Exploring the use of innovation measurement in retail organisations: a multiple case study

ABSTRACT Innovation and its management in retail organisations is an emerging field, yet the measurement of innovation in this context remains challenging and under-researched. The study investigates innovation in large Swedish retail companies, focusing on monitoring and measurement of innovation performance. Through a qualitative, multiple case study, our findings reveal that retailers are aiming to become more active innovators. They are exerting a greater control over the innovation process, implementing a range of financial and non-financial indicators that encompass inputs, outputs, and the process itself. Innovation projects serve as vehicles for experimentation and learning about how to improve the structure of innovation efforts. However, challenges still arise when strategic objectives are not aligned with performance measurements across the organisation. Nonetheless, existing performance measurement practices can support innovation capability development in retail, especially in relation to culture and competence development and utilising external knowledge for open innovation. The paper advances theory by offering insights into how retailers gain more control of innovation through measurement, contributing to the growing landscape of strategic innovation management in retail. We propose a model, informed by theoretically built framework and empirical data, to guide retail organisations in defining new measurements that promote a holistic perspective to building innovation capability.


Introduction
Once thought to be detrimental to innovation (Damanpour 1991), management control systems such as performance measurement are now considered to have a positive role in innovation and the management of the innovation process (see, e.g., Barros and Ferreira 2019;Bedford 2015;Davila, Foster, and Oyon 2009).Due to the complex and multidimensional nature of innovation, its measurement has been characterised by a diversity of approaches, prescriptions, and practices in literature (see, e.g., Adams, Bessant, and Phelps 2006;Dewangan and Godse 2014;Dziallas and Blind 2019).
Publications on innovation metrics have mostly studied manufacturing industries, accounting for 74% of the total, although there is an increasing shift towards the service industries (Dziallas and Blind 2019).While measuring innovation is a critical issue that has received significant attention, innovation measurement in retail and consumer servicesremainspoorly understood.
In the first qualitative study that explored measurement of innovation in retailing (Hristov and Reynolds 2015), measuring innovation performance in retailing proved to be a difficult task due to the open, complex and hybrid nature of innovation in the sector.Retailers engage in innovation activities that are characteristic of both production and services sectors and involve both technological and non-technological innovations (Reynolds and Cuthbertson 2014;Sundström and Radon 2014).The nature of competitive retail markets leads retailers to often exhibit more incremental and continuous innovation practices (Sundström and Radon 2014).Thus, although retailers are not inherently less innovative than other sectors, conventional ways of thinking about and measuring innovation may fail to fully capture the intensity and nature of innovation efforts in retail (Reynolds and Cuthbertson 2014).
Managerial accounting research advocates a contingency approach to measuring innovation, where there is a need to explore the role of measurement in relation to the specific tasks, contexts and settings in which innovation occurs (Brattström et al. 2018).However, there has been very limited research on its role in retail organisations.Despite innovation and its management in retail organisations becoming an emerging topic in retail research and practice, performance measurement for innovation in this context remains notably under-researched (Hristov and Reynolds 2015;Wood et al. 2008).
To date, Hristov and Reynolds' (2015) study remains one of the very few articles investigating this subject, and there is a lack of in-depth case studies that document actual insights from retail organisations attempting to measure innovation.The following research question thus emerges as important for further investigation: how do retailers currently measure innovation and what are the key issues faced in the use of innovation measurement?The paper has as its basis a qualitative, multiple case study involving three large retail companies in Sweden that have engaged in various forms of innovation projects.The study's purpose is to build on limited extant research and increase our understanding of the nature of innovation efforts in large retail companies through the analysis of attempts to monitor and measure innovation performance.It contributes to theory by providing a contemporary understanding of the growing phenomenon of strategic innovation management in retail firms, signifying a shift from previous understandings of how retailers innovate, as well as with an empirically based and prescriptive model of what measurements to consider in building the innovation capability of a retail organisation.
Defining innovation in retail is a complex endeavour due to retailers engaging in innovations that are characteristic of both production and service sectors, both product and process innovation, as well as technological and non-technological innovation, hence demonstrating the complex and hybrid nature of retail innovation (see, e.g., Sundström and Radon 2014).Hristov and Reynolds (2015) argue that innovation in retailing possesses a range of sector-specific meanings and hence proposes retail innovation typologies according to impact (i.e.strategic innovation vs operational innovation) and according to application areas (i.e.customer-related innovation, support-related, or organisation-related).Due to the retail environment being inherently characterised by hypercompetition, uncertainty, and low appropriability, most retailers tend to innovate incrementally and continuously rather than through radical or breakthrough innovation (Reynolds et al. 2007;Sundström and Reynolds 2014;Hristov 2007).Successful retail formats -and retail innovation in general -seem to emerge from opportunistic and incremental processes, sometimes based on intuition and 'capturing the moment' rather than detailed planning and rational analysis (Reynolds et al. 2007).Though it is important to bring along the characteristics of retail when researching innovation in retail, or as here, innovation measurement in retail organisations, it is also critical to take into consideration how innovation is a social process including idea creation as well as implementation (Garud, Tuertscher, and Van de Ven 2013) and that it requires collaboration and experimentation.Van De Ven (2017) also stresses that innovation is better to be looked upon as a journey than a controlled process, rather to be manoeuvred than to be controlled (Van de Ven 2017).This in particular intrigues the research on innovation measurement with the tension inherent in leading the work within organisations that needs to be guided but at the same time is difficult to fully control.Neely et al. (2005) define performance measurement as 'the process of quantifying the efficiency and effectiveness of action.'Traditional thinking about management control systems, including performance measurement systems, was that these were perceived as detrimental to innovation, constraining the freedom, creativity, experimentation, and flexibility of developers (Davila, Foster, and Oyon 2009).The new paradigm, however, is that these systems can play a positive, important role in innovation and the management of the innovation process (see, e.g., Barros and Ferreira 2019;Bedford 2015).This new way of thinking is consistent with innovation not being a random event but rather a manageable organisational process (Davila 2005).

Innovation measurement in organisations
Saunila and Ukko (2013) report that performance measurement has positive effects on different aspects of an organisation's innovation capability.Innovation capability is an organisation's ability to continuously transform knowledge and ideas into new products, processes, and systems for the benefit of the firm and its stakeholders (Lawson and Samson 2001).Innovation performance measurement has been shown to be instrumental in building innovation capability through creating and supporting routines (Nilsson and Ritzén 2014;Pavlov and Bourne 2011).It can have a triggering role to take actions, supports reflective sessions on results, and motivates to set goals and provide information on goal fulfilment (Nilsson and Ritzén 2014).Kaplan and Norton (1992), in proposing the balanced scorecard, emphasise that an innovation and learning perspective should be included in performance measurement systems since traditional financial measures are insufficient in addressing the demands of today's competitive environment.However, innovation encompasses multiple performance objectives that require different metrics, which is a topic that has acquired considerable attention in research (see, e.g., Adams, Bessant, and Phelps 2006;Dziallas and Blind 2019).Measurement of innovation is considered particularly challenging because the innovation process is complex, multidimensional, and often has an unpredictable outcome (Van de Ven 1986).Key problems in innovation measurement are related to the underlying conceptualisation of what is being measured, and the general feasibility of different types of measurement (Smith 2005).Adams et al. (2006) highlight the following innovation management measurement areas: inputs, knowledge management, strategy, organisation and culture, portfolio management, project management, and commercialisation.Dziallas and Blind (2019) suggest the following company-specific dimensions for innovation indicators: strategy, innovation culture, competence and knowledge, organisational structure, R&D activities and input, and financial performance.Innovativeness measures have also been classified in terms of inputs (R&D investments, R&D staff qualification, patents, etc.), capabilities and processes (culture, leadership, knowledge, strategy, etc.) and outputs (number of innovations, percentage of revenues of innovative products, etc.) (de Carvalho et al. 2017).Davila et al. (2012) suggest that innovation measurement framework includes indicators for inputs, process (selection and execution), and outputs/outcomes.Moreover, they suggest that innovation measurement can vary depending on the level in the organisation, i.e. project-level measurements are different from measures at a higher, strategic level.Richtner et al. (2017), meanwhile, propose that a balanced set of measures would include measures of the overall portfolio, the innovation process, and individual projects.Bringing together these sources, it points to an important opportunity to use performance measurement for building innovation capability in firms; however, this usage should be accompanied with that innovation is both a strategic issue and a work that takes place on a more operational level, e.g. in projects.Also, innovation is about experimentation and risk taking and results might show in different timelines than incremental development work.Reynolds et al. (2007) conclude that innovation retail innovation seems to emerge from opportunistic and incremental processes and that it is less of detailed planning and analysis that precede innovation results.Reynolds et al. (2007) also identify that sophisticated key performance indicators (KPIs) are often only applied after these events.Reflecting on the narrow use of innovation performance measurements in retail literature and practice on measures of novelty of product innovation, Adams et al. (2008) proposed that retailers should also consider innovation measurements in the following areas: innovativeness, e.g. a count of new products or services launched; strategic importance of innovation, e.g.contribution of the innovation in meeting the firm's strategic objectives; technological application, or assessing whether or not the firms use new materials, technologies, etc., in its service offer; and reinforcement of innovation orientation, e.g. a culture that facilitates innovation.

Innovation measurement and retailing
In Hristov and Reynolds' (2015) report on the challenges of defining and measuring retail innovation, it is suggested that the main difficulty of measuring innovation in this sector lies in the iterative and transversal nature of retail innovation across different functional budget and areas.Due to limitations in management account systems or lack of coordination among various 'budgeting silos,' the less tangible benefits of innovation (e.g.organisational learning, total impact on the business and brand equity) are considered only 'secondary' effects of innovation and are harder to understand and measure, leading retailers to mostly focus on the measurement of tangible performance outputs.Furthermore, they observed that retail managers often only measure what is easy or conventional to measure arising from innovation, e.g.short-term financial benefits, instead of long-term behavioural or organisational transformation.According to the authors, this could explain the disconnect between strategic innovations versus more operationally focused methods of innovation measurement where retailers are more equipped to measure short-term, tangible performance targets.Kumar and Venkatesan (2021), noting the need for transformation of retail metrics and analytics in general, emphasise that traditional metrics for measuring the performance of brands, products, and firms in a geographical area do not present an accurate picture of retail performance anymore, nor do these reflect how the industry is currently set up and is continuously evolving.According to Sides et al. (2019), retailers must make fundamental changes to how success in retail is defined and measured.They propose retailers to use metrics that are holistic and inclusive (i.e.metrics that relate to all organisational units and applicable to all business models and retail formats), value-driving (i.e.metrics that incorporate areas that add value to the organisation), operational (i.e.reflective of operations and not simply financial ratios), and balanced (i.e.metrics that are balanced between growth and profitability and focus on recent performance).However, newer research reveals that while financial measures are still predominantly used in retail -not just for innovation performance but retail strategic and operational activities in general -there has been a rising awareness of non-financial, more complex and intangible measures such as customer experience or customer satisfaction (Cakir, Bezbradica, and Helfert 2019).
When looking at the relationship between measurement and innovation in general, the actual metrics themselves might not be as important as how they are implemented within and beyond the organisation (Brattström et al. 2018).Research on management control systems has shown that measurement can be used in either 'diagnostic' or interactive ways (Tessier and Otley 2012).For instance, Bedford (2015) suggests that the interactive use of control systems has been associated with performance in companies that are more focused on exploratory innovation, although companies that want to pursue both exploration and exploitation could benefit from the dynamic tension created by the interactive and diagnostic use.Given the diversity of innovation metrics that exist, trying to find the perfect metric might be a futile task.Instead, Richtner et al. (2017) suggest that the key managerial challenge is first understanding the problem that measurement is meant to solve in the organisation, and then designing and implementing an innovation measurement framework that is appropriate to the organisation's needs.For measurement in general to provide its potential benefit, it needs to take into consideration contextual and environmental factors (Chiesa et al. 2009;Henri 2006).Most importantly, measurement should be aligned with an organisation's strategy (Melnyk, Hanson, and Calantone 2010;Micheli and Manzoni, 2010).

Methodology
Given the exploratory nature of the study, we use a qualitative multiple case study research methodology.This research approach was chosen to facilitate an initial yet indepth investigation of how innovation performance was being measured in retail organisations.The case study research methodology may be used when there is a need to understand social phenomena, including organisational and managerial processes (Yin 2003).Moreover, using a qualitative method allows us to gather rich descriptive data not limited to innovation measurement practices but also other processes and activities related to innovation in the retail organisations.Qualitative case studies are often considered in innovation management research to be well suited in addressing the complexities existing in this field through their ability to offer rich, contextualised explanations of actions and interactions over time (Elsahn et al. 2020).
For case selection in this explorative research paper, the study adopted the principle of purposeful sampling (Patton, 2002), wherein instead of aiming for generalisation of results to a larger population, cases were chosen on the basis of their relevance to the topic under investigation, which is innovation capability in retail.The Swedish retail market is predominantly dominated by domestic brands and a small number of large chains, particularly in certain sectors such as food retail and furniture retail.Thus, the high concentration of the food retailer market in Sweden, for instance, implies that the strategies of one or two retailers can have a substantial impact on the entire market (Anselmsson and Johansson, 2009).Hence, we looked at retail firms in Sweden that have a significant market share in their respective categories, and in particular those with a well-established history in physical retailing but are in the transformation journey towards omnichannel retailing, and are showing increased innovation ambitions, i.e. investing in large innovation initiatives that are new to their company.
Three retail companies -a grocery retailer (GR), a convenience retailer (CR), and a home furnishings retailer (HR) were thus chosen for this multiple case study.The case companies are all major retail chains and have presence in different markets, although we limit the study to their business units in Sweden.These companies were purposively selected in part because as large, established retail corporations, they face a greater pressure to become more innovative, to challenge their current paradigms and avoid 'incumbent inertia' (Lieberman and Montgomery 1988) as the industry continues to be disrupted.The study is based on a research project that followed the retail companies as they engaged in various innovation initiatives in their organisation.From November 2019 until May 2021, data were collected regarding how innovation, through these initiatives, was managed and controlled in the case companies.
Two main types of data collection methods were employed (see Table 1 for case subject details).First, semi-structured interviews form the basis of our primary data source.Semi-structured interviews were conducted with managers and employees directly involved in various innovative activities and projects, as well as related internal stakeholders.The semi-structured interview sessions lasted from 60 to 90 minutes and were recorded and transcribed verbatim, with approval of the interviewees.Interviews were based on a loose and evolving interview guide that inquired about innovation and development-related work and how it is managed in the organisation, including the selection and implementation of relevant performance measurements.
Participant observation was also used to provide supplementary material in building the case study and hence serves a secondary purpose.This includes direct observations of practices and interactions during physical and online meetings, workshops, and site visits to the headquarters of the different case companies.It takes into account both natural conversations and unstructured interviews.Field notes were taken to capture our observations and reflections, as well as records of the details and content of the meeting or workshops.As the field notes that were taken in the context of our case studies were mostly descriptive information, the following template was used: date, location, activity (e.g.meeting, workshop, etc.), participants, time duration of activity, descriptions (e.g.meeting agenda, information and other materials presented, etc.), questions for followup, and personal reflections.In addition to the field notes, we also used public and private documents as reference to collect additional information necessary in building the case study.
In 2021, at the final stage of data collection, we also included a checklist of innovation measurements (see Table A1 -Appendix) in the interview guide to provide further structure to the discussions.Based on the reviewed literature on retailing and innovation measurement, relevant innovation performance measurements and indicators were compiled.The indicators are mapped into a 2 × 2 matrix, a simplified version of similar frameworks, building on the works of Davila et al. (2012) and Hristov and Reynolds (2015).Financial and non-financial indicators are categorised from an organisational perspective, i.e. measurements at project-level or strategic level.These indicators were derived from extant literature on innovation metrics, including Adams et al. (2006), Davila et al. (2012), Saunila and Ukko (2014), Nilsson and Ritzén (2014), Dewangan and Godse (2014), Richtnér et al. (2017), Dziallas and Blind (2019).The checklist was not meant to be an exhaustive catalogue but rather functions as a simple guide that can be used in the interviews.Respondents were asked to provide commentary on metrics that were being used in their organisations, with follow-up questions also asked to investigate the respondent's insights on innovation performance measurement in their organisation.
Case study data were analysed following an abductive logic of theoretical thematic analysis.A preliminary analytical framework consists of articulated 'preconceptions,' but over time was developed continuously according to what was discovered through empirical observations, analysis, and interpretation (Dubois and Gadde 2002).An open coding process was performed for each interview transcript, using software (ATLAS.ti)during initial coding to improve traceability and transparency, resulting in a set of firstorder codes which were ranked in frequency and relevance.This was followed by sorting and synthesizing of the data, wherein emerging patterns were identified, and constant comparison with theoretical constructs was performed.A more theoretically driven stage followed, which provide less of a rich description of the data and more of focused analysis on some aspects of the data (Braun and Clarke, 2006).A more focused literature review was performed, looking into the topics of innovation performance measurement across different organisational levels and across different stages of innovation, as well as classifications and typologies of innovation indicators (e.g.financial or non-financial).These were used to analyse whether emerging themes in the data suggested concepts that help explain the phenomena being investigated.Categories were then reduced to a manageable size, developed and related based on similarities and differences found between them.Throughout this process, the field notes and archival material were used to fill in additional information that can help build the case study.
While the need for validity and triangulation in case study research does not guarantee repeatability of the observations and interpretations, triangulation is nonetheless important in clarifying meaning by 'identifying different ways the phenomenon is being seen' (Stake 1994, p. 241).To promote credibility and enhance the overall study quality, triangulation of data sources, data types and researchers was used as a primary strategy to view the phenomena being studied from multiple perspectives.As an explorative case study, we follow the within-case analysis with a cross-case search for differences and similarities of patterns in relation to how performance measurement for innovation was used across the three cases.

Case 1: Grocery Retailer (GR)
The Grocery Retailer's private label innovation and concept development team (hereby simply referred to as 'innovation team') is a newly established team, situated in the Category and Purchasing department.Their mission was to contribute to the long-term development of the retailer's private labels by creating innovative product concepts that can be commercialised by relevant stakeholders in the Category and Purchasing department.The study follows how the innovation team worked on developing new methods and processes for innovation befitting their mandate, which included establishing the control and measurement aspects of their innovation activities.Embracing an open innovation model, the team is also mandated to leverage external networks and partnerships in order to acquire competences necessary for long-term innovation across the food value chain.The team's output must complement the existing private label range but must also contribute to strengthening the brand image.Because of the cross-functional nature of the projects in their portfolio, they also have a coordinator role internally.
The team developed an 'Innovation Playbook' that formalised how they will carry out their task of exploring new opportunities (aligned with GR's overall strategic areas), the innovation process, as well as goals and other performance measurements.During the work of developing the playbook, the innovation team went through a collaborative process of developing their KPIs such as time measures, number of ideas for innovations, selection criteria, and degree of innovation, among others.The playbook not only helped clarify the innovation activities and ways of working for the innovation team itself, but it also served as a tool for them to communicate their innovation work to other parts of the organisation, which was important in securing support from internal stakeholders.
The team acknowledges that they are entrusted a certain level of freedom to explore product concepts that are much more innovative than the current assortment and to be fast-moving and have agile innovation process.However, it has not been easy to carry out their mandate in practice because their innovation output needs to ultimately be integrated to the Category and Purchasing department's activity.Thus, reaching the team's KPIs will be insufficient on its own as they must also contribute to their parent department's more traditional targets (e.g.sales, market share, customer satisfaction, etc.).Given that the scope of their work involves creating innovative concepts specifically within strategic areas like sustainability and health, the impact of their projects could be more towards the long term (e.g.brand image) rather than short term (e.g. sales and market share), which was something difficult to reconcile with the Category and Purchasing department's profitability goals.

Case 2: Convenience Retailer (CR)
For Convenience Retailer, the study follows how senior management, led by Business Development Manager, was in a process of designing a common view of innovation and establishing a better system to manage innovation-related activities, including identifying relevant performance measures for the entire company in general.This was done in relation to the development and launch of two innovation projects, one is an unmanned store, and the other is a sustainability store, both new additions to their portfolio of convenience store formats.
Choosing relevant performance measures in CR has been closely linked to the design and formalisation of the entire innovation process.Senior management acknowledged the importance of ensuring that input of new ideas will have high potential, so they developed appropriate criteria for the selection and validation of ideas to implement.Previously, there was no standardised approach to pitching project ideas to department heads and senior management.Senior management thus developed a Business Model Canvas (BMC) for anyone in the company to use as a tool to present a new project idea in a more structured way.With several criteria related to value creation and value capture, the canvas could aid senior management in the screening process.
There was also a need to better plan and monitor the outcomes of the development processes, as well as the process itself.They carried out project management training and adopted standardised processes, while exercising caution to not add layers of bureaucracy, in line with their corporate culture.CR has also developed an organisation-wide project management 'Playbook' for innovation and development projects, but also more specific department playbooks which detail how each department will carry out their activities, whether related to innovation or not.These documents have led to an improved understanding of the working methods and roles in the organisation.In addition to establishing formal guidelines for various forms of innovation and development projects, they also facilitated greater transparency around the division of responsibilities in innovation work with particular focus on the process from idea generation to implementation.Throughout the implementation of the documents, senior management emphasised that these are meant to be flexible, to be tested in practice and revised as needed so that they can be adapted to the needs of the document users.
In general, management has been willing to explore different metrics that can enable more exploratory work, such as experimenting with new sustainability metrics.They have also introduced '20% innovation time,' a loosely implemented initiative for employees to spend 20% of their time on development or innovation-related work.While this has been a welcomed initiative, details of its implementation are an ongoing point of discussion, which further emphasise the need for definitions, delimitations, and measures for monitoring.
The challenge for this retailer has not been a lack of innovation support or willingness to innovate from senior management.Rather, the challenge has been setting up clearer guidelines to streamline and focus development work, enable better practices for monitoring and follow-up, but also ensure projects have sufficient impact in terms of value creation.

Case 3: Home furnishings Retailer (HR)
HR created a project team that was tasked to develop completely new expansion strategies in cities through new, innovative retail solutions and store concepts that challenge HR's status quo.The study follows how this project was managed, including the performance measurement involved in the development of the expansion strategies for the new store formats.For the most part, management ultimately used existing templates and indicators for market and business development.The project itself has not involved the development of a new innovation process, although a large part of the work has been in identifying new customer needs and market opportunities for new customer channels with increased accessibility as goal, as they shift from their longestablished expansion strategy of building bigger stores in the city outskirts to increasing presence in city centres.Nonetheless, innovation appears to be an important issue to HR, evident in management's communication, internally and externally, as well as significant investments primarily in innovative initiatives related to sustainability.
Given the organisational location of the project within the department responsible for expansion strategies (i.e.establishment of physical stores), the biggest challenge for the project team has been to take new solutions into account while fitting these into existing requirements for the current physical store formats.An example is the use of financial goals according to their current governance principles, in particular profitability and return on investment (ROI).The original KPIs for the project had been growth and market shares; however, the direction was changed to instead focus on profit, placing even higher profitability targets than initially planned.This hampered the project to some extent and required revision of plans to make the new retail concepts 'fit in.'More exploratory retail concepts were dropped from the project plans, while other less risky concepts were prioritised.Moreover, the documentation requirements based on existing project management templates were perceived by some in the team as too detailed, extensive, and bureaucratic for a project that was supposed to be a way for HR to be innovative, exploratory, and adaptable to a rapidly changing consumer base.There were reflections on how learnings obtained from customer insight measures (which are considered important KPIs across the organisation) could be utilised even more in the development process of the new store formats, instead of focusing only on the extensive documentation and internal processes.
Time measures, number of cross-functional initiatives and project cost versus budget were other measures utilised.Changes and delays in the project have also been caused by COVID-19 pandemic; however, the pandemic also had a positive impact on urgency of decision-making regarding implementation of new delivery points, carried out at a rapid speed previously unseen in the company.

Innovation measurements used by the case companies
All case companies use a combination of several innovation indicators and measurements found in extant literature (see Table 2).
In accordance with Hristov and Reynolds' (2015) study, financial indicators used by retail companies in this study were mostly related to rate of return measures, profit margins, and sales and market shares attributed to innovations.Financial effectiveness at project level is also measured (e.g.project cost vs budget).
Non-financial indicators were also used, both at the strategic level and project level.At the project level, all retailers monitor customer insights and emphasise its importance, similar to Hristov and Reynolds' (2015) findings.Beyond that, there was more variation between the retailers regarding their other project-level indicators.This could be attributed to differences in the type of retail innovation projects that we followed across the three cases, thus needing different types of project-level measurements.For GR, the innovation projects were related to new product concepts for the private label line, which followed a more established new product development (NPD) process and one type of output (products).For CR and HR, the innovation projects were related to new store formats which required even more cross-functional work and involved a broader range of innovation output, including new services, processes, and technologies.
At the strategic level, GR and CR look at the portfolio balance, use various indicators for measuring the performance of new products, new brands and retail formats, but also  have indicators related to planning and monitoring the innovation process, e.g.ways in which they formalise the process.HR, on the other hand, has extensive monitoring for project management tools of existing store formats but does not necessarily have differentiated processes and indicators for the development of new, innovative retail channels.Our findings reveal that all retailers use many qualitative, non-financial indicators related to innovation inputs.These include measures on culture, communication, leadership, and employee well-being and satisfaction, knowledge and competence development, having a mix of innovation sources, including external knowledge and collaborations, and market and technology research.

Discussion
Based on the empirical findings and building on extant literature, this section presents a discussion of how retailers currently measure innovation and the key issues faced in the use of innovation measurement.

Retailers take control of the innovation process
Innovation performance measurement, and management control of innovation in general, is instrumental if retailers aim to have a better understanding on how and where to be innovative, i.e. strategic areas that need resources for innovation, and how to carry out innovation efforts efficiently and effectively.Large retailers are aiming to become more active innovators by engaging in innovation beyond continuous, operational developments.A key finding is that internally in these organisations, there is a conscious shift towards greater structure, control, and documentation of the innovation process, signalling a change from previous understanding of how retailers innovate (Reynolds et al. 2007;Schaffers et al. 2011;Sundström and Radon 2014).There have been calls for retailers to approach innovation more maturely and strategically beyond enhancing the customer experience on an ad hoc basis (Pantano, Pedeliento, and Christodoulides 2022).The findings demonstrate how these large retailers are moving towards the direction of placing innovation on a higher strategic priority.They were involved in various degrees of codification of processes related to development and innovation, establishing clearer working methods, roles and responsibilities, and implementation of performance indicators that assess the inputs and the innovation process itself, and not just measuring the output.

Retail innovation projects drive the development of control mechanisms
Innovation projects themselves serve as the driver for developing a greater structure of the innovation process through identification of needed control mechanisms such as innovation performance measurement.In the case of GR, the new innovation team in charge of developing new product concepts for private labels was given the time and resources to engage in an iterative learning process of developing and documenting the innovation process that is most suitable for their context, experimenting with KPIs that can enable them to enact their mandate effectively.In the case of CR, because the two major innovation projects on new retail formats were of a greater scale than usual development projects, with more resources involved, senior management acknowledged the need for greater discipline and clarity into the creative process of innovation.Planning and monitoring the innovation process through, e.g., codified playbooks and templates not just addressed the current projects but were established as new guidelines across the organisation.The study demonstrates how retailers are using innovation projects as the vehicle for experimenting and learning about how to structure innovation efforts and how to better capture value from the innovation process through performance measurement.

(Mis)alignment of performance measurement and strategic objectives
Although the study shows various innovation indicators being used in all the retail cases, the inherent difficulty of measuring innovation in retail was manifested as managers faced challenges in implementation.Tensions could be attributed to lack of alignment of metrics with the company's strategic objectives.An example is when sustainability is established as a key driver for innovation but is not being supported by the current system of performance measurement, including instances when inconsistencies in targets exist.In the case of GR, the innovation team was equipped with resources and competences to engage in explorative innovation, such as using new forms of external collaborations in innovation projects.However, there was a disconnect between some of the metrics that the team developed for themselves, which capture long-term value creation (e.g. in relation to sustainability goals), with the profit-oriented goals of internal stakeholders that are involved in the execution of their innovation projects.Hristov and Reynolds (2015) have reported on retail's tendency to prioritise measuring short-term financial benefits over other types of value, which is confirmed in this case study.The project managers who have an official mandate to contribute to innovation in the organisation still feel that they are held back in their mission due to challenges related to their respective companies' existing control systems in general, especially in relation to financial targets, such as in the case of GR and HR.Non-financial metrics that capture other types of value are used, such as those related to customer experience, customer satisfaction, brand and store image.However, short-term profitability is still of utmost importance, which can be argued to cause tensions when pursuing sustainable competitive advantage through innovation and sustainability.The findings suggest that when retailers pursue innovation that offers a broad, strategic impact such as new types of retail formats which encompass multiple application areas (e.g.offer-related, support-related, and organisation-related innovations), it is insufficient to only rely on existing performance measurements (financial indicators) that are used in current projects that are more incremental or operational in nature.This is consistent with Bedford's (2015) argument that flexible, interactive control systems and measurements are associated with firms that are pursuing exploratory innovation.
As companies try to make their innovation efforts less ad hoc and more systematic, the use of performance measurement as support to management is important because measuring the effectiveness and efficiency of actions increases the likelihood of achieving goals (Nilsson and Ritzén 2014;Neely, 2005).For retailers to strengthen their innovation capability -not least given the nature of the sector, where using metrics and analytics to drive profitability and growth is not unfamiliar -these indicators can be reviewed at both strategic and project levels to assure a clear understanding of the innovation strategy and that compatible metrics reinforce the strategy.Retailers can especially benefit from assessing the innovation portfolio balance as well as considering measures for willingness to invest in innovation, which are important indicators for organisations that have the strategic intent to become more innovative.

Developing innovation capability through performance measurement: towards a holistic perspective
There is consensus in extant research that companies should have a holistic perspective of innovation and its measurement.This entails not just measuring for the output, but also taking into consideration the whole process as well as the organisational context for innovation which contributes to an organisation's innovation capability.This includes having a coherent innovation strategy, showing commitment to innovation through structure, culture, competences, and resources for innovation, and utilising external sources of knowledge and innovation.In addition, it must be remembered that innovation is about risk taking and experimentation, also critical factors to mirror in measuring of innovation.These factors need consideration in the innovation strategy, and they are fundamentally taken care of at project level in organisations.An innovative organisation carries out innovation activities that pervade the operational activities in projects, that is, pervasive innovation actions are initiated, carried out, and consequently need to be controlled.
Given the exploratory nature of this case study, it is premature to conclude if retailers have established routines towards innovation-oriented learning.Nonetheless, our findings show that existing practices on performance measurement may support the development of innovation capability, especially in terms of culture and competence development as well as the opportunity to use external knowledge and engage in open innovation.We observed various forms of partnerships with external actors such as sustainability-related and technology-related start-ups, incubators and innovation platforms, in addition to retail's established proximity to end consumers, proving retailers' position as potential 'innovation hub' in the value chain (Sundström and Radon 2014;Hristov, 2007).The continuous use of qualitative, non-financial metrics related to culture, leadership, and employee well-being and satisfaction (e.g. through employee surveys), as well as access to diverse sources of innovation input -including market research and customer-related measures, external networks and collaborations -indicate that dimensions of innovation capability exist and can be further strengthened.This draws attention to the critical role of senior management in ensuring a holistic approach to innovation performance measurement, and innovation management overall.Ultimately, they are responsible in the strategic integration of performance measurement on all levels, assuring consistency and compatibility of metrics used across the organisation.As large retailers attempt to become mature, strategic, and systematic innovators, there is a need to clarify and communicate the organisation's commitment to innovation which ultimately hinges on management's vision and strategic mandate and implementation of holistic managerial control systems, including aligned performance measurements that contribute to the development of innovation capability.
Based on the theoretically built framework used in the case studies, and on the empirical data, we believe that the model, shown in Figure 1, may serve as an initial guideline for retail organisations in defining new measurements supporting the holistic perspective that is critical when building innovation capability.The model presents the elements of a holistic managerial control system for innovation capability in retail organisations and stresses the importance of establishing a variation of indicators that relate both to output and to input/processes in the organisational context.Also, the model emphasises that indicators of different kinds on a strategic level are most important to build a coherent strategy, while indicators on a project level are most important to build pervasive innovation actions.Both are critical in developing the innovation capability of the retail organisation.

Concluding remarks
This study sets out to investigate how retailers measure innovation, a topic that is increasingly becoming relevant in practice yet vastly under-investigated in research.Our multiple case study extends existing theory by offering a contemporary understanding of how retailers are adopting a greater control of the innovation process by implementing various ways in which innovation is monitored and measured.Our analysis sheds new light on how retailers use a variety of financial and non-financial indicators, capturing not just the innovation output but also assessing the inputs and the innovation process itself.Innovation projects also serve as the vehicle for experimenting and learning about how to better structure innovation efforts.However, retail companies are still challenged when there is misalignment between strategic objectives and the performance measurements used across the organisation.Nonetheless, existing practices on performance measurement may support the building and development of innovation capability in retail, especially in terms of culture and competence development as well as the opportunity to use external knowledge and engage in open innovation.The study also draws attention to the critical role of senior management in ensuring a holistic approach to innovation performance measurement, and innovation management overall.By synthesising insights from atheoretically built framework used in the case studies and the empirical data, the paper's contribution lies in the development of a model that accounts for how retail organisations could define new measurements that support the holistic perspective that is critical when building innovation capability.
There are study limitations that must be acknowledged, which open up opportunities for future research.The qualitative multiple case study approach using three large retail companies in Sweden was less aimed at the development of generalizable theory but rather to have an exploration of a vastly under-researched subject.Our investigation suggests that the connection between performance measurement and innovation capability in retail contexts offers an interesting and important research path and should be studied in detail.We suggest looking at how innovation measurement can be embedded into the regular practice of various retail functions to support organisational routines for learning and capability-building, a topic that could benefit from a multilevel and longitudinal research approach.Her research is focused on strategic innovation management in retailing and how large retail organisations can develop innovation capability through organisational structures, systems, and climate.
Sofia Ritzén is a professor in Integrated Product Development at KTH, the Royal Institute of Technology, Stockholm, Sweden.Her research focuses on innovation capabilities in organizations and innovation management in relation to sustainability.Focal areas of research have lately been how to use goals and metrics in order to achieve both radical and incremental innovations.Her research is empirically based with collaborations in manufacturing, retail and with public organizations.
Malin Olander Roese is a researcher and senior lecturer in Innovation Engineering, Faculty of Engineering at Lund University.Her research brings together firms' strategy with customer orientation and innovation capabilities, focusing on strategic change in various industries.

Appendix
Table A1.Checklist of innovation indicators.

STRATEGIC-LEVEL PROJECT-LEVEL Financial indicators
• Measures for willingness to invest in innova- tion.(e.g.share of research and innovation budget from total company budget, average expenditure per selected area) • ROI of innovation activities, sales and market share measures for innovation (e.g.new-tomarket and new-to-business-sales, profit growth from innovations) • Financial benefit projections (e.g.revenue targets for innovation, percentage of innovations that met targets) • Percentage of projects in total portfolio going through a defined project management system with defined milestones • R&D intensity (e.g.number of ideas/initia- tives going through a particular stage-gate) • For new products: number or percentage of new products or product categories within a retail product range • For new brands and retail formats: brand tracking and store image metrics • Employees involved in R&D and innovation initiatives, and the time invested in innovation initiatives compared to normal operational tasks • Planning and monitoring of the innovation process, e.g.ways in which the innovation process is formalised • Measures related to culture and communi- cation (e.g.creativity, entrepreneurial orientation, tolerance for failure, etc.).
• Measures for employee well-being and satisfaction, • Measures related to leadership (e.g.top management support, openness to change) • Knowledge and competence development, and innovation-oriented learning (e.g.support for education of employees, qualification of employees, staff training, number of employees or managers with innovation training, presence of innovation champions) • Feedback and reward systems • Mix of innovation sources: use of internal and external knowledge and information sources and collaboration, e.g.knowledge sharing activities with research institutes, network size (total size of strategic alliances), customer involvement (organisational support for collaboration with customer groups, user communities), percentage of innovation projects outsourced, assessment of supplier capabilities, etc.
• Quality of information for innovation, e.g.
market and technology research resources • Number of innovation projects started, on- going, and completed (includes conceptualisation projects for products, services, organisational process improvements, etc.) • Time measures (e.g.time for idea genera- tion in the project, speed to market (from ideation to commercial launch), time milestones to track market penetration and profitability) • Number of cross-functional initiatives and workshops started, on-going and completed • Number of ideas for new innovations (pro- duct/process/etc.) • Percentage of ideas found viable for commercialisation • Rate and quality of experimentation (e.g.number of prototypes per new product/ process/etc., number of customer tests before launch) • Degree of innovativeness, e.g.newness to company or market • Duration of the innovation output's life cycle (e.g. from product launch to delisting) • Customer insight measures (e.g.customer satisfaction) • Product quality and reliability (e.g.custo- mer evaluation, complaints rate) • Activities that enhance the brand image (e.g.number of awards and publications for the innovation) • Intellectual property rights (e.g.patents, citations, applications, licenses, trademarks)

Table 1 .
Details of case subjects.

Table 2 .
Innovation indicators used by case companies.
• Product quality and reliability (e.g.customer evaluation, complaints rate) ab

•
Percentage of sales protected by patents, trade secrets, other exclusive know-how • Project cost vs budget • ROI of the project outcome, sales and mar- Innovation portfolio balance (e.g.# of stra- tegic arenas to focus innovation efforts on, number of projects aligned with the innovation strategy, number of identified unmet needs and problems, number of new markets surveyed, annual portfolio review killing rate, % of projects closed by post-mortem)