Omnichannel retailing and post-pandemic recovery: building a research agenda

Purpose – The study aims to propose a comprehensive overview of the Omnichannel phenomenon by identifying its theoretical foundations as well as future research directions. Design/methodology/approach – InordertosystematizeOmnichannel-centeredcontributionsandidentify futureresearchdirectionsforpost-Covid-19,thisstudyadoptedamixed-methodstudy,combiningasystematicliteraturereview,abibliometricco-citationanalysisandapaneldiscussionbyfieldexperts. Findings – InStudy1,theauthorstracedextantliteratureonOmnichannelbacktoitstheoreticalfoundations, which led to the identification of four research areas in which the concept of Omnichannel is rooted. Contributions pertaining to the aforesaid research areas were discussed and submitted to a panel of experts (Study 2) after the lockdown periods. The experts gave various insights into both the past and future of Omnichannel research. Finally, a framework synthesizing theoretical foundations of Omnichannel, literature gaps and opportunities for future research is provided. Originality/value – To our knowledge, this is the first attempt to combine mixed methods study in Omnichannel research and to involve a panel of experts in order to discuss the findings of a literature review and evaluate future research directions. This choice allowed us to investigate both incumbent academic and managerial challenges raised by Omnichannel and to provide guidance for the post-pandemic recovery. in four


Introduction
In recent years, companies have been facing an increased complexity of the market environment, due to the continuous developments of digital technologies that have generated rapid changes in consumers' lifestyles and their relationships with brands and stores. Many companies have started developing Omnichannel strategies, a process that requires that they integrate and adapt their resources to face the emerging challenges such as maintaining Omnichannel retailing: a research agenda substantial consistency of brand values, attributes and overall image across the different services, experiences and channels provided (Payne et al., 2017) and developing seamless experiences for their customers (Von Briel, 2018). While before the pandemic Omnichannel was a differentiating factor, employed by companies in order to distinguish themselves from competitors by offering customers a wider range of services and experiences, the Covid-19 outbreak clearly increased its importance and potentiality (Verhoef, 2021).
The various social restrictions following the beginning of the pandemic, including the periods of lockdown, deeply affected the daily practices of companies and consumers alike (Farrell et al., 2020). On many occasions, retailers were forced to close their offline channel and urgently resort to delivery services, rethinking their online offer in order to continue their activities despite the crisis (Wang et al., 2020). At the same time, and out of necessity, many consumers were forced to change their behavior and habits, converting to online channels and touchpoints. Some customers experienced new services for the first time, giving rise to behavior that could be maintained after the pandemic (Arora et al., 2020). At the same time, exposure to online touchpoints may have increased consumers' technology readiness, the use of mobile devices and the adoption of social media. All these factors are extensively investigated in literature as factors that have widely impacted customers' perception of Omnichannel retailing (Hickman et al., 2020;Herhausen et al., 2019;Pagani et al., 2019).
Due to the need to integrate channels and develop Omnichannel strategies, companies today are facing the challenge of revising their operational, strategic and brand-related procedures in order to adapt to new circumstances (Picot-Coupey et al., 2016;Chen et al., 2018). Implementing channel integration and touchpoint alignment is, however, complex and risky (e.g. the subsequent emergence of channel cannibalization issues - Hansen and Sia, 2015).
From a theoretical point of view, as pointed out by Saghiri et al. (2017), Omnichannel is a complex concept with a very wide scope. This has led academic research to develop different streams of study that express different perspectives to analyze this phenomenon. However, such streams, ranging from management to consumer behavior, adopt different theoretical foundations, with different objectives, leaving scholars with a variety of contributions and the need for a comprehensive and consistent view of Omnichannel retailing. Considering such fragmentation, understanding the Omnichannel naturebased on academic literatureis, therefore, a primary aim of our research. We suggest that understanding this phenomenon through its mainstreams is necessary for redefining its boundaries and identifying present and future research directions necessary for post-pandemic recovery.
As a result, three research questions have been identified to guide our efforts: RQ1. What are the main streams of research with respect to the Omnichannel phenomenon?
RQ2. What are the theoretical foundations of Omnichannel research?
RQ3. What emerging opportunities for new research on post-Covid-19 Omnichannel can be identified?
To address these questions, we chose a mixed-method approach, combining a systematic literature review on Omnichannel conducted with bibliometric techniques (Study 1), and a panel discussion by field experts on the review results and emerging topics (Study 2) conducted after the recent lockdown periods. Study 1 identifies and compares the multiple research streams concerned with Omnichannel to uncover the intellectual foundations that each has employed to deal with the concept. Study 2 is the first, to our knowledge, to introduce a discussion with experts to validate and expand literature review findings and research directions concerning Omnichannel in the post-pandemic. The overall aim is to build a and customer loyalty in the long term (Srivastava and Kaul, 2016). The above examples  suggest the need to clarify the intellectual foundation of extant academic literature about  Omnichannel, to connect relevant topics and theories and develop a research agenda. A comprehensive list of Omnichannel definitions from the early days to the most recent conceptual developments is contained in Appendix 1.

Research design and methodology
This work adopts a mixed-method research synthesis (MMRS) approach (Sandelowski et al., 2012) which allows researchers to exploit the benefits of both qualitative and quantitative methodologies, for a deeper and more robust understanding of complex phenomena, identifying critical aspects and eventual discrepancies to obtain "more complete, concrete and nuanced answers [. . .] to complex research questions" (Heyvaert et al., 2013, p. 671). In this study, MMRS helped provide a rigorous desk analysis combined with a discussion with field experts with notable academic and managerial backgrounds, a methodology that has been successfully adopted by previous studies in marketing and management (Hall, 2011;Pohlmann and Kaartemo, 2017;Mortazavi et al., 2021). Table 1 reports the steps undertaken in the two studies (methodology and results), which will be further discussed in the following paragraphs.

Study 1: systematic literature review
The systematic literature review was conducted as follows: First, we identified a dataset of articles that represent the various fields of the academic debate on Omnichannel research. Then, we extracted the corresponding bibliographic metadata and applied quantitative techniques (bibliometric descriptive analyses), in order to set out the most relevant contributions and provide our qualitative interpretation (content analysis). 3.1.1 Literature search. The selection of the articles to be included in our literature review database was made according to the PRISMA diagram (Moher et al., 2009) to ensure effective reporting of sources. The data were retrieved from the electronic database ISI Web of Knowledge Core Collection (WoS). This source is considered "the gold standard for citation analysis" (Harzing and Alakangas, 2016, p. 791), as it provides standardized reference items and a rich set of  (Fetscherin and Heinrich, 2015) belonging to multiple disciplines (Merig o et al., 2015). The search was conducted by title, subject and abstract terms. The keywords used included the word "Omnichannel" in its different forms, such as "Omnichannel" and "Omni-channel" (with delimiters), to allow us to identify a wider set of contributions. Only articles with an available abstract and written in English were considered, and the dataset was then filtered by the exclusion of grey literature, non-academic literature and conference proceedings. The dataset output from WebOfScience was fully hand-checked by a member of the research group, while two other researchers revised the work. During this process, we harmonized the spelling of "Omnichannel" and also accounted for other relevant spelling errors in names and/or journals.
The research was conducted on the entire timespan available in the WoS database -1985-2020, up to the 20th of November 2020, when the data were extracted. At the end of the procedure, 314 articles had been identified, mainly pertaining to the WoS classification fields of "business," "management" and "operations research management science," and the corresponding bibliographic metadata were extracted and processed.
3.1.2 Bibliometric analyses. Bibliometric analyses were performed so that specific research fields could be identified and classified. The following three techniques were applied to the abovementioned 314 articles, with the purpose of identifying the most relevant contributions in the literature, to be subjected to a thorough qualitative analysis and interpretation: (1) A descriptive mapping of the collection of papers, using techniques based on the papers' metadata including authors' keywords, which required us to refine the database by harmonizing keywords' spelling. We first identified the evolutionary process of Omnichannel literature through the scientific production year by year, and the most influential papers, journals and authors. According to Ferreira et al. (2016), developing a map of conceptual frameworks employed with reference to a certain topic allows a holistic view of the topic itself to be obtained, the understanding of its relationships with other key research subjects/areas to be enhanced, and attention to be focused on emerging research gaps.
(2) A co-citation analysis. This is a statistical method that uses references as the main unit of measurement of affinity and proximity between papers to identify publications clusters that represent the intellectual foundations of a scientific discipline or school of thought (White and Griffith, 1981;Ramos-Rodriguez and Ruiz-Navarro, 2004;Galipoglu et al., 2018). Huber et al. (2014) highlight that this method has proved particularly useful in the marketing literature. Co-citation analysis is based on the concept of "citation frequency," a unit of measurement given by the total number of citations of a single document across all references in the sample (Garfield, 1979). Co-citation analysis calculates "citation frequency" by using pairs of papers to check whether both papers are cited in a third article (Aria and Cuccurullo, 2017). This process is then iterated for every paper in the sample. The analysis for this paper was conducted using a Louvain Clustering Algorithm that has been developed for group detection in very large networks (Blondel et al., 2008) and has been successfully used for clustering procedures in bibliometrics. Moreover, being a non-hierarchical algorithm, it aids discovery of the total number of clusters without pre-assumptions (Traag et al., 2019). Co-citation analysis also provides the betweenness centrality index for each paper of the different clusters. In information networks, betweenness centrality identifies which objects in the cluster represent "bridges" among other objects, namely, "nodes." In bibliometrics, papers with a high centrality index conceptually link theories and research contributions (Abbasi et al., 2012).
Analyses (1) and (2) were conducted with the support of "bibliometrix" bibliographic software, an R-based, validated tool for science mapping (Aria and Cuccurullo, 2017). This tool enables Omnichannel retailing: a research agenda analysis on three levels: "sources," "authors" and "documents," which correspond to the identification of "conceptual," "intellectual" and "social" structures of scientific disciplines, respectively (Aria and Cuccurullo, 2017). While the "conceptual" structure refers to the relationships between theoretical concepts, identified by the words used (i.e. keywords, titles and abstracts), the "intellectual" structure expresses the relationship between papers (and theories); finally, the "social" structure refers to the identification of authors and research groups devoted to a certain field. 3.1.3 Content-based analysis. The co-citation analysis yielded the most-cited papers and identified research clusters. A content-based analysis (Seuring and Gold, 2012) was performed on such papers to map theoretical developments and research directions within each cluster. Topics were identified per cluster, providing a deeper understanding of the research fields involved and their interrelations, as will be shown in Section 5. Topics were identified by the researchers separately at first and then compared in order to increase the reliability of the research (Kolbe and Burnett, 1991).
3.2 Study 2: discussion with field experts Results from Study 1 were discussed with a panel of experts, with the aim of gathering valuable insights about the major issues, most important theories and methodologies for future research in the post-Covid Omnichannel domain. Omnichannel has in fact become increasingly relevant for both academics and practitioners, in relation to the multiple changes driven by the pandemic in channel management (Verhoef, 2021) and consumption behavior (Guthrie et al., 2021).
The panel was selected to comprise academics with expertise in the Multi-, Cross-or Omnichannel topics worldwide. Data collection started in the post-lockdown period of April 2021 and was performed in two rounds by means of an online survey. For the first survey round, 17 potential participants were contacted; upon completion of the questionnaire, they were asked to suggest other experts' names. The second round involved 15 academics, for a total of 32 contacted experts. A total of 18 experts completed the survey, all of whom had past experience in Multi-or Omnichannel (2-10 years) and in retailing (11-20 years). They all work in different continents: Europe-Spain, Austria, UK, France, Belgium, Ireland, Switzerland and Finland are represented; two are located in Australia and New Zealand; one in the USA. Their studies involved both B2C and B2B, across retailing and consumer behavior. Five of them had published in academic journals almost exclusively, while the others had also written books and editorials for practitioners. Thirteen experts provide manufacturing and retail firms with their expertise regularly: eight had consulted on specific Multi-, Cross-or Omnichannel issues, such as channel integration, in-store customer experience, channel diversification and the development of e-commerce channels. They possess the necessary competencies to analyze bibliometric outputs from our research and discuss the theories and frameworks identified. Besides, because of their professional relations with retailers and practitioners, they are also able to understand their perspectives, needs and objectives.
The questionnaire was composed of six sections. In the first and second sections, the experts were asked to review a summary of the results that had emerged from Study 1, specifically on research clusters and Omnichannel theoretical underpinnings (e.g. Tell us which of the four you think should receive more attention from researchers for better understanding the phenomenon of Omnichannel distribution); also, they were asked to suggest which theories today could contribute the most to the field's development (What are the main reasons why so little theory has been used in the context of omnichannel research? In your opinion, what theories would be useful in helping to understand and investigate the phenomena IJRDM related to Omnichannel? Do you think that theories need to be developed for Omnichannel researchirrespective of which of the four clusters described above they pertain to?) In the third and fourth sections, the experts commented on the most appropriate methodologies and on industries and settings that warranted further attention. In the fifth section, they suggested which topics and research areas academics should focus on (What are the most important phenomena/themes/research questions thatfrom your point of view-need to be researched in the future? And why?). Finally, in the sixth section, based on their experience with companies, they provided a list of Omnichannel issues that were most pressing from the practitioners' perspective in the post-pandemic period (Based on your experience with companies across the supply chain, what Omnichannel issues "keep managers awake at night"? And which is the most pressing issue that needs solving from a practitioner's point of view?).
Two members of the research group separately examined and coded the collected data, solved issues with a third researcher and discussed the results with the research group as a whole.

Study 1 results: bibliometric descriptive analyses
It should first be noted that although the research was conducted on the entire timespan available in the WoS database -1985-2020, the output has a timespan of January 2011-November 2020 (Table 2). A conspicuous increase in the number of publications was identified in the period from 2016 to 2020.
Starting from 2011 to 2020 sample, we proceeded to identify the "core" journals and authors, most dedicated to the subject, by applying, respectively, the science mapping rules of Bradford's Law (Brookes, 1969) and Lotka's Law (Pao, 1985). Bradford's Law arranges journals in descending order of the number of articles they have published on a subject, thereby identifying three "zones." Journals in Zone 1 (Table 3)    We then identified the authors with the highest number of published papers on Omnichannel (Table 4) by applying Lotka's Law, which describes the frequency of publication by authors in a given field and distinguishes between "occasional" authors who have written only one article and "core" authors.
From this analysis, it emerged that about 86% of the authors had published only one paper on Omnichannel. In total, 119 authors had more than 1 paper, and the 12 most prolific (4, 5 or 6 papers) are identified in Table 5. The author that has contributed most to the field is Santiago Gallino (6 papers).
The most frequently cited papers were then identified. In Table 6, we show these 15 publications, alongside local citations (i.e. number of citations by other papers in the retrieved database) and global citation count (i.e. total number of citations received).
Finally, we examined the 314 papers to identify the most widespread research methodology and investigated settings. Of these, 82% feature one or more empirical studies and most studies employ quantitative methodologies (66%), mainly surveys and math's models (the latter in logistics and operations studies). Qualitative methodologies are applied in 33% of the papers, mostly case studies and in-depth interviews and 1% apply a mixed-method research design.
Overall, 53% of the Omnichannel papers included a focus on a specific industry, namely, fashion, apparel and accessories, food and grocery, multi-category retailing and consumer electronics. Services, both public and private, appear almost unexploredwith the exception of bankingas well as B2B and wholesale commerce. Most of the studiesespecially in the consumer behavior areause industries as a mere setting. For example, fashion, apparel and accessories or consumer electronics are often used for investigating the Millennial consumer segment. Other papers do not explicitly refer to an industry, such as in the logistics domain, where mathematical models are developed based on a hypothetical company's needs, Verlinde, Sara 4 9 12 Wollenburg, Johannes 4 5 IJRDM organization and assets. It must be noted that less than 1% of all the papers are multiindustry studies comparing behavior or strategies across sectors.

Study 1 results: co-citation analysis and content analysis on identified research clusters
In this section, we provide the result of the content analysis conducted on 50 papers, categorized into 4 research clusters, resulting from the co-citation analysis. Cluster 1 is centered on consumer behavior; Cluster 2 is labelled after the topics of strategy and management; Cluster 3 deals with Omnichannel from the perspective of channel management issues; and Cluster 4 is focused on channel integration. For each cluster, core papers were identified based on a high betweenness centrality index. Some of the clusters include papers focused on the development of research methods and procedures: following Gurzki and Woisetschl€ ager (2017), it is reasonable not to consider such papers in the analysis, to improve the interpretation of the results. The details of each of the 50 papersauthors, publication year, title and sourceare provided in Appendix 2, as well as the network and density visualizations of the four clusters.  Omnichannel retailing: a research agenda characteristics, drivers and effects on channel and company sales, in the short and longer run (e.g. Kumar and Venkatesan, 2005;Konuş et al., 2008). The need to assess the size and impact of the multichannel segment in the customer base and to develop effective multichannel management approaches such as customer relationship management emerges . Earlier works are centered on the offline-online dichotomy, while later works explore a new scenario in which the customer's purchase process becomes a dynamic journey across a wider range of channels and touchpoints. Consequently, effective management of the customer journey calls for an innovative approach, namely, customer experience management Lemon and Verhoef, 2016;.
Another common trait of the papers in this cluster is the attention paid to the managerial consequences of the behavior investigated, such as channel migration and cannibalization. The "reasons why" and implications of these studies show an explicit managerial focus. For example, studies centered on customer segmentation are strongly related to the managerial need to understandand eventually influencedifferent consumers' choice drivers and behavior across channels (e.g. Verhoef et al., 2007;Neslin and Shankar, 2009;Avery et al., 2012;Pauwels and Neslin, 2015;. As far as theories are concerned, in Cluster 1, various established consumer behavior theories are employed to explain the specific multichannel behavior that each paper addresses, e.g. theory of reasoned action, in Verhoef et al. (2007); utility-based models, in Gensler et al. (2012), Konuş et al. (2008) and Melis et al. (2015); social exchange theory, in Venkatesan et al. (2007). With their works, Lemon and Verhoef (2016) and Baxendale et al.
(2015) mark a moment of disruption. Rather than a new theory, these authors propose their novel approach based on touchpoints, which will represent the basis for many future studies.

Cluster 2: strategic management
Cluster 2 is concerned with management issues. It is composed of nine papers and has a mean year of publication of 2014. Papers in Cluster 2 question how to compete effectively in the new environment by combining channels, as is well described by Brynjolfsson et al. (2013), which is the main node of the cluster (centrality 15.78).
Channels are considered to be information and product fulfilment combinations (Bell et al., 2014) that allow for the identification of different Omnichannel solutions such as Buy-Online-Pickup-In-Store or "click-and-collect"  and showrooming Gao and Su, 2017b;Bell et al., 2018). The companies investigated in these studies face the transition from managing a dichotomous configuration (online-offline) to multiple combinations of channels and touchpoints. Differently from Cluster 1, here, the evaluation of the impact of such solutions is not constrained to channel or company sales but focuses on their effect on competitive advantage. It is within this strand, for example, that Ofek et al. (2011) address channel profitability, and  study the strategic role of salespeople as they deal with showrooming behavior.

Cluster 3: channel management issues
Cluster 3 is named channel management issues and includes 14 papers, 4 of which are literature reviews. The mean year of publication for this cluster is 2015. Papers in this cluster are concerned with channel management issues on two levels: on the one hand, there are works that address specific operational issues related to the company's adoption of a new channel, such as inventory planning (H€ ubner et al., 2016a), last-mile distribution (Bernon et al., 2016) and back-end fulfilment (Ishfaq et al., 2016); on the other, we find papers that aim to show the consequences of such adoption on a higher levelnotably, the impact on the entire organization of the retail firm of the choice to go multi-or omni-channel and the subsequent need to reorganize and integrate IJRDM processes and flows. Cao's work (2014), by showing that cross-channel provides more value but requires changes across several aspects of the retail business modelretail concept, flow management, HR organization and management and relationship managementis an example of the latter. Cao's holistic vision will be highly influential for many authors in the coming years.
Earlier works focus on the specific issue of adding a second channel in companies that have long been single-channel (online-to-offline and vice versa), while later papers try to address the greater complexity that has arisen since the onset of Omnichannel. The work of Beck and Rygl (2015) acknowledges such complexity by proposing a taxonomy of the various channel management situations that companies need to address. Ailawadi and Farris (2017) propose new metrics for gauging the performance of the new multi-and omni-realities.
It is worth noting that the majority of works in this cluster adopt an interpretivist approach by means of qualitative methodologies, such as Delphi studies, and single or multiple case studies. For example, Piotrowicz and Cuthbertson (2014), the central paper of the cluster (centrality 10.84) gather practitioners' opinions through focus groups, and the abovementioned 2014 work of Cao is based on a thorough construction of the case study of a Chinese Omnichannel retailer.

Cluster 4: channel integration
Finally, Cluster 4 is centered on channel integration, from the consumer perspective (channel integration perception) and from the management perspective (channel integration execution). It includes nine papers and has a mean year of publication of 2011. Studies in this cluster explore the impact of channel integration on key company outcomes, such as customer retention (Bendoly et al., 2005) and satisfaction (Herhausen et al., 2015), and highlight the central role of consumer perception of such integration. Several consumer behavior theories are invoked to explain how perceptions of integration are formed, e.g. Mental Accounting Theory, in Bendoly et al. (2005), the Unified Theory of Acceptance and Use of Technology, in Juaneda-Ayensa et al. (2016) and Technology Diffusion Theory (Herhausen et al., 2015). As far as integration execution is concerned, studies rely on resource-based view approaches (Cao and Li, 2015;Oh et al., 2012). This cluster stresses the role of information technology in both Omnichannel perception (Juaneda-Ayensa et al., 2016) and execution (Saghiri et al., 2017). In sum, works in this cluster suggest that the journey to Omnichannel requires integration on two sides simultaneously: back-end operations and front-end perceptions.

Study 2 results: the experts' perspective on omnichannel in the post-Covid-19
This section presents the key findings emerging from Study 2. Eighteen experts who took part in the study were asked to discuss various aspects of Omnichannel, also considering the developments emerged during the pandemic period. As Seetharaman (2020) points out, the pandemic has represented an opportunitythough forcedfor companies to develop new business models by integrating or abandoning channels or touchpoints, with the possibility of maintaining such changes in the long term. The Omnichannel transition has thus become even more strategic, due to consumers increasingly relying on the online and mobile channels during their journeys and being demanding of seamless experiences (Soto-Acosta, 2020). The results will be displayed according to the topics addressed in the questionnaire: validation of the Omnichannel research clusters; theories and theory development on Omnichannel; research designs and settings for future studies; and major challenges and needs of Omnichannel research.

Validation of Omnichannel research clusters
The experts were presented with the research clusters that emerged from Study 1, which represent Omnichannel's theoretical foundations -Consumer Behavior, Strategic Management, Channel Management Issues and Channel Integrationand were asked to Omnichannel retailing: a research agenda choose which of them should receive priority in the post-pandemic society as an area for future research. Interestingly, 11 experts chose Channel Integration, as it is the key to seamlessness. It is considered a "basis for superior customer value" and "to create long-term sustainable business models". Channel Integration embodies the "holistic perspective to Omnichannel marketing" and it can be the key to integrating the four research areas: "understanding Omnichannel demands seamless, highly integrated approaches" [. . .] "academic and practitioner knowledge needs to consider an integrated and holistic approach to the topic." Consumer Behavior in Omnichannel environments is considered a "basis for strategy development and/or redesign," and it is noted that "companies have only a limited understanding of complex customer journeys." The experts also suggested that Omnichannel is an interesting setting in which to investigate recent consumer behavior disruptions: "we need to understand what is changed with the pandemic, especially for the young segment." As for Strategic Management, the experts suggested that "understanding companies' issues is crucial to provide effective managerial implications," and that a stronger understanding of companies' perspective also has implications for academic research: "understand how management conceptualize the undergoing changes, in order to propose valid theoretical frameworks." Overall, the comments on the Channel Integration, Consumer Behavior and Strategic Management areas are in line with what emerged from the analysis of theoretical foundations (Study 1, sections 5.1 and 5.2). Finally, very little attention has been devoted to Channel Management Issues.

Need for theory development in Omnichannel
The experts provided several explanations for the fact that, in the Omnichannel domain, theories are scarcely used, thus confirming our findings (Study 1, Section 5). Specifically: (1) Recency of the Omnichannel phenomenon, originated by rapid and disruptive changes in marketspaces that contribute to quick obsolescence of frameworks.
(2) Interdisciplinarity of Omnichannel, which requires interconnection among different disciplines (i.e. HR, information management, business informatics and data analytics), some of which "marketing scholars are not familiar with." (3) Practitioner orientation of Omnichannel. Since retailing is mostly practitioneroriented, Omnichannel research is perceived as "very descriptive and operational in nature." (4) Omnichannel as a mere setting: A few experts suggested that Omnichannel may be a mere condition triggering certain behavior on the consumer side.
(5) Research design and data collection issues due to Omnichannel complexity. Researchers may face issues in investigating more than two channels at once and in measuring their integration. Also, it may be difficult to obtain Omnichannel data, quantitative or qualitative, from companies.
Omnichannel does not need new theories, according to six experts. Existing theories are perfectly fit, on the company side (e.g. resource-based view, transaction cost theory and principal-agent theory) as well as on the consumers' (e.g. technology acceptance model, goal theory, information processing theories and congruence theory). Among the 12 experts who believed that specific theories should be developed for Omnichannel, 5 suggested that these theories should emerge from each of the research areas in which Omnichannel is rooted; the others, instead, called for a general theory of IJRDM Omnichannel that would consider the interactions among all the systems and actors involved. For example, one suggested that holistic research approaches be adopted "to provide an exploratory and inductive understanding of the complexities of 'Omni'." The absence of a consensus on whether Omnichannel needs theories of its own confirms the multiple perspectives that could be adopted to address this topic.

Designs and settings for Omnichannel research
The experts were presented with findings from Study 1 referring to the adoption of methodologies in the Omnichannel domain (par 4.) and asked to offer suggestions of methodologies for future studies. Almost all the experts suggested that qualitative methodologies are preferable for exploring Omnichannel phenomena, given their novelty; mixed methods were recommended, too. Among the quantitative methodologies, the researchers suggested field experiments and big data analytics.
As regards the research settings, the experts suggested that the following industries should be considered for future Omnichannel research: home furniture, luxury, automotive and personal care. Among services, health services, hospitality and tourism, construction and education were mentioned. The experts also proposed that B2B and C2C relations be analyzed, the latter "since consumers are engaging in Omnichannel selling and purchasing of second-hand goods from each other." A few experts suggested that more interesting than an industry focus was a retail business model perspective: "Most research has a focus on brick-and-mortar retailers, while pure players transitioning to omnichannel models are of specific interest and scarcely studied."

Major challenges and need for Omnichannel research
The experts were asked to cite the most promising directions for future academic research and crucial Omnichannel issues for practitioners, especially considering the post-Covid period and the changes it has elicited. Their suggestions can be arranged under five themes: 6.4.1 Evolution of the customer journey. In Omnichannel contexts, researchers' attention today should be devoted to understanding the customer journey. Research should include the design and redesign of journeys through the selection and management of touchpoints. The pandemicand its consequent restrictionshave proven how circumstances can alter customer journeys, even preventing customers from accessing certain touchpoints or channels and forcing them to explore new alternatives. Also, as most customer journeys are complex and extended, they may include different actors than the retailer, and steps such as complaints and returns: "this process [. . .] from attitude formation to after-sales behavior starts months before an actual purchase and has not really been investigated." Studying the customer journey is regarded as well-suited for the integration of competencies from different research areas, involving consumer behavior, operations, and logistics.
6.4.2 Channel-related consumer behavior and experience. It is crucial for companies to understand how Omnichannel influences consumers, in order to effectively manage the experience it offers. In this context, sub-topics of interest are customer loyalty and engagement, Omnichannel service quality and Omnichannel brand management. The pandemic, has, in fact, further exacerbated free-riding behavior across channels, such as webrooming and the quest for convenient prices. The experts suggested further research on pure digital players, to understand how they address customers' needs, which in turn questions the role of physical stores. They suggested the question of "what contexts (products, markets, geographic and cultural contexts) enable or constrain good and bad consumer experiences in an omni-world." Studies addressing the consumer experience with Omnichannel might try to answer the question: do consumers really need Omnichannel? Or, conversely, does becoming Omnichannel for a company necessarily affect consumer perceptions positively?
Omnichannel retailing: a research agenda 6.4.3 Omnichannel strategy implementation. Omnichannel companies are facing the challenge of implementing efficient and effective information systems that are crucial for internal and external data management. It is a huge and risky investment. Information systems are intertwined with last-mile logistics and delivery issues; the experts highlighted the extreme complexity of managing operational activities, especially in the post-pandemic period, and the need to implement customer satisfaction measurement and to reach a high level of personalization in the delivery service. Today service recovery issues have also become strategic. They should be considered from both the logistic efficiency and the customer experience perspectives, since returns and/or delivery failures can disorient consumers and activate entirely different journeys. Related to the above issues is the emerging need to have constant control over the Omnichannel performance. Specifically, companies should develop metrics dedicated to measuring each channel's contribution to performance and thus reconsider operations internal and external costs. Assessing the economic viability of Omnichannel may bring about the possibility of some companies finding it non-viable or unprofitable. The experts pointed to the challenges of increased competition and related price pressure, which were further reinforced in this post-Covid period. They described Omnichannel as "a high-transparency context," in which it is difficult to make a profit when prices are being driven down; also, competing in such a context requires "huge marketing and management costs for the company's digital presence." Finally, the experts suggested attention should be paid to the role of channel partners, and the tradeoffs Omnichannel companies must face when deciding whether to outsource or internalize processes.
6.4.4 Human resources management. The experts consider the increasing relevance of Omnichannel as an opportunity for an organizational renewal of companies, in a continuous effort to "innovate while maintaining the core business." First, Omnichannel today calls for the definition of new roles for HR. HR contributes to the customer experience and is crucial for service quality perception: new skills are required, and companies must hire accordingly or train personnel specifically to accomplish those goals. Second, managing HR in an Omnichannel context also means companies must "balance staff across physical and digital channels, optimizing staff deployment." The experts also posed the question of which organizational cultures and leadership techniques might be more effective in coordinating and motivating HR to have a positive impact on "omni-success." 6.4.5 Digital transformation challenges. The experts suggested that, while studying the Omnichannel phenomenon in the post-pandemic period, it would also be strategic to address issues related to the adopting of new technology.
On the one hand, researchers and practitioners need to understand the opportunities and threats of Artificial Intelligence, Augmented and Virtual Reality, Marketing Automation software and the use of drones in logistics and delivery services. On the other, it is necessary to find a balance between physical assets and digitalization, as a basis for implementing the "Phygital" business model. The experts also mentioned that they were considering critical privacy and legal issues related to technology developments, the possibility of exploiting big data for management and marketing purposes, and the increasing role of data science in retailing.

Discussions of results and stimuli for Omnichannel research
This study aimed to shed light on the current status of Omnichannel research and its foundations by uncovering the most promising issues for future research in this area. Based on the co-citation and content-based analysis (Study 1) and on the survey of a panel of experts (Study 2), several insights emerge (Figure 1).

IJRDM
With reference to RQ1, Omnichannel is found to be a complex phenomenon that is rooted in four key areas: Consumer Behavior; Strategy and Management; Channel Management; Channel Integration. This highlights the variety of contributions and intellectual foundations that emerge around Omnichannel, while also underlining how the different research perspectives of Omnichannel studies are interconnected. There is no widely accepted definition of Omnichannel, and many studies are still focused on the difference between multichannel and cross-channel, which emerged as antecedents of Omnichannel, the latter representing a further phase within an evolutionary process related to technological advancements. Channel Integration appears to be the most promising area for future research in the Omnichannel domain. In fact, channel integration is a necessary requirement for developing Omnichannel systems that are capable of offering seamless services and experiences for consumers. Studying the role of Channel Integration connects both the consumer and company perspectives.
As far as RQ2 is concerned, theoretical developments in the Omnichannel domain are found to be rather scarce. No theory has been developed specifically for Omnichannel and the theories employed are usually existing ones adapted to the specific phenomenon under study. This opens a debate on whether Omnichannel should merely be considered a new setting that calls for new testing of existing theories or a new domain that deserves specific theoretical developments. Experts are divided on this. Those who deem new theories necessarily think that new theories rooted in the different intellectual foundations of Omnichannel are more likely to emerge than a comprehensive general theory of Omnichannel.
Finally, from the analysis of papers in the Omnichannel domain and from the knowledge shared by the interviewed experts, we can develop a research agenda based on the five pillars identified (RQ3), in this post-Covid-19 phase which can guide practitioners and academic researchers. For each pillar, we have identified relevant research questions that are listed and briefly discussed below.
Omnichannel journeys: with the role of the physical store shifting from point of sale to point of experience, and with the emergence of new online sales channels such as livestream shopping and conversational commerce applications, it will be challenging for companies to understand today how to keep their brand identity and values when integrating all the different channels so as to offer a seamless but authentic experience. Specifically, designing Omnichannel journeys might require companies to develop ongoing interactions and coordination with a substantial number of channel partners and service providers. How can companies manage such a complex network of relationships Omnichannel retailing: a research agenda and/or control these partners'/providers' actions? What are the consequences when "the price companies must pay" to be Omnichannel is to relinquish control to the hands of more powerful players (e.g. platforms)? Given such complexity, on the one hand, it would be of primary importance to understand whether a high degree of channel integration implemented so as to deliver an Omnichannel experience is in fact beneficial for all companies. On the other, further studies should investigate whether the effects of channel integration might vary across different types of customer journeys (long versus short journeys, journeys with service failures) and across subgroups of customers (e.g. generational cohorts, brand detractors vs. brand lovers, deal-prone consumers and different cultural subgroups).
The evolvement of the Omnichannel customer experience: there is a need to explore the longitudinal dimensions of the customer experience, to appropriately investigate the role of time in touchpoint exposure. For example, do Omnichannel customers display Omnichannel behavior in every situation? Is there an Omnichannel "fatigue" to such an extent that Omnichannel behavior can be abandoned? Are some consumer segments more likely to adopt (or abandon) Omnichannel behavior (e.g. deal-prone consumers) than others? Do friends or relatives acting as shopping companions influence the experience with touchpoints? These are relevant questions for understanding how to achieve customer engagement and customer loyalty in Omnichannel.
Omnichannel transition issues: it has become important to understand whether the transition from single channel to Omnichannel is a gradual process that entails going through a multi-channel phase or instead if a direct leap is preferable, and from what perspectives. Are barriers and obstacles to the Omnichannel transition specific or common to bricks-andmortar retailers and online pure players? Does it pay off in the long term to become Omnichannel? How can Omnichannel success be measured with reference to the short versus long term? Moreover, Omnichannel is not simply a matter of introducing (or not) certain services such as BOPS: it is rather a problem of defining appropriate thresholds in the delivery of these services. What should guide the setting of such thresholds? Besides, Omnichannel is also a matter of the brand transitioning to digital. How can companies accomplish this while preserving the brand identity? How might the alignment between brand identity and brand image in the Omnichannel environment be measured? Companies should understand when it is beneficial and when detrimental to involve channel partners and service providers in the deployment of the Omnichannel transition. If channel partners are to manage several phases of the customer journey, how can companies display a consistent brand image from the pre-purchase to the after-sale stage?
The human factor in Omnichannel: both employees and consumers need to be trained to navigate the Omnichannel experience. Salespeople and customer service representatives are important human touchpoints that can be leveraged to educate consumers and point them towards taking the company-designed seamless journey. What training and what incentives are most effective in the Omnichannel employee-customer interaction? The human components will be key to enriching the value and the social dimension of the experience that the customer is living within a certain channel in order to prevent competitive showrooming or webrooming: is this so in any channel and setting? As far as customer education about Omnichannel is concerned, the role of content marketing and personalization could be further explored.
Augmented and intelligent Omnichannel environments: the disruption brought by emerging technologies and their increased use by consumers is attracting attention to how Artificial Intelligence and Augmented and Virtual Reality will change the customer journey and the environment in which it takes place. Focusing on the Omnichannel environment requires the adoption of a holistic approach, which takes into account the implementation of new technologies by companies and their adoption by consumers. Special attention should be IJRDM devoted to identifying the dark side of applying new technologies that have the potential to negatively influence consumer privacy and security.
Research methodologies and settings: exploring the abovementioned research directions might require the adoption of new approaches and methodologies. The novelty of Omnichannel environments calls for rigorous qualitative methodologies with which to explore the emerging issues and to be combined with quantitative methodologies in mixedmethod designs. Employing field experiments or big data analysis is also suggested so that we can better understand how customer purchase behavior and customer engagement are influenced by Omnichannel. As far as the settings are concerned, many industries are still unexplored, especially B2B contexts and services such as health, hospitality and tourism and education.
Omnichannel has been found to be a growing and promising field of research. This paper has attempted to summarize the research directions and gaps and to foster relevant, farreaching research.
The major limitation of our study concerns the methodology used in the selection of the literature to be reviewed. In order to focus on the Omnichannel phenomenon, we only selected papers explicitly referring to it; as recently observed, Omnichannel is strongly related to other concepts such as "customer journey," "channel integration," "customer experience," etc. It might, therefore, offer an opportunity to trace back to how studies dedicated to such topics can contribute to the Omnichannel domain. Nevertheless, this study opens various avenues of fruitful research in Omnichannel for practitioners and academics in the post-pandemic recovery. Online/in-store integration and customer retention

Journal of Service Research
Cao, LL. and Li, L. 2015 The impact of cross-channel integration on retailers' sales growth

Journal of Retailing
(continued )