France, Europe	Supermarkets	1 chain, 38 stores	Retail management, business and operations	Not described	Investigate the influence of store neighbourhood attributes and competition on supermarket efficiency, as measured by data envelopment analysis (DEA) models.	Internal benchmarking is a two-step process. First, the retailer defines the performance criteria. After having defined performance criteria, retailers cluster the shops into several segments that are supposed to be homogenous to support the stores’ performance evaluation.
Wu, 2009 [77]	Taiwan, Asia	Convenience stores	1 chain, 20 stores	Retail management, business and operations	Not described	Evaluate the efficiency of convenience stores in Taiwan owned by the same chain via application of data envelopment analysis methods.	Not described

DEA Data Envelopment Analysis, GIS Geographical Information System, INFORMAS International Network for Food and Obesity/Non-communicable Diseases Research, Monitoring and Action Support

Sampling

Details for how food retailers were recruited or included in the sample were provided in 27 studies (59%). Convenience sampling approaches were described in 19 studies (41%) based on data availability and store characteristics. Purposive sampling was described in 8 studies (17%), where retailers or stores were chosen according to market share and influence, to provide useful comparisons and ensure relevance of findings.

Definition of benchmarking

A definition of benchmarking was provided in 11 articles (24%, Table 2), with the majority describing competitor benchmarking (comparing with direct competitors; n = 4) or internal benchmarking (comparing within one organisation; n = 4). Industry benchmarking (comparing within the same industry) was described in two studies, and one study described a ‘benchmark’ rather than the process of benchmarking. Only one article (Cortés Rodriguez et al., 2023, internal benchmarking) described a cycle of developing targets, monitoring and continuous improvement [47].

Benchmarking frequency

Of the 39 studies that collected data to develop or test a benchmarking tool, establish benchmarks or evaluate performance (85%), 30 (77%) analysed data at one timepoint only, and nine studies analysed data at more than one timepoint (range 3–24 timepoints, noting data may have been collected for testing of methods or tools, or the development of a framework). One study was found to be evaluating a benchmarking process of continuous improvement; re-evaluating goals annually but reported overall performance change over a three-year period [47]. Ten studies (26%) collected benchmarking data for the development or testing of a tool, framework or methodology, with eight studies reporting data analysed for this purpose only at one time point and two studies reporting data at three time points. Three studies (7%) were aimed at establishing benchmarks, with all of these using data collected at one time point for establishment. Seven of the included studies (15%) described a concept or methodology only, and therefore benchmarking data were not collected as part of the study.

Mapping against the six steps of benchmarking

Table 3 summarises which of the 6 steps of benchmarking were undertaken.

Table 3.

The six steps of benchmarking

First author, year (reference)	Step 1. Determining which functions to benchmark		Step 2. Identifying performance variables and measures		Step 3. Evaluating and comparing performance				Step 4. Developing recommendations and actions to ‘meet and surpass’		Step 5. Implementing and monitoring actions		Step 6. Continuous improvement process
						Type of measure	Benchmarking target	Comparator
Álvarez-Rodríguez, 2019 (b) [38]	✓	Literature review	✓	Sustainability/Environmental impact	✓	Quantitative	Best performing in group	Internal	✓	Produced targets	✕		✕
Álvarez-Rodríguez, 2019 (a) [37]	✓	Literature review	✓	Sustainability/Environmental impact	✓	Quantitative	Best performing in group	Internal	✓	Produced targets	✕		✕
Aggelopoulos, 2023 [36]	✓	Literature review, Engagement with management of retail company	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✓	Analysis led to a series of management actions (implied feedback process)	✕		✕
Aggelopoulos, 2024 [35]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✕		✕		✕
Athanassopoulos, 1995 [41]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✕		✕		✕
Athanassopoulos, 2003 [40]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	External	✕		✕		✕
Barros, 2006 [42]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	External	✕		✕		✕
Baviera-Puig, 2020 [43]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✕		✕		✕
Bolton, 2003 [44]	✓	Literature review, Store scanner data& interview with marketing director	✓	Planning/strategies/management processes	✕				✕		✕		✕
Bradley, 2016 [45]	✓	Literature review	✓	Sustainability/Environmental impact	✕				✕		✕		✕
Chatzipetrou, 2016 [46]	✓	Literature review	✓	Quality/quality costing	✕				✕		✕		✕
Cortés Rodríguez, 2023 [47]	✓	Literature review	✓	Planning/strategies/management processes	✓	Quantitative	Gold standard	Internal	✓	Fed back to stores	✓	Employee surveys re: impact on practice, governance entity monitoring all initiatives, monthly meetings following the Plan Do Check Act cycle, monitoring scorecards	✓	Long term (3 year), annual and intermediate goals monitored via monthly meetings and scorecards to determine any corrective measures needed
de Melo, 2022 [39]	✓	Literature review, brainstorming with retailer & direct observation in retail setting	✓	Efficiency of cashier service process	✓	Quantitative & qualitative	Best possible value	Internal	✓	Suggested actions which were deemed appropriate by retailer	✕		✕
de Waal, 2017 [48]	✓	Literature review	✓	Quality/quality costing	✕				✕		✕		✕
Fenyves, 2020 [49]	✓	Literature review	✓	Efficiency	✕				✕		✕		✕
Froulik, 2023 [50]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	External	✕		✕		✕
Gauri, 2013 [51]	✓	Literature review	✓	Planning/strategies/management processes	✓	Quantitative	Best performing in group	External	✓	Produced targets	✕		✕
Greene, 1984 [52]	✓	Literature review	✓	Food environments	✓	Quantitative	Best performing in group	External	✕		✕		✕
Gupta, 2010 [53]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	External	✓	Produced recommendations but did not feed back to stores	✕		✕
Hendrix, 2023 [54]	✓	Literature review	✓	Planning/strategies/management processes	✓	Quantitative	Gold standard (from literature)	Internal	✕		✕		✕
Horacek, 2018 [55]	✓	Literature review	✓	Healthiness	✕				✕		✕		✕
Jensen, 2019 [56]	✓	Literature review	✓	Calorie efficiency	✓	Quantitative	Best performing in group	External	✕		✕		✕
Kasture, 2019 [57]	✓	Literature review	✓	Nutrition commitments and practices	✓	Quantitative	Gold standard	External	✓	Fed back to stores	✕		✕
Min, 2010 [58]	✓	Literature review, customer questionnaire	✓	Planning/strategies/management processes	✓	Quantitative	Best performing in group	External	✓	Produced recommendations but did not feed back to stores	✕		✕
Niemann, 2018 [59]	✓	Literature review, interviews with retailers	✓	Planning/strategies/management processes	✕				✕		✕		✕
Pulker, 2019 [60]	✓	Literature review	✓	Nutrition commitments and practices	✓	Qualitative	Gold standard	External	✕		✕		✕
Ramos, 2023 [61]	✓	Literature review	✓	Planning/strategies/management processes	✕				✕		✕		✕
Reiner, 2013 [62]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✕		✕		✕
Roig-Tierno, 2018 [63]	✓	Literature review	✓	Competition	✓	Quantitative and Qualitative	Best performing in group	Internal	✓	Produced targets	✕		✕
Sacks, 2019 [30]	✓	Literature review, consultation with researchers	✓	Nutrition commitments and practices	✓	Quantitative	Gold standard	External	✓	Proposed method to feed back to stores	✕		✕
Sacks, 2020 [24]	✓	Literature review, Consultation with experts to adapt existing tool	✓	Nutrition commitments and practices	✓	Quantitative	Gold standard	External	✓	Fed back to stores	✕		✕
Sellers-Rubio, 2006 [64]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	External	✕		✕		✕
Sellers-Rubio, 2009 [65]	✓	Literature review	✓	Efficiency	✕				✕		✕		✕
Shakoor, 2017 [66]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✕		✕		✕
Smith, 2000 [67]	✓	Literature review, consultation with experts	✓	Competition	✕				✕		✕		✕
Sorensen, 2017 [68]	✓	Literature review	✓	In-store shopper behaviour	✓	Quantitative	Best performing in group	External	✕		✕		✕
Stasova, 2022 [69]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	External	✕		✕		✕
Swinburn, 2013 [23]	✓	Literature review	✓	Food environments	✓	Quantitative	Gold standard	External	✓	Proposed method to feed back to stores	✕		✕
Trivedi, 2017 [70]	✓	Literature review	✓	Planning/strategies/management processes	✓	Quantitative	Best performing in group	Internal	✕		✕		✕
Van Dam, 2022 (a) [71]	✓	Literature review	✓	Nutrition commitments and practices	✓	Quantitative	Gold standard	External	✕		✕		✕
Van Dam, 2022 (b) [72]	✓	Literature review	✓	Nutrition commitments and practices	✓	Quantitative	Gold standard	External	✕		✕		✕
Vandevijvere, 2017 [73]	✓	Literature review	✓	Healthiness	✓	Quantitative	Gold standard	External	✕		✕		✕
Vandevijvere, 2019 [74]	✓	Literature review, interviews with retailers, community meetings	✓	Food environments	✓	Quantitative	Gold standard	External	✓	Proposed method to feed back to stores	✕		✕
Vaz, 2012 [75]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✓	Proposed to develop targets for each store	✕		✕
Vyt, 2008 [76]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✓	Produced targets	✕		✕
Wu, 2009 [77]	✓	Literature review	✓	Efficiency	✓	Quantitative	Best performing in group	Internal	✕		✕		✕

Step 1: Determining which functions to benchmark

All studies completed step 1 of benchmarking, as was required as part of the inclusion criteria. All studies completed a literature review as part of step 1 with nine studies (20%) complementing the literature review with consultation (with retailers, researchers, communities, and/or customers) to determine which function to benchmark. One study also used store scanner data.

Step 2: Identifying performance variables and measures

All studies completed step 2 of benchmarking. Most variables were related to efficiency (n = 17; 37%), retail strategies or management approaches (n = 9, 20%) or nutrition commitments and practices (n = 6, 13%). Eighteen (39%) studies consulted with retailers, researchers, communities and/or customers to inform performance variables and measures.

Step 3: Evaluating and comparing performance

Thirty-six studies (78%) completed or described step 3 of the benchmarking process, which involved collecting benchmarking data and analysing results at the store, chain or company level. Seven studies (15%) collected benchmarking data but did not analyse the results at the store, chain or company level and a further three studies (7%) did not collect benchmarking data or describe this step. From the studies that completed or described step 3 (n = 36), thirty-three (92%) described a quantitative measure, one described a qualitative measure, and two described a measure using mixed methods. The majority of studies completed an external comparison (n = 20, 56%) and most were comparing against the best performing in a defined group (n = 24, 67%).

Step 4: Developing recommendations and actions to ‘meet and surpass’

Sixteen studies (35%) completed step 4 or described the methods for step 4 of the benchmarking process. Of these, recommendations were given as benchmarking targets rather than actions in six studies (38%). The remaining studies outlined recommendations on actions to improve performance. Eight studies (17% of total studies) fed back or proposed methods to feed back the developed recommendations to stores.

Step 5: Implementing and monitoring actions

One study described the method of completing step 5 of the benchmarking process [47]. This involved the development of a steering committee that utilised the Plan-Do-Check-Act cycle and monitoring scorecards. It is noted that this study was an evaluation of the implementation of a company’s continuous improvement strategy.

Step 6: Continuous improvement

The same study above described the method of completing step 6 of the benchmarking process. This involved long term metrics and goals aligned with strategies that were monitored by a steering committee and corrective measures taken if required [47].

Qualitative insights from benchmarking in food retail

Mechanisms of Action

Two studies outlined a formal logic model or theory of change regarding how the benchmarking approach was intended to drive performance change in food retail. These studies were both published by the INFORMAS network; one described the overall INFORMAS framework, and the other described the Business Impact Assessment on Obesity and Population-level Nutrition (BIA-Obesity) tool, which use performance comparison and provision of recommendations to support retailer actions to create healthy food environments [23, 30]. Of the studies describing the mechanisms through which benchmarking can facilitate change (n = 23, 50%), the most common mechanisms were to inform retail decision making and priorities (n= 13, 28%) [36–39, 43, 45, 48, 50, 59, 63, 65, 68, 69], increase retailer awareness of own performance as well as best practice (n= 6, 13%) [39, 41, 45, 58, 59, 61], improve retailer accountability and transparency (n= 4, 9%) [23, 24, 30, 74], and facilitate collaboration between internal or external stakeholders (n= 4, 9%) [39, 47, 59, 76]. Each of the following modifiable factors were described as mechanisms of action for benchmarking driving change in various single studies: motivation and curiosity among retailers to understand their own performance [62]; internal allocation of resources to implement benchmarking recommendations [30]; integrated benchmarking within a continuous improvement cycle [47]; and creation of a supportive and participative workplace environment focused on achieving goals and objectives aligned to a vision/mission [47].

Impact on study outcomes

One study described the impact of benchmarking and continuous improvement on study outcomes; no other studies assessed performance change in the same food retail setting in response to a benchmarking intervention. Cortés Rodríguez et al. (2023) described the implementation of the 10-step Hoshin Kanri methodology for lean management in a food retail company operating in Spain and Portugal, involving a process of setting targets, monitoring and identifying areas for improvement based on assessment of the previous year’s performance [47]. Across a three-year cycle, with monthly and yearly reviews, authors described improvement of performance across different process areas. These included improvements in customer loyalty via provision of high-quality operations and services, improvements in supply chain productivity, learning and growth processes particularly among workforce and financial benefits in allowing the company to operate using fewer resources, thus allowing for added value services for customers at lower costs. Additional discussed benefits of benchmarking via Hoshin Kanri included improved communication between departments, higher level of staff commitment and teamwork and lean management to create a more agile organisation, culminating in improved economic results and quality improvements [47].

Benefits of benchmarking processes

Thirty-six studies (78%) described benefits or advantages of benchmarking processes. Of these, seven studies articulated observed changes within the study context, while the remaining studies referred to theoretical or proposed benefits of benchmarking. Twenty studies referred to benefits to food retail settings. Reported advantages of benchmarking included the ability to identify drivers and determinants of retail performance [40, 43, 46, 59, 63, 65, 68, 70], benchmarking findings informing retailer decision making for performance improvement [35–37, 39, 42, 45, 48, 51, 56, 62, 66, 68, 75–77], and fostering collaboration between researchers and retailers as well as within companies for whole-of-organisational change with clear communication of responsibilities [24, 30, 47, 57]. Further to this, benchmarking was proposed as an accountability mechanism which can elevate the importance of nutrition for retailers [30] and/or be a process to improve the customer orientation and perception of food retailers [47]. Benchmarking, when framed as part of a continuous improvement model, facilitated the establishment of goals and objectives [73], with corresponding processes necessary to reach these including structures for reporting and governance, time and resources [47]. Twenty-eight studies described advantages of benchmarking related to research methodologies, such as demonstrating the utility of a benchmarking tool [30, 35–39, 43, 48, 54, 56, 57, 61, 62, 76, 77], facilitating the integration of different evidence sources for holistic performance assessment [38, 39, 68, 76], as well as academic contribution in furthering research evidence in food retail performance improvement [23, 30, 35, 36, 40, 41, 45, 46, 57, 59, 62, 63, 71].

Enablers and barriers to benchmarking processes

Twenty-seven studies (59%) discussed at least one enabler or barrier to benchmarking processes. Seven studies (15%) discussed benefits of directly engaging retailers for data collection reliability, ensuring relevance of findings and the ability to better tailor recommendations [24, 30, 37, 38, 57, 71, 75]. Conversely, poor engagement from companies was commonly discussed as a barrier [24, 71, 72], with potential reasons for this including time commitment and lack of trust with researchers [57]. The risk of better performing stores engaging with benchmarking processes was also discussed, with the potential to skew findings [57]. Sacks et al. (2019) discussed restricting involvement of retailers in tool development and funding as an enabler of impactful benchmarking free of conflicted interests and highlighted the risk of food companies leveraging involvement in benchmarking research for public image [30]. Access to food retail data was a common enabler of benchmarking processes [35, 60, 64, 68], including having flexible benchmarking tools that are adaptable to available data [76]. Where this was not evident, it was discussed as a barrier [30, 51, 68, 71, 76, 77]. The importance of presenting data in an approachable and relevant manner for performance comparison and targeted recommendations was highlighted [30, 38, 39, 72], such as acknowledging local contextual factors by integrating geographical information system (GIS) analysis [63]. Developing a robust benchmarking methodology was highlighted across many studies, including factors such as having a clear scope [45], standardisation and consistency in methodology [24, 30, 36, 45, 49], a clear vision statement and intention [47], as well as ability to overcome previous methodological limitations [45]. Conversely, methodological barriers discussed included challenges in making comparisons with the data, such as across retail contexts or sectors where certain indicators may be more or less applicable [30, 42, 67, 77]. Authors also described limitations within their specific benchmarking tools as barriers to benchmarking initiatives [30, 42, 43, 63, 67, 76]. Repeated data collection [37, 39], as well as lowering cost [30, 63] and ensuring simplicity of data collection tools [36, 56, 61], were acknowledged as enablers of the benchmarking approach. If data did not capture in-store or company changes over multiple timepoints, this was described as a barrier to effective benchmarking [35, 36, 43, 71, 72].

Transferability of benchmarking approaches

The transferability of benchmarking approaches was discussed in 24 studies (52%). Nine studies commented on the relevance of benchmarking methods to be extended to other sectors [37, 38, 45, 56, 62, 63, 70, 76, 77], most commonly in other service industries [38, 70, 76], while others described the application of benchmarking methodologies in retail settings with similar characteristics, such as similar product categories or store size [37, 59, 60, 62]. Authors commented on the expansion of benchmarking approaches to other food retail contexts [43, 47, 59, 77], such as looking at individual stores as opposed to chains [48], online food retail environments [68], or analysis of different product categories [44, 62]. As well as performance improvement, the use of benchmarking by retailers to undertake evaluation was also expressed [68]. Transferability of benchmarking approaches to other geographic or cultural contexts was highlighted by five studies, including the capacity of benchmarking to measure the same company’s performance across different markets [30, 55, 58, 74, 77]. Several studies described the application of benchmarking approaches in policy settings to assess effectiveness of policy commitments, particularly related to sustainability [38] and public health nutrition [23, 30]. For example, Sacks et al. (2019) suggest that the BIA-Obesity tool could be integrated into United Nations monitoring to assess private sector voluntary targets for NCD prevention [30]. Notably, the application of benchmarking tools and findings for advocacy was described in only one study [55]. Six studies describe challenges in extrapolating findings outside of a study’s context due to factors including small sample sizes [35, 59, 76], narrow product range used [62], and differences in food retail sector composition in different contexts [64, 68].

Recommendations for improvement

Opportunity for future improvements in benchmarking was described in 31 studies (67%). The most common themes described across the studies related to the inclusion of additional or more detailed metrics to strengthen assessment of performance (n= 17) [35, 36, 41, 42, 45, 46, 51, 53, 54, 61, 62, 64, 65, 67, 68, 70, 72], including a larger sample size (n= 5) [42, 51, 59, 63, 73], analysing a greater variety of retail settings with different organisational structures or extending the benchmarking model to suppliers (n= 5) [52, 55, 59, 65, 76]. Three studies highlighted the importance of repeated assessments to assess change over time [24, 57, 58], while extending the benchmarking initiative to different product categories was suggested in two studies [43, 76]. Three studies recommended company engagement to ensure relevance and utility of results, including undertaking research to aid uptake of benchmarking processes among retailers [24, 40, 53]. For example, Sacks et al. (2020) pointed to the importance of investigating drivers and leverage points for food retailers to take action [24]. Related to this, one study suggested diversifying employee perspectives in benchmarking beyond the store or company management to obtain a well-rounded assessment of performance [48], and a further two studies recommended expanding analysis to look beyond company commitments to assess practices [24, 57]. To ensure the most effective benchmarking method was adopted, two studies suggested contrasting results between different benchmarking tools [55, 76]. One study recommended each of the following to optimise benchmarking approaches: collecting data over a long period of time [35], communicating results through alternative channels [52], setting aspirational performance targets [67], prioritising dissemination of best practices in poor performing stores [75], as well as the potential to integrate smart technologies for optimised management of internal benchmarking processes [43].

Discussion

This review draws on forty-six studies which applied benchmarking approaches in food retail settings. It is apparent that there is significant heterogeneity in how researchers conceptualise benchmarking approaches, both in definition and application, with only one quarter of articles defining the benchmarking concept. Notably, benchmarking was often conceived as a process of performance comparison, rather than as part of a continuous improvement methodology. Alstete (2008) acknowledges this misunderstanding between performance comparison and ‘true benchmarking’ in qualitative interviews with 42 business professionals regarding conceptualisation of benchmarking in the business sector, where 95% identified a misconception between performance comparison and ‘true benchmarking’ as a broader process improvement initiative [78]. These differences were reflected in qualitative interviews, asking participants about the conceptualisation of benchmarking in the business sector [78]. As such, it is proposed that the performance improvement potential from benchmarking is not likely to materialise when organisations stop at performance measurement (step 3) and do not proceed to developing recommendations and actions (step 4), implementing and monitoring actions (step 5) or a continuous improvement process (step 6) [79]. We observed only one study (Cortés Rodríguez et al., 2023) which had completed all six steps of benchmarking, undertaking repeated measurements within the same food retail organisation, with favourable improvements in study outcomes [47]. This is an important research gap identified by this review, and limits the potential of benchmarking to improve performance if interventions do not go beyond measurement. Several other studies mentioned the importance of repeated assessment as recommendations for improvement within benchmarking. An evaluation of the INFORMAS initiative from 2012 to 2020 highlights the importance of repeated benchmarking to analyse trends, explore links to health outcomes and evaluate the effectiveness of policy, and is a key future direction of the project [6]. To maximise the potential of benchmarking initiatives in health-enabling food retail environments, interventions should clearly articulate a definition of benchmarking, describe the steps, and ensure benchmarking is applied as part of a continuous improvement process.

The understanding of benchmarking as performance comparison is also reflected in the observation that only eight studies (17%) delivered feedback or described the process to feedback to food retailers as part of step 4 of the benchmarking process; to develop recommendations and actions to ‘meet and surpass’. One notable example of this in public health nutrition is the BIA-Obesity tool from INFORMAS, which involves the development of company-specific recommendations for actions to improve the health of the food environment in consultation with company representatives, with a scorecard privately fed back to retailers prior to public release [30]. Action planning has been found to be a critical factor in enabling change in audit and feedback interventions in healthcare [80]. Analysis of a Cochrane review of audit and feedback interventions in healthcare settings identified that feedback was most effective when accompanied by both explicit goals and an action plan to address performance gaps [81]. Additionally, in an expert consensus in audit and feedback interventions, a number of ‘best practices’ have been proposed regarding the feedback components, including multi-modal presentation of performance data, delivery from a trusted source and ensuring comparison data is relevant for the stakeholder [82]. In our review, we identified studies that recommended engagement with retailers to ensure relevance and utility of results and research to aid uptake of benchmarking. However, no publications explicitly described how they considered the relevance of these feedback elements to stakeholders in their methodology. While this omission may reflect word limitations, scope and requirements of scientific journals, it highlights an important opportunity to strengthen benchmarking methodologies to consider the design and delivery of feedback in a way that most effectively facilitates performance change in food retail settings.

Understanding the mechanisms for how an intervention drives change is proposed by implementation science researchers to improve evaluation in public health [83]. However, only two studies identified in this review described a program logic model or theory of change underpinning their benchmarking initiative; one for the BIA-Obesity and one for the INFORMAS framework overall. From this review in food retail settings, benchmarking processes were most commonly proposed to drive change via informing retailer priorities and decision making, accountability and transparency, increasing awareness of own performance and best practice and facilitating stakeholder collaboration. It has been indicated from audit and feedback studies that a lack of research attention on the underlying theory and conceptual basis can hinder understanding of differences in effectiveness between interventions [84]. This aligns with findings from a systematic review of audit and feedback in healthcare, in which only 14% of studies reported use of a theory within any aspect of the study, and only 9% of studies reported intervention designs informed by theory [84]. Acknowledging that food retailers have finite capacity for continuous improvement, designing theory-informed benchmarking that clearly articulates pathways for change can facilitate effective deployment of resources, justify any required investments and strengthen understanding of the key ingredients for successful implementation [82, 85].

This review highlighted great diversity in the role that food retailers played in benchmarking initiatives. In many studies, researchers accessed publicly available datasets with no engagement with retailers. Where retailers were engaged, benchmarking initiatives involved contrasting approaches to the involvement of retailers in guiding study approaches and tools. For example, while Sacks et al. (2019) described limiting the involvement of food retailers in the development of benchmarking tools for health promotion as a strength of the design [30], the study by Chatzipetrou & Moschidis (2016) disseminated the survey tool to supermarket managers and integrated feedback in its development targeting improved quality costing [46]. Overall, engagement with food retailers was highlighted as an enabler to successful benchmarking, particularly in terms of data availability and reliability and potential to drive performance change. Research points to the importance of engagement with food retailers in designing and developing effective health-enabling interventions in food retail environments [10, 86, 87]. While co-design processes are increasingly being applied as a means to strengthen health-enabling food environment interventions, no studies in this review were identified to have adopted a co-design methodology. There was an observed trend towards studies in the public health nutrition field disclosing greater detail around the role of retailers relative to studies in retail management, business and operations field. In contrast to the body of evidence advocating for co-creation of interventions with retailers, this trend may also reflect increased research attention on the potential conflicts of interest that can arise from public health researchers collaborating with food industry stakeholders [88, 89].

While benchmarking has been widely applied in healthcare, this is often in the context of established national minimum standards such as the National Safety and Quality Health Service (NSQHS) in Australia, which are actively enforced [90]. In food retail settings in Australia, aside from mandatory food safety regulation, there is no regulatory framework outlining standards for health-enabling retail practice, other than the recent Code of Practice for Remote Store Operations (Code of Practice) introduced by the Commonwealth government [91]. This reflects a global trend of limited policy implementation to regulate food environments at the retail level, with the exception of school food environments, to promote population health [15, 92–94], and thus could limit the potential to engage food retailers for health improvement. There are, however, examples of leading practice such the United Kingdom’s Food (Promotion and Placement) (England) Regulations 2021 to reduce the promotion and visibility of unhealthy foods to consumers [95]. In remote Australia, where many food stores are owned by community, partnership between remote retail stores and researchers has culminated in implementation of a benchmarking model following a continuous improvement cycle across 29 community stores to improve the adoption of health-enabling standards at scale [5]. This benchmarking approach has informed the Commonwealth’s Code of Practice health standards, and their implementation in remote stores nationally.

Limitations

The authors are researchers in the field of public health nutrition, while the review drew from the fields of retail management, business and operations, which may potentially have impacted the technical depth to which the included studies could be scrutinised. The review, is however, on the implementation of benchmarking and drawing out insights and learnings of relevance for public health, which closely aligns with the authors’ expertise. Adopting a systematic scoping review methodology provided the flexibility to explore emerging trends and outcomes of interest to the design and implementation of benchmarking initiatives in food retail [96]. Expanding the scope of the review beyond benchmarking for health-related retail outcomes also enabled the collation of cross-disciplinary insights into how benchmarking initiatives can drive change in food retail settings. There is no accepted gold-standard benchmarking process with a set of criteria or steps. For the purpose of this study, we adopted a 6-step approach informed by existing research, however not all studies conformed neatly into this framework for comparison. The review assessed publicly available data published in academic journals, so any benchmarking initiatives undertaken by food retailers not available in the public domain, or those not published in peer-reviewed journals were not integrated in this review.

Conclusion

This systematic scoping review of the use of benchmarking in food retail environments revealed significant heterogeneity in the definitions and application of benchmarking methodologies, with only one study completing all six steps of benchmarking. Very few studies describe a theory of change, or articulated feedback strategies to facilitate performance improvement among food retailers, with a varying degree of involvement of food retailers identified across the included studies. Based on our findings, we recommend that benchmarking should be conceived as a tool for continuous improvement, rather than simply performance assessment to facilitate feedback and action among food retailers. The identified benefits of engaging directly with retailers as part of benchmarking should be considered by researchers and practitioners designing benchmarking initiatives in future, though further research is warranted on the design and communication of feedback. Embedding benchmarking processes into existing workflows and regulation presents a valuable opportunity to further drive performance change at scale. This review provides important insights to optimise benchmarking methodologies and, in turn, maximise the potential of future food retail benchmarking interventions to create and sustain health-promoting food retail environments.

Authors’ contributions

JB developed the initial research idea. JB, MFa and EvB led the design of the study, with input from all other authors (AH, MC, MFe, EM). MFa and EvB screened articles, extracted and synthesised data, supported by JB. MFa and EvB drafted the manuscript, with all other authors (JB, AH, MC, MFe, EM) reviewing draft and final versions of the manuscript.

Funding

The work was funded by JB’s Investigator Grant #2017170 from the National Health and Medical Research Council.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.