August 20, 2023


One of the most disruptive developments in the retail industry in recent times has been the increasing dominance of ecommerce. Although ecommerce itself is not a new concept, advancements in ecommerce and its related fields such as deliveries and logistics

List of contents





One of the most disruptive developments in the retail industry in recent times has been the increasing dominance of ecommerce. Although ecommerce itself is not a new concept, advancements in ecommerce and its related fields such as deliveries and logistics, facilitated by AI and automation, are resulting in new practices and pressures that have significant implications for the wider retail sector and its workforce. The following sections outline specific innovations in AI and automation that are being used in ecommerce. However, beyond these technologies and practices, the impact of ecommerce on the retail industry, as discussed by focus group participants, is far-reaching.

The trend towards ecommerce was already in full swing before the pandemic hit, but Covid19 has further accelerated online sales and demand. In the UK, for example, internet sales as a percentage of all retail sales increased significantly from around 20% at the beginning of 2020 to between 30% and 35% for the rest of the year. However, for some retailers, particularly smaller ones, the cost of setting up an online platform with all the necessary technological and logistical infrastructure is too high, making it difficult for them to compete. This leaves them with no option but to outsource parts of the process to established retailing platforms such as Amazon or other large players. This presents a number of challenges for retailers, workers, and unions.

On the other hand, ecommerce and omnichannel retailing puts immense pressure on retailers to keep prices low in order to compete with other online stores. Consumers can use their smartphones to compare prices and examine goods in-store before making online purchases, which can squeeze profit margins for brick-and-mortar stores that already have higher rents and staffing costs compared to online only retailers. Real-time price comparison also has implications for in-store staff as they need to be aware of competitors’ offers and be prepared to deal with customers requesting price matches.

Nevertheless, ecommerce and the shift towards omnichannel retailing provide some smaller retailers with access to a wider market, compensating for the decline in foot traffic seen in many town centers. A participant in the focus group commented on this positive aspect of ecommerce.

Figure 1: Technologies processes and applications
Sales and customer data • E-commerce and outsourcing of services

• Cashless/contactless payment systems

• Ordering, inventory and stock replenishment

• Work/staff planning and scheduling

• Task allocation, targets and reward

• Surveillance and monitoring systems

• Predictive marketing and personalization

• Autonomous warehouses

• Automated/self-service HR

• Augmented/virtual reality and product visualization

• Counterfeit detection

 Networking, the web and IoT
Loyalty cards and memberships 
 Advanced image processing
Store sensors and remote sensing 
 Natural language processing
Cameras and CCTV 
 Machine Learning and AI
Robots and autonomous vehicles 
Mobile devices and wearable tech 


Smaller retail organizations are currently facing a disadvantage compared to established e-commerce platforms and larger retail chains with resources and brand recognition. The upfront investment in potentially risky technologies requires substantial amounts of capital. Moreover, the barriers to entry into the industry are high due to low margins and wages in smaller stores. Therefore, many retailers choose to outsource parts of the e-commerce process, such as marketing, logistics, and front-end user interface. However, this often means handing over sales and product data, as well as access to customer data and profiles. This can lead to the host platform launching competing product lines. The dominance of platforms in the retail industry can generate data inequities, especially if left without anticompetitive regulation. Established e-commerce platforms and larger retailers have advantages over smaller organizations, such as brand reputation, online visibility, existing sales and customer data, and in-house technical expertise. The rise of dark stores and Qcommerce poses a major threat to smaller bricks-and-mortar stores and larger established chains. Dark stores offer extremely rapid delivery through click-and-collect or home delivery. Qcommerce firms are disrupting the industry with on-demand delivery and the use of automation, order processing, and labor management technology for delivery drivers/riders. The sector’s economic muscle increases as e-commerce sales volume increases, opening up more funds to develop advanced systems. Regulation of dominant market positions may be possible under the Digital Markets Act and Digital Services Act forthcoming under European law. However, Amazon seems unlikely to face serious regulatory threats to its market position in the near future.

The participants in the focus group identified a significant challenge in the form of distinct regulatory structures for online and in-person retail. These regulations encompass various aspects such as payment of local taxes and business rates, as well as rent for prime real estate. These factors serve to elevate the cost of in-person purchases compared to online ones, which, when combined with increased overheads, creates a difficult situation for in-person retailers. The participants of the focus group suggested that policy interventions were necessary to even out the regulatory environment, especially with the significant economic power imbalance between ecommerce platforms and larger retailers versus smaller retailers, which account for the majority of employment in the retail sector. The pandemic has further accelerated the growth of online sales, leading to concerns about employment security and working conditions in the retail sector. The shift from in-person retail to logistics work has also raised concerns about job quality and gendered employment, as warehouse and delivery work are physically demanding and dominated by male employment. The increasing prevalence of flexible contracts among ecommerce players is leading to a fragmentation of the workforce, making union organizing and recruitment challenging.

On the whole, although ecommerce offers various advantages to customers, such as greater flexibility and convenience as well as price competition, it has resulted in significant costs for numerous retailers and retail staff. Many are worried that the amplified competitive pressures attributable to the escalating significance of ecommerce and the related imbalances have led to the closure of several retail organizations or necessitated a review of their offerings. In order to choose between a ‘low road’ strategy of competing on price and reducing work conditions or a ‘high road’ approach of competing on customer service through skilled and well-compensated workers and providing a unique offering, retailers have to make a decision.



Digital payment systems comprise a suite of technologies, including mobile phone payments, contactless card transactions, and scan-and-go checkout systems that use smartphone apps. These systems offer convenience and flexibility to consumers but also raise concerns about privacy and security due to surreptitious data gathering. Initially, these systems were the preserve of large retail chains, but they are now prevalent in small/local retailers and farmers’ market stalls due to their decreasing cost and increasing applications linked to mobile phones. Despite this accessibility, smaller retailers remain at a disadvantage due to the possibilities for data capture, analysis, and processing enabled by heavily promoted scan-and-go and online account-based payments used by large retailers. Although efforts to address this disparity are small, they are evident. Self-checkout systems and related technologies are also expanding, enabling the development of staff-less stores that can be operated with very few employees, which raises concerns about potential job loss through automation. However, it also potentially enables the opening of stores in remote and sparsely populated areas. New digital credit players are significantly facilitating the ongoing sales growth of the sector, increasing convenience for consumers but can encourage overconsumption and indebtedness, particularly among economically disadvantaged groups, and especially when combined with the kinds of sophisticated marketing systems. Cashless payment systems offer security for firms against theft and error, but these advantages also hold for workers in stores who find handling cash and calculating change stressful.



Automated inventory management systems consist of various technologies that enable digital monitoring and examination of stock and inventory, as well as analysis of sales data and automatic replenishment order placement. These systems collect inventory data from multiple sources, including RFID tags, weighing scales, cameras, and scanners, and process and analyze it algorithmically. The benefits of these systems for firms are numerous, including faster unloading and sorting in warehouses, reduced errors and labor costs associated with stock checking, more efficient shelf stacking and picking for home delivery, demand forecasting, optimized labor scheduling, and enhanced logistical efficiencies and automated reordering.

For consumers, the implementation of digitalized inventory systems can lead to increased product range, availability, and information on stock, including lead times and expected delivery times. Medium to large-scale retailers are increasingly adopting these systems, as they offer reduced labor costs and enhance just-in-time throughput of products. Moreover, digital logistics platforms that offer inventory and supply chain management services for incoming inventory and e-commerce sales are increasingly opening up the field to smaller firms.

However, the effects of these systems on workers remain ambiguous. While shelf stacking robots are still far from being mass deployed due to cost and competence issues, warehouses and logistics centers serving both bricks-and-mortar stores and e-commerce fulfillment are becoming increasingly subject to robotic automation. Nonetheless, the number of warehouse and logistics staff in its robotized fulfillment centers remains substantial, not only in technical roles but also in low-wage picking jobs.

Moreover, while grocery warehouses lend themselves to the deployment of large-scale automated systems, in high-value and lower-throughput sectors like furniture and consumer electronics retail, demand for such technologies is likely to remain lower, even as cost reductions facilitate more widespread access to automation. Although automated consumer-facing reordering systems may pose some threat of technological unemployment for sales and checkout workers, such applications currently account for only a tiny minority of overall purchases and seem unlikely to become widespread in the near future.

Automated inventory and ordering systems can effectively eliminate the need for workers to perform time-consuming inventories, allowing them to focus on more skilled sales tasks. However, this shift in responsibilities requires employers to invest in developing the necessary skills for their staff to navigate digital logistics systems. Workers must adapt to using handheld terminals or smartphones to access logistics data and product details in order to respond to customer inquiries. Unfortunately, the increased availability of information through web and smartphone apps can potentially undermine the role of retail workers as intermediaries between consumers and producers.

While automated inventory management has the potential to improve supply chain efficiency and reduce carbon footprints by eliminating wasted space during transport and reducing distances traveled by deliveries and returns, attendant practices currently offset potential benefits. For instance, retailers increasingly place small batch orders with suppliers, and customer-facing one day, two-hour, or even 15 minute delivery services (as in the case of q-commerce) facilitate increased emissions. Moreover, inefficiencies built into on-demand delivery systems further exacerbate the problem.



Algorithmic work planning and staff scheduling is becoming increasingly prevalent in larger retail establishments. Employers can benefit from such systems as they provide a more precise forecast of staffing rosters during peak demand periods, both in-store and for home-based ecommerce deliveries. The use of such responsive digital scheduling systems can ensure that customer demand is met more efficiently, resulting in a smoother shopping experience during peak times.

However, the use of automated shift planning software can have a powerful negative impact on workers’ conditions and work-life balance, especially when combined with flexible contracts such as “zero hours.” This combination allows employers to deploy labor during peak demand periods while reducing labor costs during troughs, effectively shifting the burden of risk for reductions in consumer demand from the firm to the worker. Moreover, scheduling systems can lead to work intensification and a sense of being “always on” when combined with little advanced scheduling.

The kind of work intensification experienced in ecommerce is now permeating brick-and-mortar stores. The automation of work planning and scheduling can dehumanize the employment relationship by limiting workers’ abilities to change or contest schedules. The system appears “objective,” making it difficult for even managers to question it. While algorithmic scheduling could theoretically act to limit managerial use of scheduling as a system of discipline, reward, and favoritism, evidence suggests that such “automatic” scheduling systems can be used to favor and discipline workers.



Retailers are increasingly utilizing digital technologies for “algorithmic management” alongside their planning and scheduling applications. This involves the algorithmic allocation of tasks, targets, rewards, and bonuses to workers through sophisticated software analytics packages. Machine learning and AI enable the gathering, cleaning, and analysis of “big data” from a wide range of sales functions, including sales calls, webinars, customer interactions, and preparation. Data for such systems is gathered from various sources, such as in-store cameras, sensors, wearable devices, and browser tracking cookies. Such systems offer substantial benefits, including highly refined customer data and the ability to closely monitor and improve staff performance beyond human management. For instance, a multinational jewelry retailer reportedly uses detectors to monitor footfall and sales data, which is used to adjust staff sales targets in real-time. However, there are concerns that such systems could lead to work intensification and create intense competition between workers, undermining cooperation and trust between staff and leading to unintended consequences. Additionally, these systems involve increased data gathering on staff and customers and raise ethical and legal questions about privacy and control, as well as whether they comply with existing collective agreements.



The allocation of tasks and planning of workforce systems entail the collection and utilization of sales, staff, and customer data on an unprecedented scale, often in real-time. This can comprise data from wearable technology, in-store (or warehouse) cameras and sensors, as well as data from sales and stock information and RFID tags. Although workplace surveillance and monitoring are not uncommon, there were concerns among focus group participants that AI and automation were facilitating staff surveillance and monitoring on an unprecedented scale that could lead to decisions affecting staff without human managerial input. For employers, work planning and the assurance that staff are working as intended bring obvious benefits. However, surveillance and monitoring systems have the potential to undermine employee trust and engagement. For workers, the increasing deployment of intensive monitoring and surveillance systems raises serious concerns over work intensity and job quality. Dynamically modified targets, such as those discussed above, can exert intense pressure and act as a source of stress upon workers, particularly when imposed without managerial oversight or recourse. Moreover, there are concerns regarding both robustness/quality and transparency. In many cases, algorithmic management systems are like ‘black boxes,’ with little information available either to those being managed or even to line managers over how decisions are arrived at, making it more challenging to challenge decisions as they appear to be objective, based on science and facts, rather than human biases. This can result in an element of distrust and increase arbitrary managerial overruling, making it more difficult to fix issues. The rollout of such systems raises the possibility of attempting to ‘game’ algorithms in ways that do not benefit firms or workers. Monitoring specific metrics, such as sales targets, time spent performing specific tasks, etc., can lead employees to concentrate on achieving those targets at the expense of other tasks, resulting in unexpected patterns of behavior that may not always yield the intended results for employers. This may be a particular issue in ‘black box’ automated surveillance systems that use machine learning, as it is not always clear which variables carry the most weight. Finally, such systems can further intensify work and raise questions about pernicious or illegitimate forms of worker surveillance, eroding worker-employer trust, particularly if monitoring results in sanctions for workers, with negative consequences for employee engagement, motivation, and potentially staff absence and turnover. Concerns about homeworking have been magnified by the Covid19 pandemic. Some people are apprehensive that AI is (or could be) used to monitor incoming and outgoing communications in the context of remote working. Systems are already widely used to monitor whether remote workers are working by examining login/logout times, keystrokes, and screen captures. However, AI makes monitoring a far greater number of inputs, raising concerns about surveillance overreach privacy and fostering distrust between managers and workers.



The combination of advances in machine learning and the abundance of consumer data has led to the development of increasingly sophisticated automated marketing systems. In the realm of ecommerce, the use of tracking cookies and site-specific accounts allows for easy customer monitoring and profile building. This data is collected and analyzed for various purposes such as measuring time spent on a particular item, click-through rates, and repeat purchases. The resulting personal profiles of customers are used to target them with special offers and to monitor customer demographics and popular items. In-store technologies, such as loyalty cards and smartphone scan and go payment systems, are also being used to gather data and create customer profiles. These systems can inform dynamic pricing that entices customers to buy, and can also benefit workers through increased sales and job security. However, there are risks associated with these systems, such as overspending and overconsumption, as well as concerns over transparency and privacy in data collection and sharing. Many consumers are unaware of the extent of data gathering by firms in both ecommerce and in-store settings.

The ramifications of these technologies on competencies and occupations are not easily foreseeable. On one hand, these systems imitate customary methods of upselling and cross-selling, a proficient task carried out by a salesperson selling high-value items. However, due to the reduced cost of suggestion algorithms, predictive marketing systems can be employed on a wide range of products and can be performed remotely from the physical store. Consequently, these systems, in some respects, replace the conventional role of retail workers in providing personalized recommendations to clients, which can potentially erode the professional identity of sales staff. Conversely, algorithmic customer suggestions can serve to enhance the abilities of workers. If algorithms generate a range of possibilities, the sales personnel can guide the client through the various options available to make an informed decision, necessitating investment in the sales workers’ product knowledge and digital skills.

Given the available opportunities for data gathering online, along with current technological capabilities, participants of the focus group expressed concern that predictive marketing systems provide an advantage to larger retailers with an existing online presence over smaller brick-and-mortar retailers. While it was recognized that some smaller retailers are venturing into this area and that some are leveraging innovative individual and collective store loyalty programs to do so, there were apprehensions about data ownership and sharing. There were queries about who possessed and had access to data, particularly when smaller retailers utilized large online platforms to market their products. Additionally, there were worries that online platforms could benefit directly from the capability to sell anonymized data to third parties.



Robots and automated computer systems have been utilized for some time, particularly in the manufacturing sector. However, recent advances in robotics and machine learning have expanded the potential uses of these technologies. In the retail sector, robots and intelligent machines are used for automated sorting, fault detection and quality checking systems, autonomous vehicles, and warehouse robots, as well as robotic process automation (RPA) systems that can perform various customer service, marketing, and HR functions. Although these technologies automate many physically demanding or repetitive tasks, which can increase productivity and potentially make jobs less monotonous, their introduction is not without issues.

One characteristic of robots and automated computer systems is that they do not get tired, and in some cases, they can perform tasks faster than human workers. This can lead to productivity gains for employers and cheaper prices for consumers. However, robots and sophisticated RPA systems are expensive and require large amounts of data to develop. Therefore, for some uses, their introduction only becomes cost-effective at scale, and many smaller retailers may be priced out of making the most of these technologies or may have to pay for third-party providers for some services, both of which have implications for their ability to compete on price. Furthermore, the benefits of deploying such technologies at scale will likely favor larger players, prompting further consolidation of the industry.

For workers, the impact of these technologies is unclear. On the one hand, where robots are used to automate physically demanding or repetitive tasks, this can improve health and safety and free up workers’ time to do more specialized tasks, leading to job upskilling. On the other hand, there are concerns related to jobs and health and safety. In relation to jobs, there are three main concerns. First, automation may lead to less demand for human workers and fewer jobs. Second, the removal of repetitive and less complex tasks may free up time for workers to do more stimulating tasks but could lead to less diversification of tasks, leaving only complex and demanding tasks for workers. This could lead to an intensification of work. A related third concern is that any job upskilling needs to be accompanied by appropriate training and increased pay commensurate with the work. There were concerns among focus group participants that these conditions were not always, perhaps even rarely, met.

Advancements in robotics have led to an increase in their ability to perform a wider range of tasks with greater accuracy. However, the feasibility of investing in robots is still limited as there are certain tasks for which human labor is more affordable, flexible, and cost-effective. Furthermore, the unique qualities of human workers are difficult to replicate with robots. Although robots can perform challenging tasks, human workers are still necessary to work closely with them. In highly roboticized environments, such as the Amazon warehouse in the United States, the injury rate is reported to be three to four times higher than the industry standard.

Another challenge that automation presents is the inadequacy of most countries’ tax systems in accounting for a future in which technology and automation contribute significantly to productivity. Robots are not subject to the same taxes as human workers, and corporation tax is often lower than income tax, which encourages the use of labor-saving technologies. In the absence of an increase in the tax paid on profits, a reduction in income tax paid by workers could lead to a shortfall in tax revenue, particularly if working hours decrease. This could have consequences for welfare states, particularly if there is significant technological unemployment. However, the focus group participants were uncertain about whether automation would lead to job creation or destruction. Some participants noted that Amazon, a retailer at the forefront of automation, continues to expand its workforce, suggesting that productivity benefits may lead to more job opportunities rather than fewer.



The progress made in AI and ML has given rise to a number of HR processes and functions being automated by retail companies. An instance of this is Walmart in the USA, who developed a machine learning algorithm to rank applicants for store level vacancies, and Carrefour in Europe, who have implemented a self-service system for workers to access various HR services such as pay slips, schedules, and training resources. Carrefour provides smartphones for all store workers to use exclusively for contacting human resources and obtaining information on products, their job, and training. This has improved the speed and efficiency of communication between staff and management, as well as enhancing staff access to HR services. However, there are concerns regarding worker surveillance, monitoring, and control, particularly when AI and ML are utilized for HR decision-making, which raises issues of transparency and privacy in regard to employee data. The opaque nature of algorithmic decision-making poses a challenge, as workers and even managers have no knowledge or control over the information used by the system to make decisions or recommendations. In addition, work smart devices could be utilized to monitor staff whereabouts and communications in new ways during the pandemic, leading to concerns about privacy and trust. These issues should be addressed through greater transparency and engagement with worker representatives.



Numerous retailers have recently implemented various applications that utilize virtual or augmented reality (AR) and machine vision. These applications include phone or tablet-based product visualization systems, both in-store and online, that permit customers to view outfits in different contexts, as well as apps that can identify body shape and suggest clothing items. Machine vision applications are also used to detect counterfeit products or fraudulent customer behavior such as missed scan detection at checkout. AR-based systems, such as those used by Walmart and Tilly’s, can also be employed as marketing tools, creating treasure hunts and mini-games to attract consumers to stores.

These systems offer numerous benefits for both retailers and consumers. The latest “digital catwalk” technology creates web-based events in which consumers may participate as spectators and shop simultaneously. The visualization of products can potentially assist customers in making more confident purchasing decisions and reducing the number of returns, both for online and in-store purchases. Consumers are provided with significant convenience by such augmented reality technologies, avoiding trips to stores and the inconvenience of returning items that do not meet their expectations. In-store AR-based content can also be used as a marketing tool, enhancing the customer experience, and potentially leading to increased sales, reduced costs, and increased profitability.

However, some concerns were raised by focus group participants. Firstly, while product visualization can be developed for in-store or online use, such systems arguably have greater potential for online retail, as imaging software allows customers to try on clothes digitally online, increasing the reach of ecommerce and enabling it to replicate some of the features of in-store retail. Secondly, the cost of such systems and technical expertise required to develop them could potentially exclude smaller retailers from these developments in the short term. Additionally, product visualization systems could potentially reduce employee customer interactions and the opportunities for staff to deploy customer service and sales skills, tasks that have traditionally been the domain of human workers. Finally, all of the applications in this section use cameras and generate video data, raising ethical concerns about privacy and data usage.



  1. Brynjolfsson, E., Rock, D. & Syverson, C. (2017). Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics. NBER Working Paper 24001.
  2. CIPD (2019). People and Machines: From hype to reality. London: Chartered Institute for Personnel and Development.; Susskind, R. & Susskind, D. (2015). The Future of the Professions: How Technology Will Transform the Work of Human Experts. Oxford: Oxford University Press.; Savage, N. (2020). Another set of eyes for cancer diagnostics, 579,
  3. Hutton, G. (2021). Retail sector in the UK.  HoC  Briefing  paper SN06186. London: House of Commons Library.  https://researchbrief
  4. UNI Global Union & UNI Europa (2020). Accounting  for  Workers’ Rights When Regulating Amazon & Other Giants. Submission to EU Commission’ Consultations for Regulation of Digital Platforms (DSA/ NCT). Available at: 2020/09/Final_UniGlobalUNIEuropaDSASubmission.pdf  [accessed 6/1/2022]
  5. (UNI Global Union & UNI Europa, 2020) (n16) BBC (2021, June 2). Amazon warehouse injuries ›80% higher‹ than competitors, report claims. Available at: https://  [accessed  7/1/2022]
  6. Ocado Group. (n.d.). Our partnership with Groupe Casino. Available at: groupecasino  [accessed  6/1/2022]
  7. Wallop, H. (2020, November 28). ›Christmas slots went in five hours‹: How online supermarket Ocado became a lockdown winner. The Guardian,., Available  at: 2020/nov/28/christmasslotswentinfivehourshowonlinesuper marketocadobecamealockdownwinner  [accessed  20/12/2021]
  8. Deloison, T., Hannon, E., Huber, A., Heid, B., Klink, C., Sahay, R. & Wolff, C. (2020). The Future of the Last-mile Ecosystem: Transition Roadmaps  for  Public-and  Private-sector    World  Economic Forum. Available at: of_the_last_mile_ecosystem.pdf  [accessed  19/12/2021]
  9. Schildt, H. (2020). The  data  imperative:  How  digitalization  is  reshaping management,  organizing,  and  Oxford: Oxford University Press.
  10. (Wood 2016) (n4)
  11. Wood, A. J. (2020). Despotism on Demand. New York: Cornell Uni versity Press.
  12. Chung, D.J., Huber, I., Murthy, V., Sunku, V. & Weber, M. (2019, July 19). Setting better sales goals with analytics. Harvard Business Review. Available at: goalswithanalytics  [accessed  3/1/2022]
  13. Darr, A. (2018). Automatons, salesfloor control and the constitution of authority. Human relations, 72(5), 889–909.
  14. Levy, K. & Barocas, S. (2018). Refractive surveillance: Monitoring cus tomers to manage workers. International Journal  of  Communication, 12,  1166–1188.
  15. El Pais (2021, July 22). Mercadona paga una sanción de 2,5 mil lones de euros a Protección de Datos. Available at: sancionde25millonesdeeurosaprotecciondedatos.html [accessed 23/12/2021]
  16. (BBC, 2021) (n17)
  17. Hunt, W. & O’Reilly, J. (forthcoming). Rapid Recruitment in Retail: Lev eraging AI in the hiring of hourly paid frontline associates during the Covid19 Pandemic. Digital Futures at Work Research Centre working paper. Brighton: Digit.
  18. Hove, B. (n.d.). Carrefour: digital empowerment and innovation in HR. Available at: carrefourdigitalempowermentinnovationhr [accessed  7/1/2022]
  19. fashion/; returnsto/
  21. theft20196?r=US&IR=T; helpsretailersmanageselfcheckoutaccuracyandloss/
  22. fashion/