Out of the fog: fog computing-enabled AI to support smart marketing management

Abstract

Marketing and consumer research use a variety of data and electronic measurement devices for research, theory-building, and applied decision-making. Managing data deluge produced by ‘smart devices’ and internet of things (IoT) actuators and sensors is one of the challenges faced by managers when using IoT systems. With the advent of the cloud-based IoT and artificial intelligence, which are advancing a ‘smart world’ and introducing automation in many application areas, such as ‘smart marketing,’ a need has arisen for various modifications to support the IoT devices that are at the center of the automation world, including recent language models like, ChatGPT and Bart, and technologies like nanotechnology. The article introduces the marketing community to a recent computing development: IoT-driven fog computing (FC)—an emerging concept that decentralizes operations, management, and data into the network utilizing a distributed and federated computing paradigm. Although numerous research studies have been published on ‘smart’ applications, none hitherto have been conducted on fog-based smart marketing. FC is considered a novel computational system, which can mitigate latency and improve bandwidth utilization for autonomous marketing applications requiring real-time processing of ‘big data’ typical of smart marketing ecosystems.

1 Introduction

Fog computing is considered a formidable next-generation complement to cloud computing (Sofla et al. 2022).

Society is moving toward a fourth stage of human development and is currently in the midst of a critical transition from information society to intelligent society (Alshurideh et al. 2023; Sen 2023). In this transition, computing and immersive technologies are constantly undergoing transformative, even disruptive, changes. This development follows Industry 4.0, with its focus on intelligent or smart devices. Smart devices involve data analytics capabilities empowered with technological advancements such as internet of things (IoT), machine learning (ML), deep learning (DL), artificial intelligence (AI), virtual reality (VR), big data analytics (BDA), and eventually Intelligent Device-free Sensing (IDFS), and nanotechnology (Huakun et al. 2024; Shah et al. 2024), all of which will become an eminent part of Industry 4.0. These technologies, which range from automation and computation of marketing systems to autonomous and hyperconnected marketing systems, where human input is not required (Swaminathan et.al. 2020), promise to revolutionize marketing as we know it.

Computing is regarded as a critical driving force in the development of human systems. Due to the enormous volume of marketing data, human analysts are unable to adequately uncover and compute useful information for marketing decisions. Recent developments in intelligent computing are expected to precipitate a major breakthrough not only in intelligence-oriented computing but also in intelligence-empowered computing (Caruelle et al. 2022; Puntoni et al. 2021). Intelligent computing provides autonomous, scalable, efficient, reliable, universal, secure, and transparent computing services to support large-scale and complex computational tasks in today’s ‘smart world’ (Liu et al. 2019; Simões et al. 2018). It has greatly broadened the scope of computing by extending it from traditional data computing to increasingly diverse computing paradigms, such as perceptual intelligence, cognitive intelligence, autonomous intelligence, affective computing, and human–computer fusion intelligence (Alshurideh et al. 2023). With the expansion of IoT and wide proliferation of wireless networks, the number of edge devices (e.g., sensors, scanners) and the massive data generated from them have been growing fast. Also, many services, such as online gaming and image processing rely on AI to achieve the required smartness and autonomy. Above all, the digital world is advancing rapidly, and developments in networking technologies, such as low-power wide area networks (LPWAN), 4G long-term evolution (LTE), wireless broadband (WiBro), 5G, and nano datacenters (NaDa or nDCs), are leading to the emergence of sophisticated data services. Billions of devices, ranging from user gadgets to more complex devices, are generating massive amounts of data for research and applications. Pervasive usage of smart, interconnected devices is estimated to reach 58.2 billion units by 2025 (Costa et al. 2022). This exponential growth is nourished by the proliferation of mobile devices (e.g., mobile phones and tablets), with smart sensors serving different vertical markets, such as smart cities, smart healthcare, and smart marketing.

Typically, data generated by IoT devices are transferred to and processed centrally by services located on distant clouds. However, with the dawn of IoT technology, which traverses the internet to the cloud where it is saved, several challenges have emerged (Das and Inuwa 2023; Hassan and Rashard 2023). These include but are not limited to issues related to high latency, loss of reliability, network congestion, poor quality of service (QoS) and quality of experience (QoE), and so on (Hazra et al. 2023). In other words, although current devices are getting smaller and more powerful in terms of features and capabilities, they are still incapable of executing autonomous and intelligent tasks such as those often required for smart applications like marketing. To address these challenges, advanced technologies and software are needed to manage the growing systems and ‘big data’ of IoT services. Clearly, the ideal place to analyze most IoT data is not the remote cloud but rather closer to the devices that produce and act on that data. For this reason, the IoT-driven fog computing (FC) platform supported by software-defined networking (SDN; Al-Shareeda et al. 2024; Kumhar and Bhatia 2022; Nunez-Gomez et al. 2023; Tran-Dang and Kim 2023) was introduced to bridge the distance and address some of the aforementioned challenges. The term “fog computing”—derived from the phrase “fog is a cloud that is closer to the ground”—was coined by Cisco Systems to describe the need for computing capabilities at the network edge to address the challenges posed by the increasing volume and velocity of data generated by IoT devices. Thus, FC aims to reduce latency, conserve network bandwidth, improve efficiency, and enable real-time processing and analysis of data. In November 2015, an alliance of industries and academia, including Intel, ARM, Microsoft, Dell, Cisco, and Princeton University, launched the OpenFog Consortium to introduce and promote the use of FC. By 2018, FC had become a leading platform for developing IoT frameworks (The Economist 2018) among governments and academic institutions. At the time, the OpenFog Consortium Architecture Working Group (2017) defined FC: “A system-level horizontal architecture that distributes resources and services of computing, storage, control and networking anywhere along the continuum from Cloud to Things.”

1.1 Study objectives

Inspired by recent successful applications of FC to various smart ecosystems, in the present study we suggest future applications to smart marketing and other marketing domains. To advance these objectives, precisely during an era in which marketing scholars are calling for more conceptual work (e.g., MacInnis 2011; Sridhar et al. 2023, Vargo et al. 2020), we first provide an overview of FC and related concepts relevant to marketing. Second, we elaborate on intelligent/smart marketing. Third, we present some illustrations of FC marketing applications. Fourth and finally, we identify a number of fundamental research challenges for marketing in order to fully exploit FC for marketing management and research. This conceptual study goes beyond a literature review by offering compelling observations of marketing in the real world. Employing a multi-perspective approach, it aims to deliver valuable insights into FC offering perspective on several impacted marketing areas. Positioned as an important contribution within the emerging FC-focused literature, this article, we hope, will serve as a valuable reference for marketing scholars and practitioners and foster future theoretical and practical research innovations in FC-driven marketing. To assist readers, in Web Appendix Table 1, we provide a glossary of key terms used in the article.

2 Fog computing

Nowadays, the pointer of the trending paradigm is pointing at fog computing. As efficiency and quality of service stand as important objectives in the world of computing, it is viewed as a promising application (Das et al. 2023).

Just as the natural phenomena of fog are perceived when a cloud is close to the earth, FC was conceived to describe the computational principle that aims to bring the benefits of the cloud closer to end-devices or users. For this purpose, it is considered a highly distributed and decentralized computing model, combined with the cloud, but with processing also being executed at the edge of the network, facilitating operation of applications not previously feasible due to the high latency in communication between the cloud and a vast multitude of devices (Hazra et al. 2023; Li et al. 2023; Tomar et al. 2023; for a recent review, see Das et al. 2023; Srirama 2024). Compared with macro-data cloud centers, the fog consists of lightweight micro-cloud or nano-datacenters (m/nDCs). Therefore, FC presents the same functionalities that the cloud offers to users but with greater proximity and scope. The usual fog network rests on a cluster of heterogeneous interconnected sensors and devices used for computation, communication, and storage for latency-restricted IoT applications. In this way, FC provides a practical platform of device-to-device connectivity that enriches local processing and networking services between the different data centers and end-users of early cloud computing technology. In FC, edge devices have computational and storage capabilities that allow them to perform data-processing tasks locally. These edge devices can filter, preprocess, and analyze data before sending relevant information to the cloud for further processing or storage. By doing so, FC reduces the need to transmit large volumes of data to centralized cloud servers, which can be resource-intensive and cause delays (Vambe 2023). FC is particularly useful in scenarios where low latency and real-time processing are critical, such as industrial automation, smart cities, autonomous vehicles, healthcare monitoring, video surveillance, and smart marketing (Hussein et al. 2023). Thus, the FC paradigm offers fundamentally different solutions to computational issues and enables more efficient problem-solving than what is possible with traditional computations. By leveraging FC, managers can harness the benefits of both local edge processing and centralized cloud computing, creating a more efficient and responsive computing infrastructure (see Fig. 1, with some elaboration in Web Appendix A). In an FC system, multiple servers can cooperate with each other securely (e.g., by utilizing a blockchain platform) to serve terminal devices and thereby improve utilization of fog servers (Ometov et al. 2022). Notably, what is happening undoubtedly with FC is that the static, single-cloud management model is evolving into a completely new model, which is much more dynamic and variegated, and located on different, heterogeneous, and usually mobile fog premises, furnishing distinct services that can be linked together, as well as capacities that can be jointly offered (for a comparative analysis of cloud computing and FC, see Web Appendix B). In sum, FC provides the following major benefits for marketing applications: 1. Lower latency: By processing data closer to the source, FC reduces the time it takes for data to travel to the cloud and back, resulting in lower latency and faster response times. 2. Bandwidth efficiency: FC reduces the amount of data that needs to be transmitted to the cloud by processing and filtering data locally. This helps conserve network bandwidth and lessens the costs associated with transmitting large amounts of data to the cloud. 3. Improved reliability: Since FC distributes computing tasks across multiple devices, it provides redundancy, and improves overall system reliability. Even if some devices fail, others can continue to function, ensuring continuity of service. 4. Privacy and security: By keeping sensitive data local and performing computations at the edge, FC offers enhanced privacy and security compared to transmitting data to the cloud. It enables data to be processed and analyzed without leaving the local network. 5. Multimodal integration: FC has the ability to combine and integrate an array of data-generating modalities and information from IoT, AI (including language models such as ChatGPT), text, images, audio, video, databases, software, collaborative filtering, and more, in comprehensive, accurate, and decision-support outputs. Indeed, newer generative AI algorithms have evolved, which can process data in their natural form, thus enabling mining of unstructured data, such as raw text and images. Moreover, methods in AI, including deep learning and reinforcement learning, have been developed in which specific algorithms, such as convoluted and recurrent neural networks, have gained prominence for being able to analyze images, audio, and even video. Thus, FC-enabled AI offers the opportunity to combine resources and modalities into a hybrid digital data system (Hegarty and Taylor 2021)., which supplies comprehensive information for various ‘smart’ applications.

Fig. 1

FC major benefits

2.1 Fog computing architecture

FC architecture (Fig. 2) is comprised of highly dispersed and diverse devices with the intent of facilitating IoT applications that require computation, storage and networking resources allocated at different geographical locations. FC is a layered construct for facilitating a ubiquitous approach to a shared continuum of scalable computing resources (Costa et al. 2022; Manzoor et al. 2022). For example, the business layer receives the user-based applications from the application layer and creates various service models, graphs, and flow models for handling the data efficiently and effectively, while the device layer establishes the relationship between various sensors and actuators for receiving the data from the environment, and performs specific tasks based on the upper-layer commands. Nowadays, the system expedites the use of distributed, latency-aware services and applications, and consists of fog nodes (physical or virtual), located between smart end-devices and unified (cloud) services. Figure 2 depicts FC in the larger situation of a cloud-based ecosystem facilitating smart end-devices.

Fig. 2

FC architecture

This architecture is able to process data in a matter of milliseconds. Of course, the connectivity options vary by use case. An IoT sensor in a store, for example, can likely use a wired connection. However, a mobile resource, such as an autonomous vehicle, or consumer-grade GPS device, will require an alternate form of connectivity. The architecture of an FC environment in iFogSim consists of multiple layers, with each layer responsible for specific tasks to assist higher-layer operations. In the architecture, the bottommost layer consists of IoT devices, which interact with the real world and comprise data sources. IoT sensors act as data sources for applications and are scattered over different locations, sensing the environment (e.g., the market) and emitting observed data to the upper layers through gateways for additional processing. The iFogSim library can be downloaded at https://github.com/Cloudslab/iFogSim. Since it is written in Java, the Java Development Kit (JDK) will be required to customize and work with the toolkit. The iFogSim library can be executed on any Java-based integrated development environment (IDE), such as Eclipse, Netbeans, JCreator, JDeveloper, jGRASP, BlueJ, IntelliJ IDEA or Jbuilder (Web Appendix C). After downloading the Zip file, the compression toolkit is extracted, and a iFogSim-master folder is created. AbdElhalim et al. (2019), for example, have used iFogSim to stydy the latency of user fog nodes offloading tasks in a very dense fog-to-cloud computing ecosystem. In their study, the authors introduced adaptive low latency networks, irrespective of the number of user nodes, based on game theory. Game theory was used to reduce energy consumption on the fog nodes as well to maintain an acceptable latency. They also developed automated distributed FC for computational offloading, applying the theory of minority games. Resource allocation was performed in a distributed manner for improved network management. Their simulation results suggested that the proposed algorithm reached the user satisfaction latency deadline as well as the required QoE. Moreover, it guarantees an adaptive equilibrium level of the fog-to-cloud computing system, which is relevant for heterogeneous wireless networks deployed in smart marketing. At present, IoT edge devices, through a plethora of technological innovations and disruptions, are getting smaller, while gaining in memory, storage, networking capacities, and processing capability important for marketing analytics. In this way, fog devices are apt to become part of mainstream marketing computing. For the time being, however, FC remains a concept in development, and its full realization has yet to be achieved. For marketing analytics, it is important to explain the major elements of the fog computing architecture—the fog nodes.

2.2 Fog nodes

The FC layer consists of several nodes, which are responsible for advancing transactions between different elements and the cloud system. They are located closer to the data source and have higher processing and storage capabilities. As such, they can process data far more quickly than would be the case were they to rely on the cloud for centralized processing. Fog nodes are either physical units (e.g., routers, gateways, servers, switches, etc.) or virtual units (e.g., virtual machines, virtualized switches, cloudlets, etc.), which are connected to the smart end-units and facilitate transmission of computing resources to them. Fog nodes act in essence as small databases for routers, and as an activity and network monitor to dispatch control and configuration information to the network for more efficient performance in data transmission, caching, and security. Since FC is an extension of the common cloud-based computing model, deployment systems important for marketing, such as private fog nodes, community fog nodes, public fog nodes, and hybrid fog nodes (more in Web Appendix D), are also supported (Li et al. 2023). Therefore, due to their distributed capabilities, most real-time and delay-intensive applications should prefer to offload their computation data in a fog device. To facilitate new network innovations while drastically simplifying FC network operations, software-defined networks (SDN) have been developed (Ahmad et al. 2020; Nunez-Gomez et al. 2023).

2.3 SDN-based fog computing

Fog computing embraces a flexible software architecture with the capacity to consolidate different design choices and user-specified guidelines. Software-defined networking-based FC is a type of computing architecture that combines SDN with FC. This approach enables the creation of a highly distributed network infrastructure capable of providing low-latency, high-bandwidth connectivity to a wide range of devices and applications. SDN, as a software architecture, helps solve the IoT-FC heterogeneity problem, enabling the formation of independent protocols, and addressing issues related to restricted hardware and proprietary software (Kumar and Bhata 2022). Unlike traditional IP systems, in which control and data planes are strongly coupled and embedded in the same system elements, SDN-based FC separates the control plane from the data plane and system elements (Das and Inuma 2023), thus allowing service management to be optimized and service needs supported from a centralized user interface (UI), all of which affords greater agility and programmability, as well as the capacity to add network automation (Ahmad et al. 2020; Anoushee et al. 2023). Figure 3 provides an abstract sketch of this three-plane architecture in which the data plane is based on smart IoT devices that sense and produce the data, while the control plane acts as a brain for the core system because it has a complete picture of the system as a whole. The SDN controller and the system devices communicate with each other by the southbound API. The application plane can include many different FC applications such as smart marketing. The application and control planes, on the other hand, communicate with each other by the northbound API. Due to advanced technologies, FC is using a rapidly growing number of devices, while computerized models like SDN help to provide higher service quality (Kumhar and Bhatia 2022).

Fig. 3

SDN-based IOT architecture

Ahmad et al. (2020) recently introduced a system to address the benefits of energy-saving by combining SDN and the fog environment in one system. To this end, they utilized a dynamic programming scheme to optimize the selection of fog nodes. In their proposed architectures, SDN is located between the cloud and fog layers. The authors showed the proposed method’s capacity to minimize the energy consumed by the fog system while maintaining service level agreement parameters. For example, a smartwatch may collect data related to an individual consumer’s physical activities, including heartbeat and pulse, and transmit it to their smartphone, whereupon an SDN-based FC system can process the data and make suggestions to the consumers regarding how to adjust their activities, diet, and lifestyle according to their physical condition. By combining SDN with FC, SDN-based FC offers marketing analytics the following benefits: 1. Network orchestration: SDN enables centralized control and orchestration of network resources, thus allowing marketers to dynamically allocate and manage network resources for marketing applications. This includes managing bandwidth, traffic prioritization, and quality of service (QoS) for different marketing services and applications running on fog nodes (Ollora et al. 2020). 2. Dynamic service deployment: SDN allows managers to dynamically deploy marketing services and applications across fog nodes. By leveraging SDN controllers, marketers can programmatically allocate resources and provide services based on demand and network conditions. This lends agility in deploying marketing campaigns, scaling up or down resources as needed, and optimizing service delivery. 3. Traffic-steering and optimization: SDN-based FC holds tremendous potential in terms of intelligent traffic-steering and optimization for marketing purposes. In this context, SDN controllers can analyze network conditions, traffic patterns, and user behaviors to dynamically route traffic to the most suitable fog nodes for improved performance and user experience, thus ensuring that marketing content and services are delivered efficiently and with low latency. 4. Network security and policy enforcement: SDN provides enhanced network security and policy enforcement capabilities. With SDN controllers, managers will be better able to enforce security policies, implement access controls, and monitor network traffic for potential security threats. SDN’s programmability enables rapid response to security incidents and the ability to apply security measures uniformly across the fog computing infrastructure (Ahmad et al. 2020). 5. Service-level agreements (SLAs): SDN-based fog computing is able to facilitate the establishment and enforcement of SLAs between marketers and network service providers. In this regard, SDN controllers can monitor network performance metrics and enforce SLA requirements (Ollora et al. 2020), thus guaranteeing that marketing services meet agreed-upon levels of performance, availability, and reliability. 6. Network virtualization: SDN-enabled network virtualization promises to allow marketers to create virtual network slices or overlays for specific marketing services or campaigns, which will permit isolation and customization of network resources, thus ensuring that marketing applications operate independently while sharing the underlying FC infrastructure efficiently. Overall, this kind of system can change the manual operation of repetitive marketing processes and replace it with performance-oriented application software built for specific marketing purposes, such as smart marketing.

3 Smart marketing

The ‘smart’ concept emerged out of the evolution of information and communications technologies (Liu et al. 2019). Indeed, ‘smart’ is often applied in contemporary culture as a prefix to technological terms to express special qualities of intelligence and/or connectivity, as in ‘smart phone’ or ‘smart card,’ that is, technologies that operate with little, or no human influence (Chang et al. 2023; Panato and Timmermans 2019). An important driver of a ‘smart’ business ecosystem, for example, is a basic infrastructure that enables a fast connect and disconnect, or ‘pick, plug and play’’ approach. However, ‘smart’ has entered into buzzwords in many facets of life: ‘smart grid,’ ‘smart city,’ ‘smart tourism,’ and ‘smart product’ are just a few familiar terms. Actually, ‘smart’ for our purposes comes from the acronym for “Self-Monitoring, Analysis and Reporting Technology,” implying it with terms such as “digital, internet, online” (Simões et al. 2018).

Smart marketing (Fig. 4) is considered an important evolution of marketing, which is expected to drastically alter the way in which consumers engage with marketing as we know it. Panato and Timmermans (2019) hold that in marketing the term ‘smart’ represents all things embedded in or enhanced by technology. Accordingly, whenever data are collected from different sensors, actuators, and machines within a marketing environment, and access to and control of the data and the devices generating it are enabled through the internet, such a scenario may be termed a ‘marketing internet of things’ (MIoT). The MIoT in this sense will focus mainly on the transfer and control of mission critical information and responses and rely heavily on machine-to-machine communications (Taylor et al. 2020). Recent developments in smart marketing include AI language models such as ChatGPT, Google’s Bard, and Meta’s LLaMA, all of which will provide new marketing data and new ways in which people interact with computers and each other (Liu et al. 2023). Language models with FC-enabled AI offer great advantages to marketing in the form of richer, faster, and more automated access to data, simpler and more efficient processes, greater accuracy and cost effectiveness, and enhanced recommendation systems (Paul et al. 2023; Peres et al. 2023). In such a reality, an enormous amount of marketing data of a highly diverse nature is apt to be generated. Encompassing cyberphysical systems (CPS), extreme automation, actuators, etc., the MIoT as a whole will be working steadfastly toward making the system faster, smarter, securer, and more robust. Indeed, MIoT is expected to enhance several methods at the level of augmented reality, chatbots, gamification, transmedia, wearable technology, etc., which will undoubtedly influence smart marketing. Hence, it is clear that ‘smart marketing’ encompasses all digital marketing tools: internet, mobile, social network, augmented reality, wearable technology, etc. (Sridhar and Fang 2019). Digital marketing has made available to researchers unprecedented data on firm and customer behavior, including structured data, e.g., numerical data on consumer purchasing behaviors and on firms’ digital marketing interventions, and unstructured data, e.g., text, audio, or even video content from consumers and firms. With the advent of the digital era, marketers have been better able to harness the data necessary to acquire, expand and retain their customer bases (Sridhar and Fang 2019). The smart marketing mix, which includes e-product, e-price, e-place, e-promotion, e-process, e-people, and e-physical evidence, is considered an influential tool in revolutionizing marketing management (Chang et al. 2023; Hoffman et al. 2022). In sum, smart marketing can incorporate sundry technologies, data analysis, and customer-centric strategies to deliver targeted, personalized, and efficient marketing campaigns, ultimately enhancing customer experiences and driving business results. Some argue that ‘smart’ implies ‘intelligence,’ because smartness can only be realized when an intelligent system adjusts to the needs of its users (Sharma and Dash 2023). Below are some key characteristics or elements of smart marketing: 1. Data-driven decision making: Smart marketing relies on collecting and analyzing data from various sources to gain insights into customer behavior, preferences, and needs. By understanding customer data, marketers can make informed decisions, create targeted campaigns, and deliver personalized experiences. 2. Personalization: Smart marketing emphasizes tailoring marketing messages and experiences to individual customers. This involves using customer data to segment audiences and create personalized content, offers, and recommendations. Personalization helps increase engagement, improve customer satisfaction, and drive conversions. 3. Automation: Smart marketing leverages automation tools and platforms to streamline marketing processes and improve efficiency. Marketing automation enables the automation of repetitive tasks such as email marketing, social media posting, and campaign management. This frees up time for marketers to focus on strategy and creative initiatives. 4. Multi-channel marketing: Smart marketing recognizes the importance of reaching customers through multiple channels and touchpoints. This involves creating a cohesive and consistent brand experience across various digital channels, such as websites, social media platforms, mobile apps, email, and search engines, as well as offline channels. Multi-channel marketing helps increase brand visibility, engagement, and customer loyalty. 5. Customer journey optimization: Smart marketing focuses on understanding and optimizing the customer journey, from initial awareness to final conversion and beyond. This entails mapping out the customer journey, identifying touchpoints, and optimizing each stage to provide a seamless and engaging experience. By understanding customer needs at different stages, marketers can deliver relevant content and experiences, which drive conversions and foster customer loyalty. 6. Performance tracking and analytics: Smart marketing relies on measuring and analyzing key performance indicators (KPIs) to evaluate campaign effectiveness and make data-driven improvements. By tracking metrics, such as website traffic, conversion rates, customer engagement, collaborative filtering, and return on investment (ROI), marketers can assess the success of their marketing initiatives and make informed decisions for future campaigns. Overall, smart marketing combines technology, data, and customer-centric strategies to optimize marketing efforts and achieve desired outcomes. 7. From human-controlled to autonomous operations: Marketing systems are quickly transitioning from human-controlled processes to closed-loop control services supporting their operations autonomously using extensive sensor and computing infrastructure. This revolutionary change is critical for supporting growing data-intensive and time-sensitive smart marketing applications. 8. Hyperconnectivity: The proliferation of networks of consumers, devices, and other entities, as well as the continuous access to other consumers, machines, and marketing entities, regardless of time or location is referred to as hyperconnectivity (Swaminathan et al. 2020). In this environment, information is always accessible and abundant for marketer and consumer decision-making. 9. New offerings: The advent of digital technologies means that marketers can augment their capabilities with new product offerings, or by creating matching platforms, which bring buyers and sellers together. Smart marketing strategy deals with firms’ judicious use of smart resources to create differentiated and sustainable value for customers.

Fig. 4

IoT smart marketing

3.1 FC applications

With the FC-enabled IoT “which seeks to support the dream of smart world, it is no longer enough to have only quality of service as it does not satisfy the user experience. As such, quality of experience has now become a key to get user’s satisfaction regarding reliability, availability, scalability, speed, accuracy, and efficiency” (Vambe 2023, p. 2).

The adoption of FC has increased rapidly due to recent developments in IoT smart devices and technologies (Al-Shareeda et al. 2024; Hazra, et al. 2023). However, FC is constantly modified, and the number of applications administered in a fog platform is expected to grow (Li et al. 2023; Tomar et al. 2023). The new structure and use cases of FC are strongly influenced by an array of innovative applications, which demand additional platform attributes that can only be delivered if the applications are deployed closer to end users, as in ‘smart vehicles,’ ‘smart supply chain systems’ (Gupta and Singh 2022), ‘smart healthcare systems’ (Patient Monitoring 2022, http://www.fogguru.eu/tmp/OpenFog-Use-Cases.zip), ‘smart banking,’ and more (Web Appendix E). In January 2023, Intel acquired FogHorn Systems, a leading provider of FC software in a deal that will help Intel to expand its FC offerings and provide businesses with a more comprehensive suite of ‘smart’ solutions. Meanwhile, shortly thereafter, in March 2023, Dell Technologies acquired Nimbix, a provider of FC solutions for the telecommunications industry (smart telecommunications), thus enabling the former company to meaningfully expand its FC offerings in telecommunications and meet the growing demand for FC in this market (https://www.skyquestt.com/speak-with-analyst/fog-computing-market).

4 FC and marketing

This section pulls the different elements of our perspective together, emphasizing key insights for marketers and marketing scholars. It is clear that FC can play an important role in almost all aspects of marketing, thereby making marketing ‘smarter.’ To monitor marketing activities and ascertain consumers’ cognitive, affective and behavioral attributes, marketing currently employs various obtrusive (surveys, experiments) and non-obtrusive data-generating methods and systems. Modern marketing routinely captures ‘big data’ through vast amounts of observations across individual subjects, stores, managers, brand SKUs, time periods, and predictor variables, producing massive multi-faceted datasets. For example, Amazon and AliExpress have accumulated consumer behavior data on millions of items selected, including consumer ratings, and geodemographic characteristics. Similar datasets have emerged recently in retailing with potential use of RFIDs (radio frequency identifications), online auctions, social networking sites, product reviews, collaborative filtering, customer relationship marketing, internet commerce, and mobile marketing. The data are collected by means of scanners, monitors, mobile devices, smart cards, eye gaze trackers (EGTs), odor measurements (e.g., Wang et al. 2023), semantic analyses, psychophysical devices such as fMRI, computerized haptic measures (Hannafeld and Okamura 2016), voice analog scales (VAS), computerized visual analog scales (VAS), vocal monitors (Hinkle et al. 2019), wireless body sensors, biometric authentication—and in the near future, mobile robots and nanodevices, as well. Importantly, research into the use of physiologic sensors, with skin conductance, heart rate, respiration, blood pressure, ECG, EEG, EMG and other capabilities, to measure emotion is increasing (Yu 2023). Various new and unobtrusive devices, which promise to considerably alter precision marketing, are presently being tested (Huakun et al. 2024). The advanced technologies embedded in these devices are able to provide computerized scalable digitalized data, some employing remote physiological measurements (Lu et al. 2023), and others able to track behaviors over time measuring “streak” behaviors performed consecutively (Silverman and Barasch 2023). Most smartphones and some smartwatches can accommodate ambient light sensors, internal motion sensors (accelerometers), gesture sensors, gyroscopic sensors, temperature and humidity sensors, magnetometers, barometers, etc. Communication interfaces commonly found on smartphones include WiFi, GPS, Bluetooth, near field communications (NFC), and infrared (IR) LED (Lee et al. 2023). The widespread diffusion of multiple wireless technologies, especially high-speed 5G networks, will further optimize the real-time sensing and collection of massive repositories of spatiotemporal data, which represent proxies for consumers’ sentiments, interactions, communications, and activities. With the assistance of decision software, data will be converted into information for management consideration and decision-making. It seems that virtually any human sensation can and will be monitored and scaled along four basic dimensions—intensity, quality, extension, and duration. FC is able to successfully integrate a vast spectrum of mental and physiological modalities and measurements, which are expected to deliver more precise information for identifying emotion and behavior—thus, creating value for market stakeholders by building a sustainable differential advantage relative to competitors. This immense amount of digitalized data (‘big data’) obtained from smart devices can be offloaded subsequently to the fog for storage, processing, aggregation, and computation, as depicted in Fig. 5. Users then will be able to shift this data to the optimum place for processing, and decisions can be based on how fast results are needed. For instance, time-sensitive marketing decisions will be made closer to the things producing and acting on the data. In contrast, big data marketing analytics dealing with historical information might need the computing and storage resources of the cloud. The capacity to triage marketing data and make critical judgments within the device’s own context will aid in extracting essential insights from the massive volume of available marketing data. For example, if millions of consumers around the world at the same time want to play a particular song on a music streaming service and that song is on a server in the United States, then processing that request would create a long que, slowing service to consumers. Fog servers, which are located close to the devices requesting the song, are able to stream that song immediately to the edge device with minimal latency. The fog node maintains a copy of that song, enabling other consumers in that same geographic area to also stream the song instantaneously.

Fig. 5

traditional cloud and fog enabled MIoT

Marketing-monitoring models can be designed with the help of MIoT and FC to minimize latency in data processing. As aforementioned, data-sharing rules and authentication are executed automatically based on smart modules and node verification. Data transmission is protected, thus minimizing the risk of data tampering in the public cloud environment. Consequently, just as FC is able, with considerable success, to treat and synthesize digital data with decision support software for smart cities or smart healthcare, monitored by some of the same devices employed for smart marketing, it should be capable of bringing together and ‘hybridizing’ various smart devices and associated big data for smart marketing and marketing analytics purposes. Moreover, it is reasonable to expect that FC-based solutions similar to those already being deployed in healthcare industries will soon be used for delivering quality of service (QoS) to consumers. Integrating MIoT and FC is also likely to improve operational efficiency, optimize power consumption, and minimize operation costs. In sum, although FC is emerging as a scalable, reliable, and cost effective solution for big data analytics, especially in the IoT domain, it has not been applied, so far, to marketing domains. Drawing on previous FC contributions to various smart domains, the present article identifies potential novel applications to marketing, as well as questions requiring further research for successful adoption in marketing domains (Ding et al. 2024).

4.1 FC applications to smart consumer behavior

Digital marketing strategy has fundamentally altered the marketing landscape, as it not only enriches the tools that marketing employ in customer relationships but also empowers consumers by providing them with unprecedented opportunities to interact with firms (Sridhar and Fang 2019). Because access to information is much easier in a hyperconnected world, consumers need not expend much effort to learn about products and services. They are using smart devices, which teach them how to manage their homes, travel, health, banking, distance education, and lifestyles. Cloud computing and artificial intelligence (AI) are enabling consumers to provide feedback via IoT devices, and to extract multiple benefits from a vast array of products ranging from LED lights to refrigerators to sophisticated lawn mowers and combine harvesters. Moreover, as devices turn smarter and AI and robots enter the marketplace, thus exposing consumers to the risk of being hacked, FC-enabled blockchains with encrypted text will secure their digital assets (Sen 2023). Also, consumers’ experience will be enriched with FC in such a way that they will be able interact in a more immersive manner with products and stores. Aliyu et al. (2023), for example, recently proposed a basic model of an FC-based smart shopping (SMSH) system, which is designed to recommend stores and products to customers according to their preferences, and to improve the shopping experience and customer loyalty. The system uses three parameters to optimize the shopping experience: (1) the distance between the shopper and the store, (2) the number of people in the desired store, and (3) whether the retailer offers a discount. The authors found that the proposed system provided up to 30% in savings and 6.12% faster shopping. However, they admitted that their model is very basic, and that to fully exploit FC, it must include other parameters influencing consumer behavior.

Contemporary consumers are immersed in a strong emotional culture, which permeates their behavior. Emotion is pivotal for understanding consumer preferences and behavior. This is being achieved in part through sentiment analysis, and more recently AI to detect consumer polarity. The latest developments in digital systems, including algorithms for emotion recognition (Caruelle et al. 2022), are being used to provide estimates of autonomic behavior. Researchers and technologists have been exploring ways to enhance digital emotions and create more immersive and authentic emotional experiences by deploying affective computing (for a recent review, see Lee et al. 2023, and Web Appendix F). FC is expected be able to compute unobtrusive easy-to-use technologies to provide objective information about sentiments that consumers might not report accurately. For example, consumers frequently receive personalized product recommendations from automated algorithms when using Amazon, Spotify, or Netflix, services where consumers usually browse the options and indicate their attitudes and emotions through the items they select. The FC recommender systems can utilize this input to generate other product recommendations. Future measures will be based on data collected from diverse sensors within, for example, smartphones, smart watches, wearable technology, and nanotechnology (Mileti et al. 2016), all with the aim of enhancing consumer satisfaction and welfare, and facilitating smarter shopping behavior.

4.2 FC-supported supply chain management

Smart supply chain management (SCM) refers to the application of advanced technologies, data analytics, and automation to enhance the efficiency, visibility, and agility of supply chain operations. Smart SCM focuses on enhancing collaboration and integration among supply chain partners. This can involve the integration of systems, sharing of real-time data, and collaborative forecasting to improve coordination and responsiveness. It leverages digital transformation and innovative solutions to optimize various aspects of the supply chain, from procurement and production to logistics and distribution. Gupta and Singh (2022) have proposed an approach for improving supply chain performance based on the use of FC to mitigate latency issues. The authors claim that increased competition and customer expectations have negatively impacted supply chain performance, operational costs, and the environment, while severely compromising safety, quality, traceability, and transparency. They show how FC, as a cloud extension, can provide support to supply chain operations through efficient resource utilization, enhanced performance, risk identification and reduction, bandwidth utilization, energy-consumption tracking, and above all real-time monitoring. Otherwise, they contend that the heterogeneity of available platforms and services makes all these supports difficult. Innovative FC, MIoT, and SDN, however, infuse novel approaches to assimilate, store, and share transactional data through collaboration and interoperability among supply chain stakeholders. The authors show how FC is able to support real-time data processing with the help of time-sensitive applications. Indeed, FC-based SCM platforms can provide centralized data storage, collaboration, and accessibility across supply chain partners, while IT facilitates real-time data sharing, improves communication, and supports scalability and flexibility. Musa and Vidyasankar (2017) have used blackberries as a case study to show how FC brings efficiency and reduces wastage in perishable SCM. Specifically, the monitoring FC system consists of, for example, fog nodes deployed in the field to provide real-time monitoring of fruit. In transit, the smart readers and truck are used at this stage as a mobile fog node. At the packing house, data from the fog nodes are used to determine the priority of pallets/cases for pre-cooling and also for shipping to the retail distribution center. At the distribution center, the environmental parameters to be monitored include temperature, humidity, carbon dioxide, and light intensity. In sum, FC can assist smart SCM to create a more agile, resilient, and customer-centric supply chain, which can adapt to rapidly changing market conditions. By leveraging advanced FC and analytics, stakeholders will be able to optimize inventory levels, reduce costs, improve delivery times, and enhance overall supply chain performance to maximize customer and supplier satisfaction.

4.3 FC-supported smart retailing/stores

Retailing is witnessing a transformation due to rapid technological development. Indeed, retailers are using smart technologies to establish smart retailing, to improve the consumer shopping experience, and to stay competitive. Smart retailing leverages various advanced technologies, such as the IoT, AI, ML, data analytics, and mobile applications, to create a more connected and personalized shopping environment (Pantano and Timmermans 2019). Some retailers have invested in introducing self-service technologies, such as self-cash desks, informative touch points, interactive displays equipped with touch screens, digital signage, and applications for mobile phones supported by radio frequency identification (RFID) tags. Others have introduced smart shopping carts equipped with edge devices, such as sensor modules, small but efficient processors for image processing and machine learning, cameras, smart screens, barcode scanners, charging ports, and speakers, which can collect and process data, model the customer’s context, and offer personalized shopping recommendations. Quickly processing the collected data at the server and responding with, for instance, aisle and shelf information, promotions, and product suggestions based on past purchases, makes for a better customer experience and greater satisfaction. Other retailers have developed entirely virtual stores where consumers can use their phones to locate and purchase products (Chang et al. 2023). Pantano and Timmermans (2019) emphasize that “the emerging idea of smart retailing would reflect a particular idea of retailing, where stores and consumers use technology to reinvent and reinforce their role in the new service economy, by improving the quality of their shopping experiences” (p. 30). Likewise, according to Dr. Mekel-Bobrov, eBay’s Chief AI Officer (2023), eBay uses generative AI “to enrich the online shopping experience to make online shopping less transactional and more experiential, more like shopping at a physical store, to create live streamed events, and to deliver personalized experiences to create higher engagement and more direct interactions with buyers.” However, eBay may consider using FC because “Fog computing's ability to process data closer to the source could potentially be beneficial for large-scale e-commerce platforms like eBay, especially in scenarios where real-time data analysis and decision-making are critical” (Tran-Dang and Kim 2023, p. 65). Some researchers (Bourg et al. 2021) link smart retailing to the broader concept of smart cities by focusing on a new approach to retail management, which considers technologies as enablers of innovation and improvements in consumers’ quality of life. The idea is that the same FC paradigm used in smart cities can be applied to smart retailing. For example, FC can provide retailers with real-time visibility into their inventory. By deploying edge devices or sensors in-store or in warehouses, they can monitor stock levels, track product movement, and receive immediate alerts for low-inventory or out-of-stock situations. This will allow proactive inventory management, minimize stockouts, and ensure optimal product availability. By capturing information about customer preferences, purchase histories, and behavior, retailers will be better equipped to deliver personalized recommendations and promotions in real-time. This will enhance the customer experience by providing tailored and relevant interactions, both in-store and through mobile applications. Additionally, FC will enable the implementation of smart shelves equipped with sensors or RFID technology. These shelves can detect product presence, monitor stock levels, and even analyze customer interactions with the products. FC will enable retailers to perform data analytics and generate insights at the edge of the network. By processing data locally from surveillance cameras, sensors, or point-of-sale systems, retailers will gain real-time insights into customer behavior, foot traffic patterns, and store performance. This information can be used to optimize store layouts, improve product placement, and enhance overall operational efficiency. The new IDFS is going to leverage various types of signals, such as Wi-Fi signals, radio frequency signals, sound waves, or even changes in ambient light, to infer information about the presence, location, movement, and interactions of consumers within a specific retail environment. Moreover, FC can be leveraged to enhance payment systems in retail stores. By deploying edge devices or terminals capable of processing payments locally, retailers will be able to reduce transaction latency and enhance payment processing security. Additionally, FC will permit new payment methods such as mobile wallets, contactless payments, and biometric authentication, providing customers with convenient and secure payment options. Finally, by deploying edge devices with advanced processing capabilities, FC can power interactive displays and product catalogs, as well as virtual try-on experiences. Integration and synchronization of digital data through FC will facilitate all of the above opportunities, thus enhancing customer engagement, improving the shopping experience, and assisting customers in making informed purchase decisions. Bourg et al. (2021) have introduced the SMARTBUY ecosystem to realize the concept of a “distributed shopping mall,” which allows retailers to join forces and unite in a large commercial coalition that generates added value for both retailers and customers. The SMARTBUY ecosystem provides two services: a geo-location marketing service, which takes advantage of a cloud-enabled city-scale cluster of WiFi apps, and a recommendation engine, which considers a multitude of contextual factors including time, season, demographic data, behavior, and user location history to accurately match like-minded customers and derive meaningful, personalized product recommendations. The FC-enabled SMARTBUY ecosystem can convert physical smart city stores into an open, geographically distributed mall by providing the logical consistency needed to conduct centralized searches over independent physical stores. These are just a few examples of how FC can be applied to smart retailing. In sum, by leveraging FC capabilities, retailers will be able to revamp their traditional brick-and-mortar business model to create technologically revolutionized stores. The latest FC-monitored smart technologies will provide tailored, context-specific suggestions and offers to customers via interactive screens or smartphones as they walk around the store. Similarly, FC will allow retailers to introduce into the store cutting-edge immersive experiences capable of genuinely transforming how customers shop.

4.4 FC-supported smart advertising

Smart advertising refers to the use of automated technology and algorithms to bid, buy, sell, and design advertising in real-time. It leverages data about customers’ demographics, interests, browsing behavior, and other relevant information to deliver personalized ads to specific audiences. It is regarded as advertising with intelligence, which aims to create highly positive customer experiences (Seyghaly et al. 2021). Seyghaly et al. (2021) have shown how the AROUND system, which is supported by IoT sensors, will assist in personalized advertising. Specifically, multiple Bluetooth low-energy beacon devices will be deployed in different locations in commercial shopping malls to track customers and offer personalized interactive functionalities to advertisers. The AROUND system is equipped with a smart recommender function, which provides individual data like, purchasing history, current mood, time spent in a given shop, or current location. All positioning data from the beacon is detected by the user’s app installed in their smartphone, and transmitted to the cloud. The authors suggest using FC “as a best option for providing a more effective, yet efficient, advertising system” (p. 8). Bulkan et al. (2023) have introduced the SuperEye, a smart advertising insertion for online platforms, which they consider an “eye-brain correlation.” The SuperEye system is powered by ML and AI, which will allow predictions about the interplay of emotions and inter-person attention differences to be tested at each point in time during exposure. The authors claim that customer engagement rates for advertisement can generate better results in terms of correct audience targeting and enhanced QoE, if the system is used with other smart algorithms. Recently, Guo (2023) has argued that smart advertising can be improved by enhancing the visual aesthetic effect. They propose an objective quantitative scoring method for image aesthetics based on a multi-scale feature extraction network, which is a computational intelligence approach. The framework consists of various “classification-based aesthetic quality evaluation[s]” based on a quantitative neural network. Guo (2023) concludes that such an approach can improve smart advertising by including in the system a deep neural network (DNN) algorithm, such as facial recognition, as well as tracing software, for effective advertisement. Another application for evaluating smart advertising is the ad display thumbnail pattern, which detects face and eye responses to each frame to improve ad design (Murtadho et al. 2019). Advertisers can supplement this application by including FC-enabled programmatic platforms or demand-side platforms (DSPs) to define their target audience and set parameters for their campaigns, such as budget, ad formats, and desired outcomes. For example, Netflix could personalize a movie poster based on individual viewer preferences by, for instance, changing the actors and headline appearing on the poster. The FC paradigm could assist in individualized advertising by including emotionally appealing images, such a favorite celebrity, or a certain type of wording, or background music, or background color, to suit the specific consumer. All of this information can be gathered from different sources, which suggests that two individuals with different preferences living in the same room might not see the same ad, even if they are searching for the same product. Personalized ads such as these will be distributed digitally to target consumers via websites, social media, or apps, depending on which they engage with more (Seyghaly et al. 2021). Also, ad effectiveness will be monitored to measure consumer engagement, that is, to ascertain how many consumers are following the link to find out more about the offer, or are sharing it, or are ignoring it altogether. All of these insights may be used to gain a better understanding of consumers and improve the development and effectiveness of ads. Also, FC can provide computer-implemented methods and computer-readable media configured to target advertisements based on emotional states (Murtadho et al. 2019).

To advance campaign effectiveness, FC can deploy conversion tracking and attribution, a platform that offers robust conversion-tracking capabilities to measure the effectiveness of ad campaigns (Seyghaly et al. 2021). Similarly, FC is capable of utilizing smart display software to design campaigns, where, for example, Google Ads automatically selects the best combination of attributes (images, headlines, descriptions) to create engaging display ads (Varmarken et al. 2020). With the assistance of Fog enabled ML algorithms, advertisers will be able to reach the right audiences and achieve campaign goals. They will also be able to provide emotional tags, indicating the desired emotional state of users who should see the advertisements linked to the tags. By assessing, on the one hand, emotions and behavior, through a combination of automated facial expression detection to yield a sample of pretested advertisements, and on the other hand, concentration of attention, through eye tracking and viewer retention, by, for example, recording zapping behavior, ad effectiveness can be honed moment-to-moment to concentrate viewer attention and retain engagement. Identifying consumer emotional trajectories can also inform pretesting practices by focusing advertisers and agencies on which emotional aspects of single frames in advertisements need improvement, and what the gains of potential improvement are likely to be. Conversely, it can provide information about the specific moments that trigger consumers to lose concentration and/or to zap. All of this can be achieved by deploying software for automatic emotion recognition from facial expressions, such as eMotion, FaceReader (Noldus), OKAO (Omron Corporation), and GfK’s (Growth for Knowledge). Advertising can also be made smart. For instance, FC can allow giant digital screens to be updated from a central location. In the same way, in shopping malls and supermarkets, advertising can be more personally customized. Similarly, through location tracking, fog-based smart advertisement and promotion systems will be able to place digital screen ads and coupons in relevant aisles with customers viewing promotions and suggestions according to their past purchase histories and current traces. Impact of TV Program Ratings and Commercials: Using gesture recognition systems (GRS) and voice/speech recognition to monitor consumers’ verbal comments while viewing TV programs and commercials, edge devices (e.g., mobile phone) will be able to obtain consumer gesture information, including program and volume changes (Varmarken et al. 2020), as well as speech input corresponding to a given program while performing the given gesture. The system will transfer the speech signal to the laptop, which works as fog device, to recognize the program, the commercial and volume. Finally, the fog device transfers the result to the edge device. The incremental learning model monitors and learns the consumer’s watching behavior via the edge device (Guo 2023). Such incremental learning will help to improve the accuracy of program rating measurement and commercial viewing behavior with the possibility of identifying which ads stand out from the clutter. In sum, the challenge inherent to these applications is to make sense out of the massive amount of data and use it appropriately to gain insights, predict behavior, and make management decisions. FC can integrate and synchronize all applications and software to provide the necessary information for smart advertising, which will in turn increase the effectiveness of ad campaigns, optimize ad spending, and deliver more targeted, relevant, and engaging ads to consumers. The immersive attributes of FC can assist in designing optimal ads beyond sight and sound but the combination of all senses (e.g. sensory and virtual) for a more dynamic consumer experience. It will offer advertisers greater flexibility, control, and insights into their campaigns, leading to improved targeting, better return on investment (ROI), and enhanced overall ad performance.

4.5 FC support of smart products

FC enabled smart products with, for example, the aid of smart packaging might leverage various sensors, indicators, or electronic components to gather data, monitor product conditions, and communicate with consumers or supply chain stakeholders. For example, sensors will track various parameters like temperature, humidity, location, and other environmental conditions. This enables real-time monitoring of the product's status throughout its journey in the supply chain, helping to ensure product quality and safety. Smart products can include tamper-evident seals or indicators that change color or show visible signs if the package has been opened or compromised. This will help to prevent tampering and ensures the integrity of the product. By monitoring environmental factors, smart will help extend the shelf life of perishable products by ensuring optimal storage conditions. The system might also include sustainable materials solutions or use of electronic components sparingly, contributing to reducing the overall environmental impact of products and packaging. Smart packaging can be integrated into the FC enabled MIoT ecosystem, allowing seamless communication between products, retailers, and manufacturers. This enables better inventory management, reducing waste and ensuring timely restocking. More on smart products see Web Appendix E.

4.6 FC-supported marketing research

Conventional market research is usually costly, time-consuming, and intrusive, and the generated data may have a short shelf life in fast-moving markets. FC can store both quantitative and qualitative data, leverage experimental studies and secondary and unstructured data, and above all, exploit field studies, combining data from emotion monitors with actual purchase data. As afore noted, empathic affective computing might be highly relevant to marketing researchers seeking to understand consumers’ emotions (Caruelle et al. 2022). Emotion detection is central to market research, as many consumer decisions are driven by emotions. As such, empathic affective computing will likely advance market research by offering a new range of methods for determining which emotions are triggered by a product, advertisement, or other element of the marketing mix. In an era of big data, consumers’ cognitions, emotions, and behaviors are collected and recorded by various devices. Particularly when consumers steadily conduct web searches and engage in online shopping, every operation and search they perform might reflect their consumption preferences and habits (Bleier et al. 2019). Information found online provides a digital footprint that implicitly segments consumers. Also, FC-enabled AI may automate some aspects of conventional quantitative and qualitative marketing research. Accordingly, FC is projected to improve marketing research through the following processes: 1. Data collection and processing: FC allows researchers to collect and process data at the edge of the network, closer to the data source. This is particularly beneficial for MIoT research projects where data is generated from sensors and devices (Aliyu 2023). Fog nodes will be able to perform real-time marketing data analysis, filter out irrelevant data, and aggregate meaningful insights before transmitting the processed data to a central server or cloud. 2. Distributed computing and collaboration: FC will allow researchers to distribute computing tasks among fog nodes, enabling collaborative research efforts across multiple locations. Fog nodes can share processing and storage capabilities, allowing researchers to work together on complex computations or data analysis tasks. This will promote collaboration, reduce the burden on centralized resources, and enhance marketing research productivity. 3. Real-time monitoring and control: FC can support real-time monitoring and control applications in research. For example, in environmental research, fog nodes equipped with sensors can continuously collect data on temperature, humidity, air quality, etc., and perform local analysis to detect anomalies or trigger immediate actions. Real-time monitoring and control will enable researchers to respond immediately to changing conditions and facilitate timely research interventions. When customers arrive, marketers often collect information to generate delay forecasts. Delay data-collection, customer loyalty profiles, and forecasting systems can be designed to improve customer satisfaction (Chang et al. 2023). 4. Edge analytics and machine learning: FC will enable researchers to deploy analytics and machine learning algorithms at the edge for real-time marketing decision-making and predictive analysis (Hussein et al. 2023). Fog nodes will be able to process and analyze consumer data locally, extracting valuable insights and patterns. This might be particularly useful in research domains where immediate data analysis is critical, such as time-limited promotion campaigns, or anomaly detection in product processes. 5. Mobile research applications: FC can support research applications on mobile devices, leveraging their computing power and connectivity. Mobile research applications are used to collect and process data locally, reducing reliance on continuous network connectivity and allowing research to be conducted in remote or resource-constrained environments. FC opens up opportunities for mobile-based data collection, surveys, and field studies frequently used in marketing research (Hazra et al. 2023). 6. Resource optimization and energy efficiency: FC can help to optimize resource usage and energy efficiency in research projects. By distributing computing tasks and data storage across fog nodes, marketing research will be able to reduce the load on centralized servers and cloud infrastructure, resulting in more efficient resource utilization. Moreover, fog nodes can implement energy-saving mechanisms, such as dynamically adjusting their power usage based on workload, contributing to energy-efficient research operations. 7. Privacy-preserving data analysis: FC will allow researchers to perform privacy-preserving data analysis at the edge. Instead of transmitting sensitive data to a central server, fog nodes can process data locally while preserving individual privacy. This is particularly relevant to marketing research, which frequently relies on personal or sensitive data, as well as voting and healthcare research, where privacy regulations need to be respected (Caruelle et al. 2022). 8. Hybrid capabilities: FC can deploy and combine data from different resources using an advanced monitoring infrastructure (AMI), in which smart grids deploy smart meters and sensors to collect various types of data. This will yield new measures of consumers’ mental state and behavior, which will facilitate screening, diagnosis, and ways of approaching consumers. 9. Concept testing: Research techniques routinely used to screen new marketing ideas do not have sufficient realism in terms of concept presentation, competitive context, and measurement of customer behavior to adequately assess the value of innovation. Lacking hard evidence of customer demand for significant product change, firms can use FC. For example, hybrid approaches employ a combination of data derived from consumers’ digital emotions, physical product interactions, monitored senses, such as touch, taste, and smell, as well as virtual reality simulations (Chang et al. 2023; Clegg et al. 2023; Dorneles, et al. 2023; Luangrath et al. 2022). All offer several advantages over conventional research techniques for testing new marketing ideas by allowing researchers to gauge shopper response to new product designs, packaging, promotions, and merchandising in realistic, competitive settings while maintaining a high degree of control and confidentiality. With the use of virtual reality (VR) simulations, researchers will be able to create photorealistic simulations of the retail environment, conduct marketing experiments, and collect detailed behavioral data without requiring retailer cooperation. Multiple concepts can be tested in a controlled and confidential setting, at a lower cost and faster than through an in-market test.

10. Adaptive Algorithms: Marketing research process can be highly dynamic and subject to environmental changes. FC can host adaptive algorithms that adjust process parameters in real-time 11. Decentralized Collaboration: Marketing research often involves collaboration among different research groups or institutions. FC can support decentralized collaboration by enabling efficient and secure sharing of data and processing resources among collaborators. 12. Questionnaire design: The emerging technologies of FC enabled language models and content analyses will provide researchers with questionnaires tailored to the investigated issue, context, and target respondents, in multiple languages (Hazra et al. 2023). 13. Compliance: FC can help researchers to comply with regulations in a number of ways, such as by ensuring data sovereignty, that is, that data are stored and processed in a way that respects the laws and regulations of specific country, while enhancing auditability by providing a clear trail of data-processing activities (Costa et al. 2022). 14. FC-supported nanomarketing: Nanotechnology and FC may play an important role in marketing, including the development of FC-enabled nanotechnologies for marketing research. This future evolution warrants some elaboration.

Nanotechnology (“technology of the small”) encompasses a number of emerging technologies that deal with structures with dimensions of less than 100 nm (10.9 m). Nanotechnology alludes to the advancement of structures that display novel and essentially improved physical, compound and natural properties, marvels, and nanoscale-sized procedures. Nanotechnology has led to the development of miniature complex tools, or nanodevices, and more efficient electronic components, which can obtain enhanced results at minimal dimensions and costs. As such, it has the potential to revolutionize many marketing technologies, such as smart cards and neuromarketing (Mileti et al. 2016). Indeed, nanoscience is establishing a new marketing frontier—nanomarketing. This is highlighted in the development of new nanodevices and materials with improved properties. Nanoscale devices will be engineered to have a wide range of properties, including strength, flexibility, conductivity, and biocompatibility with unprecedented sensitivity and selectivity, all with the potential for use in a wide range of FC marketing applications, for example, collecting data from the human body (e,g., consumers), physical entities (e.g., products, ads, shelves) and locations (e.g., stores), and transmit it to the FC for analysis. This unique data could be used to monitor emotions and behavior (Web Appendix F) and deliver personalized products and messages. In other words, FC-enabled nanomarketing will make it possible to do the following: (1) develop more advanced and sensitive market research tools, for example, nanoscale biosensors could be employed to measure emotional responses to advertisements, packaging, or products, providing marketers with deeper insights into consumer reactions, while nanoscale imaging techniques, such as scanning probe microscopy, could be used to study consumer behavior at a micro-level, both of which could help researchers to better understand how consumers interact with products and make purchasing decisions; (2) carry out nonintrusive and noninvasive studies in places visited by consumers; (3) monitor consumers’ mental processes in real time and over time; (4) combine various technologies to corroborate the results obtained by different scientific nanotools; (5) highlight ethical issues raised by the use of these novel, miniature nanodevices; and (6) improve packaging properties, such as barrier properties, antimicrobial coatings, or smart packaging that can detect spoilage or temperature changes as well as product fabrication. All of these can influence data on consumer perceptions, extend the shelf life of products, and improve marketing’s ability to understand, predict, and respond to market changes.

In sum, FC promises to transform marketing research by deploying new types of monitoring systems, generating new types of data analysis, and informing how and where value is delivered. The emergence of, for example, FC-enabled MIoT/AI/nanotechnologies alone and in combination will increase the amount of data gained through connected (nano)devices and smart monitors, will enhance researchers experience, and widen the scope of marketing analytics. The inception of FC may entail a unique set of metrics for measuring marketing effectiveness beyond what has been employed in the existing marketing research domain.

4.6.1 FC-supported marketing management

On November 28, 2022, Adidas announced that their nanotechnology enabled them to prove that the football star Cristiano Ronaldo “definitively made no contact” with the ball in Portugal’s opening goal against Uruguay. “The 500 Hz IMU sensor inside the ball allows us to be highly accurate in our analysis,” they stated. Indeed, the Al Rihla match ball contained a very small sensor suspended in its center to assist in detecting unclear touches and more data to improve the quality of game referees’ decision-making, making it the first ball to feature such an innovation. Similar innovations in marketing ecosystems will aid and improve marketing decisions. Indeed, marketing systems are quickly transitioning from human-controlled processes to closed-loop control services supporting their operations autonomously using extensive tiny sensor and computing infrastructures. This revolutionary change is critical for supporting growing data-intensive and time-sensitive smart marketing applications (Ma and Zhang 2022). As businesses seek to produce desired business outcomes more efficiently, it is hard to imagine an aspect of marketing practice that will go untouched by FC. In a sense, FC may not only personalize marketing but also democratize it. For instance, start-ups, small companies, individual entrepreneurs (e.g., artists, craftspeople, influencers, etc.), non-profit organizations and charities, as well as firms at the bottom of the pyramid might all benefit from FC. Hence, marketing is poised to gain even more importance due to the growth of digital technology and formalization of firms’ digital strategies. After all, marketing’s role is to help define, own, and manage the organization’s customer interface. Future marketing systems can be designed with the aid of MIoT and FC to minimize latency in marketing data processing. Data-sharing configurations in such systems will be calculated automatically based on node verification, while transmitted data will be protected, avoiding the risk of data tampering in the public cloud. Likewise, FC will be able to integrate the various techniques employed to detect spam or fake online reviews, including machine learning algorithms using, for example, logistic regression, random forest, naïve bays, etc. (Costa et al. 2022). In sum, FC can provide a range of exciting possibilities to facilitate and enhance smart marketing. It is paramount to reiterate that FC’s most significant contribution to marketing inheres in its ability to integrate and synthesize even convoluted, unstructured data (numerical, semantic, sensory, emotional, etc.) and immersive technologies into unified, standardized and workable interoperability (hybrid) algorithmic system, all with the goal of facilitating managerial decision-making and enhancing research capabilities in different marketing scenarios. However, to realize FC’s potentialities, managers and academicians are advised to explore future FC trends and challenges for smart marketing and analytics.

5 Future trends and challenges

“Technology and service providers that fail to adapt to the pace of fog shift face increasing risk of becoming obsolete or, at best, being relegated to low-growth markets" (Gartner Co 2023).The marketing discipline can employ FC in its various facets to generate better outputs in research and practice. It is acknowledged that this is a concept that has not been tested hitherto in marketing. Accordingly, we call here for future research and development to fully explore and realize FC’s marketing potential. Understanding new technological innovations and how they might add to the marketing mix is undoubtedly important to improving performance. FC’s full potential has yet to be realized because several challenges are still being addressed by the research community. According to SkyQuest Technology’s Global Fog Computing Market Report for 2023, the following are some of future trends for FC growth with marketing implications: Adoption of 5G technology and its synergy with FC; integration of AI and ML capabilities; edge analytics for real-time data processing and decision-making; emergence of fog-to-cloud orchestration and hybrid cloud architectures; convergence of FC with blockchain technology; growth of edge devices and sensors in IoT ecosystems; edge-native applications and services; language models; and more improvements in edge security and privacy. Intelligent Device-free Sensing IDFS will possibly represent one of the most significant advancements in the near future IoT. Current emerging techniques for IDFS include wireless device-free localization, gesture recognition, action recognition, computer vision-based sensing, etc. Acquiring accurate information about human activities and environments is particularly critical. But, handling such issues is challenging because of complex interference from dynamic environments, moving targets, ambient noise, etc. There are several approaches in the literature to address these challenges. FC has been proposed as a new model to overcome these challenges (Huakun et al. 2024).

As the market for smart devices develops, so does the number of internet users. Basically, FC was designed to connect billions of intelligent entities, which can help to establish a better future. Due to the widespread use of smart gadgets, colossal amounts of data are being generated and transmitted via the internet. Thus, the need for efficient algorithms to manage data, fog devices, and cloud servers is also expected to grow (Das and Inuwa 2023). In this regard, there are several fundamental challenges that marketing science and informatics will have to investigate. The first is how to handle and make sense of billions of observations that are being generated on a dynamic basis, while the second is how to translate the deep insights derived through analytics into new fundamental and applied knowledge (Hassan and Rashad 2023). The second is how to authenticate such a large number of connected devices via FC in view of its highly distributed nature. The third concerns, as aforementioned, how affective computing will evolve as research and development continue in the field. Numerous machine learning models have been proposed to predict emotional responses for subsequent managerial tasks. However, they are often also viewed as opaque black boxes, arousing concern as to their generalizability. Hence, the question remains as to how FC-enabled AI can help alleviate generalizability concerns (Hoffman et al. 2022) by making such models more explainable while maintaining good predictive accuracy. The fourth challenge relates to the question of how the rise of FC will improve cloud computing performance and how acceptance of smart devices will transform future marketing research and analytics and shape the computing environment around marketing. For example, smart marketing will require a multiplicity of sensors to adequately sense the competitive market and devise marketing-mix strategies according to the changing requirements of consumers. As physical sensor deployment would be too costly in this case, an important question to be addressed is whether virtual sensors (VS) might be a viable solution, which a nearby fog can readily provide (Koohang et al. 2023)? Thus, recent advances in VS networking, robotics, and communication technologies will likely pave the way for a more autonomous marketing environment (Hollensen et al. 2022). The fifth question is whether FC is capable of embracing the variety of datasets and measurement devices necessary for some of the applied marketing decision-making and choice research. The sixth concerns the concept of digital ethics (Naik et al. 2008; Sen 2023) as a potential framework for understanding the ethical implications of emerging technologies like FC and language models (ChatGPT). Theories of consumer psychology and behavior, such as the theory of planned behavior (TPB; Bosnjak et al. 2020) and theory of reasoned action (TRA; Hale et al. 2022), can contribute to the understanding not only of how these technologies may be used to persuade customers, but also of how to ensure that that they are used responsibly. The seventh challenge has to do with FC-enabled nanotechnology as a promising new field with the potential to create a new marketing paradigm: nanomarketing. As research continues, market stakeholders will surely grapple with many innovative applications and research questions, and as nanoscience develops, it is likely to be combined with advanced computing methods such as FC, which will significantly impact marketing research in the years to come. With respect to this challenge, as AI decision support systems and other applications are currently under much criticism, it is important to investigate which market stakeholders are more inclined to adopt FC-enabled AI and which feel more vulnerable in the face of this emerging technology (SimanTov-Nachlieli 2023). Finally, technological advances have resulted in a hyperconnected world, requiring a reassessment of FC-enabled marketing research from the perspectives of organizations, consumers, and society. In addition, among the many relevant questions that might be posed in future research are also the following:

RQ1: What types of marketing devices and services can be moved into this platform? RQ2: What are the fundamental future changes in marketing research stemming from FC integration in the research process? RQ3: How can marketing utilize the FC architecture to maximize customer satisfaction, market share, and profitability? RQ4: What are the trending topics and future research directions for the adoption of FC in marketing? RQ5: Given that FC incorporates elements of the cognitive process, can it integrate emotional sources, such as emotionally intelligent machines, which are important to marketing research? RQ6: How can FC contribute to improving supply chain strategic performance and mitigate latency issues? RQ7: Can all security and privacy issues important to marketing be resolved by FC? RQ8: Given that detecting and selecting the relevant fog node in a high-mobility marketing environment is challenging (Costa et al. 2022), can FC fully coordinate the many mobile fog nodes?

RQ9: What are FC’s offloading capabilities when the aim of offloading is to outsource the load exchanges between fog nodes or between fog nodes and the cloud? RQ10: How can FC augment (rather than replace) human judgment and integrate with and substantially improve standard marketing processes and resulting outcomes? RQ11: Which analytical capabilities are required to determine and leverage the effectiveness of FC? RQ12: Which prevalent algorithms are investigated in fog-based smart marketing?

5.1 Limitations

The study has some noteworthy limitations. It has been our objective to present future FC applications in marketing in a concise way and to derive major research questions. Consequently, we did not elaborate on the many subtleties and intricacies of FC technologies, which are undoubtedly important subjects for future research. We also recommend that future research explore the effects of the value of FC information, information overload, trust in smart technology, technology-mediated life, and enjoyment of technology-enriched experiences, among many other questions that ought to be investigated in the general context of FC. Although the present article has focused on consumers, many of these issues are also relevant in business-to-business contexts, and we expect further research to apply the proposed framework in B2B e-commerce marketplaces.

6 Conclusions

Fog computing promises to become a robust and powerful means for obtaining solutions to previously unsolvable computational problems (Hazra et al. 2023, p. 89).

The Fourth Industrial Revolution has characterized marketing as integrating big data about customers and products across all channels and products, and using that data to better understand its end customer experience and visibility across all functional marketing domains. It is clear that “The era of the cloud’s total dominance is drawing to a close” (The Economist 2018). Emerging technologies such as neural networks, blockchains, the metaverse (e.g., virtual reality, augmented reality), cyberphysical systems, artificial intelligence, and nanotechnologies, some of which are used independently in marketing ecosystems, are revolutionizing the marketing landscape as we know it. The explosion of sensor and IoT data at the edge will force marketing systems to rethink the location of their data processing infrastructure, “bringing down” the cloud to the ground in the form of FC. One of the most remarkable change brought by FC is the unprecedented interactive system that cuts across physical, sensory, virtual, digital data, not only for consumers, but also products, stores, locations, shelves. In other words, FC’s major contribution to marketing lies in its ability to collect and compute big-data sets in milliseconds, from an array of different sources, including location-based applications (e.g., mobile devices), user-centric services, and behavioral and emotional modalities. The framework presented here suggests that FC is developing into a fully functional and potentially disruptive opportunity for marketers, one unachievable—or even unthinkable—within earlier marketing platforms. FC presents a platform for combining their usage in a scalable manner to enrich marketers’ knowledge. In this context, FC can play a crucial role in big data analytics in terms of anticipating and providing marketers with guided valuable information to meet customer expectations. The combination of MIoT-SDN devices and decision-support software provides managers and researchers with improved systems to monitor not only consumers’ attitudes and preferences but also various marketing strategies and tactics. This hybrid model promises to enhance marketing efficiency, consumer engagement and satisfaction, and competitive positioning. In short, FC is expected to become a unifying platform, rich enough to deliver a new breed of emerging services and enable the development of new marketing applications. Research on FC remains in its initial phases (Ding et al. 2024). Throughout the article, we have provided a holistic view of FC as a smart world facilitator. FC is a relatively new paradigm that has swiftly garnered wide recognition, and found broad applications, due to its significant contribution to advanced computing technology (for an updated list of references, see Web Appendix G). However, its use in marketing hitherto, perhaps surprisingly, has been very limited. Here, we have argued that FC is well suited for marketing, where there is a constant need to quickly analyze and react to real-time data. Indeed, FC’s ability to accelerate awareness of and response to events with minimal latency makes it perfect for marketing. It seems clear, as well, that MIoT-enabled FC may have the potential to revolutionize the future of emotion and behavior measurement in marketing, using hybridizing approaches. With the evolution of the IoT and AI services and technologies, which establish a new ‘smart world’ where everything is monitored automatically, merely providing quality of service is no longer acceptable as it fails to offer a satisfactory user experience. Indeed, quality of experience (QoE), which gratifies user experience and enhances user performance, has become far more vital, and FC is considered a major technological breakthrough in terms of the advancement of QoE (Vambe 2023). Marketing management and research can survive and grow only when it continuously and rapidly adapts to changing technologies. According to Martin Key et al. (2020, p. 162), “Marketing scholars tend to ignore new technologies, and this is particularly true of new information technologies….” We hope that the present article will encourage marketing managers and scholars to adopt this nascent platform to advance marketing practices, research, and theories. In much the same way as marketing have moved into cloud computing, FC is the natural next step. Just as the fog can make an instantaneous decision to turn traffic lights green when it detects emergency lights flashing on an ambulance, or as smart vehicles use a large amount of data to enable customer's safe driving, FC can assist smart marketing by improving decisions and customer satisfaction.