Research | Creating human-technology synergies at the organizational frontlines - ScienceDirect

Highlights

Abstract

Advanced technologies, including robotics and artificial intelligence, are radically transforming organizational frontlines. With rapidly expanding agentic abilities, these technologies increasingly collaborate with employees to co-produce the service and achieve superior outcomes, highlighting the importance of creating human-technology (HUMTECH) synergies at the organizational frontlines. However, the characteristics of these synergies and the process by which they emerge remain unclear. Drawing on cognitive appraisal theory and literature on complementarities, this paper offers a comprehensive understanding of HUMTECH synergies. We delineate three distinct modes of human-technology collaboration—employee-led, technology-led, and co-led—that give rise to HUMTECH synergies. Further, we provide a model that uncovers the three-stage process underlying the creation of HUMTECH synergies. Our model outlines how the emergence of HUMTECH synergies requires the a) existence, b) recognition, and c) enactment of complementarities between employees and technology, explicating the role of human and technological readiness, metaknowledge, and various fit appraisals for superior outcomes.

1. Introduction

Technological advancements such as cloud computing, big data analytics, robotics, and artificial intelligence (AI) are rapidly transforming organizational frontlines, or “the point of contact between an organization and its customers that promote, facilitate, or enable value creation and exchange” (Singh et al., 2017, p. 4). Consider the following examples: Medical AI now collaborates with physicians in diagnosis and treatment (Longoni et al., 2019); large-language models like ChatGPT, DALL-E, and Microsoft’s Copilot work with frontline employees in real-time to solve customer problems (Frey & Osborne, 2023); advanced business-to-business (B2B) platforms like Goldman Sachs’ Marcus Invest engage with financial advisors in portfolio management (Marcus, 2024); and intelligent service robots serve customers alongside employees in hotels, coffee shops, or hospitals (Evans, 2024, Shanks et al., 2024). This new generation of technology is increasingly agentic in nature (Baird & Maruping, 2021). Unlike frontline technology that relies on predefined rules, such as early-generation ATMs, check-in kiosks, or industrial robots for package and storage (De Keyser et al., 2019, Seeber et al., 2020), agentic technology (AT) is equipped with the ability to learn, adapt, decide, and act autonomously under uncertainty (Baird & Maruping, 2021).

This increased agency allows AT to assume responsibility for a wide range of tasks that were previously exclusively handled by employees, radically transforming the way in which employees work with technology at the organizational frontlines. For example, B2B solutions such as NICE CXone and Verint Systems collaborate with employees by providing real-time insights, AI-driven coaching, and autonomous workflow optimization. These agentic abilities allow those systems to proactively assist agents, streamline customer interactions, personalize service at scale, and enhance response times, whilst retaining a human touch (Nice, 2025, Verint, 2025). Rather than replacing or augmenting employees’ input (e.g., Huang and Rust, 2018, Larivière et al., 2017, Marinova et al., 2017), AT “partners” with employees to collaborate and co-produce the service and achieve superior outcomes (Blaurock et al., 2024, Le et al., 2024, van Doorn et al., 2023).

Previous research highlights the general importance of complementarities for successful human-technology collaboration (e.g., Huang and Rust, 2022, Pinski et al., 2023, Tsai et al., 2022, van Doorn et al., 2023) with human-technology synergies—or simply HUMTECH synergies—considered the “holy grail” for service frontline scholars owing to their ability to lead to superior service outcomes (Singh et al., 2019). HUMTECH synergies broadly refer to the beneficial interplay between employees and technology, where each leverages the other’s complementary potential to achieve superior outcomes that neither could have accomplished alone. Indeed, empirical research shows that when working together, humans and technology can outperform technology or humans working on their own, but only when they effectively leverage each other’s respective strengths or compensate for each other’s weaknesses (Fügener et al., 2022). Similarly, Blaurock et al. (2024) highlight how “reciprocal strength enhancement”—where the complementary strengths of employees and technology are leveraged—is crucial to achieve service improvements and enhance employee outcomes during human-technology collaboration at the organizational frontlines.

However, despite these valuable insights, there is little understanding of what constitutes HUMTECH synergies across different modes of human-technology collaboration. Given the increased agency of AT (Baird & Maruping, 2021), the nature of HUMTECH synergies is likely to differ considerably depending on whether technology or employees take the lead (Tsai et al., 2022). Moreover, prior research implicitly assumes that complementarities between employees and technology will automatically lead to HUMTECH synergies. However, extant literature is unclear regarding how such complementarities can be established, as well as how employees and technology can leverage these complementarities to achieve HUMTECH synergies at the organizational frontlines. In short, current studies on human-technology collaboration fall short of explaining the fundamental process of how HUMTECH synergies emerge. Thus, this paper seeks to answer the following two questions:

To do so, we take a complementarity perspective (Ennen & Richter, 2010) and draw on cognitive appraisal theory (Lazarus, 1991) to develop a process model of HUMTECH synergies that addresses these theoretical voids. Our model makes two key contributions:

First, we view HUMTECH synergies as having three distinct modes: a) employee-led, b) technology-led, and c) co-led. By clearly defining HUMTECH synergy and delineating its idiosyncratic features depending on the collaborative roles of employees and technology as leaders, followers, or partners, we critically advance literature on human-technology collaboration in service settings (e.g., Blaurock et al., 2024, Le et al., 2024), and human–machine collaboration research more broadly (e.g., Fügener et al., 2022, Tsai et al., 2022).

Second, by proposing that the emergence of HUMTECH synergies not only requires a) the existence of complementarities between employees and technology, but also their b) recognition, and c) enactment, this paper is the first to provide a holistic, multi-stage theoretical framework on how HUMTECH synergies emerge at the organizational frontlines. Specifically, our model explicates the role of human and technological readiness (Danatzis et al., 2022), metaknowledge (Fügener et al., 2022), and various fit appraisals (Pinski et al., 2023) along each stage of the HUMTECH emergence process, thereby synthesizing interdisciplinary literature streams from marketing, management, and information systems to provide a more fine-grained understanding of how HUMTECH synergies emerge.

This paper proceeds as follows. First, we review the literature on human-technology collaboration and conceptualize HUMTECH synergies depending on different collaboration modes. Then, we outline the process of how HUMTECH synergies emerge and outline its real-life applicability in B2B contexts. We conclude with a discussion of our model’s theoretical contributions, implications for managers, and directions for future research.

2. What are HUMTECH Synergies?

Most previous literature on the role of technology at the organizational frontlines has centered on the debate over whether technology has the potential to augment or replace humans (e.g., De Keyser et al., 2019, Huang and Rust, 2018, Larivière et al., 2017, Marinova et al., 2017, Mende et al., 2019, McLeay et al., 2021, Robinson et al., 2020; see supplementary Table A1 for an overview). Augmentation refers to technology’s potential to boost employee performance, while replacement refers to technology’s potential to automate and substitute human input (Marinova et al., 2017, Larivière et al., 2017). Importantly, these research streams primarily focus on technology with limited degrees of agency, that lacks “the ability to initiate action and accept rights and responsibilities for achieving optimal outcomes under uncertainty” (Baird & Maruping, 2021, p. 315). Traditional examples range from early-generation ATMs, check-in kiosks, and electronic ordering systems (de Keyser et al., 2019). Consequently, these studies treat technology as a passive tool, waiting to be used by employees or the firm to either take over individual tasks or assist employees in completing these tasks more effectively (Le et al., 2024). In any case, technology is implicitly considered subordinate to humans—be it an employee or a customer—in the service encounter (Baird & Maruping, 2021).

However, many forms of technology are increasingly collaborating with employees to co-produce the service and achieve superior outcomes (e.g., Blaurock et al., 2024, Brynjolfsson and Mitchell, 2017, Le et al., 2024, van Doorn et al., 2023, Wilson and Daugherty, 2018). Specifically, AT systems, sometimes referred to as “agentic IS artifacts” are “rational software-based agents that have the ability to perceive and act, such as take on specific rights for task execution and responsibilities for preferred outcomes” (Baird & Maruping, 2021, p. 317). Examples of B2B AT systems include virtual assistants such as Olivia, which works alongside recruiters throughout the recruiting process (Paradox, 2024) or automated portfolio management systems such as Goldman Sachs’ Marcus Invest (Marcus, 2024), which collaborates with human advisors and wealth managers to manage client portfolios. Other more business-to-consumer (B2C)-oriented examples include medical AI such as Mia which works with radiologists to treat breast cancer (Kheiron, 2024) and large language models such as ChatGPT, DALL-E, and Microsoft's Copilot that collaborate with employees—often in real time—to solve various customer problems (Frey & Osborne, 2023).

What sets these systems apart from other forms of technology is their “agentic abilities” that allow AT to make rational and autonomous decisions under uncertainty in tasks with ambiguous requirements (Baird & Maruping, 2021). These agentic abilities, in turn, give rise to three key technology elements that distinguish AT from non-AT systems: a) autonomy in decision-making, b) learning and adaptability, and c) context awareness. First, AT systems can operate autonomously, having the ability to make independent decisions rather than relying on rigid, predefined rules. Importantly, depending on the extent of their agentic abilities, the decision-making latitude of AT systems can vary greatly—from making autonomous decisions in response to immediate stimuli and providing guidance based on predefined models to autonomously anticipating and prescribing actions for managing complex, outcome-oriented tasks (Baird & Maruping, 2021). Second, AT systems continuously learn and adapt, improving their responses and decisions over time based on previous interactions with customers and employees (Lecun et al., 2015). Third, AT systems are context-aware, meaning they can recognize, interpret, and dynamically respond to changes in their environments without requiring explicit human input. Unlike non-AT systems, AT systems can monitor their surroundings and anticipate “the need to act without being prompted by users” (Baird & Maruping, 2021, p. 316).

Together, these three technology elements enable AT to act as “digital employees” and “partners”—rather than mere tools—to employees (Le et al., 2024). Instead of simply replacing or augmenting human abilities, AT can assume responsibility for a wide range of tasks throughout the service process. These tasks can be either functional—related to work design, structure, and organization (Tsai et al., 2022)—or relational, referring to the emotional support that customers receive (Eckardt et al., 2021). Moreover, AT can assume responsibility for tasks that are highly interdependent, with employees and AT relying on each other’s activities to achieve a collective goal (O’Neill et al., 2023). Human-technology collaboration thus shifts the focus from the execution of individual tasks to the delegation and coordination of interdependent tasks at the service process level (Le et al., 2023, Le et al., 2024). Unlike replacement and augmentation, when employees collaborate with AT, the superiority of the former is no longer a given and AT is not always subordinate to employees. Instead, the roles and responsibilities of humans and technology are fluid and can shift back and forth throughout the service process, depending on the nature of the task and the relative strengths of each partner, even within a single customer interaction (Baird & Maruping, 2021).

Successful human-technology collaboration, therefore, critically hinges on leveraging each partner’s respective strengths or compensating for each other’s weaknesses. In other words, employees and AT need to complement and augment each other to create synergies as they collaborate throughout the service process to achieve superior outcomes. However, while prior research highlights the importance of complementarities between employees and technology for human-technology collaboration (e.g., Huang and Rust, 2022, Pinski et al., 2023, van Doorn et al., 2023), little is known about the nature of these complementarities and how they can be leveraged to achieve HUMTECH synergies at the organizational frontlines.

Synergy can generally be described in terms of leveraging complementarities, defined as the “beneficial interplay of the elements of a system, where the presence of one element increases the value of others” (Ennen & Richter, 2010, p. 207). As Subramony (2009, pp. 746-747) states: “The accumulation of the actions of various elements can create combined effects larger than what can be expected when these elements operate in isolation … [and] when two or more elements operate together to serve a common function, it is possible for the system to conserve energy and reduce risks.” In the context of human-technology collaboration, Raisch and Krakowski (2021, p. 193) state that synergies arise when “humans and machines … combine their complementary strengths, enabling mutual learning and multiplying their capabilities.” Similarly, Blaurock and colleagues (2024) outline how AT systems need to allow for “reciprocal strength enhancement” between employees and the technology in order to achieve tangible service improvements and enhanced employee outcomes. Thus, the existence of complementary resources is a necessary but insufficient condition to achieve synergies (Harrison et al., 2001). Rather, the existence of complementary strengths between humans and technology offers opportunities for synergies that—when effectively leveraged by either the employee or the technology (or both)—can lead to superior outcomes beyond what either party could achieve alone (Gass, 2023).

In line with the above, we define HUMTECH synergies as the outcome of the beneficial interplay between a human employee and technology, where the employee leverages the complementary potential of the technology (and/or the technology leverages the complementary potential of the employee) to achieve superior outcomes that either party could not achieve alone. Notably, our definition allows for both employees and technology to take the lead in leveraging each other’s complementary potential. In line with the rapid advancement of AT, recent research highlights how both technology and humans can take on the roles of leaders, followers, or partners in human-technology collaboration, with different downstream consequences for both employees and customers (e.g., Shanks et al., 2024, Tsai et al., 2022, van Doorn et al., 2023). Accordingly, we view HUMTECH synergies to arise from three distinct modes of human-technology collaboration: a) employee-led, b) technology-led, and c) co-led (see Table 1 for an overview). For each human-technology collaboration mode, we provide an illustrative archetype for developing HUMTECH synergies in the context of B2B recruiting.

Table 1. Human-technology collaboration modes.

Employee-led HUMTECH synergies. Employee-led HUMTECH synergies refer to the synergies that arise in human-technology collaboration when an employee takes responsibility for accomplishing a set of interdependent tasks to achieve organizational goals. Decision-making authority lies with the employee in terms of both task execution and coordination. For HUMTECH synergies to arise, this mode of human-technology collaboration requires human leadership in delivering task instructions to technology and adapting these instructions based on the input received from technology (Tsai et al., 2022). Task interdependence is largely sequential, requiring employees to ensure a smooth hand-off of work between technology and themselves throughout the service process (O’Neill et al., 2023). The role of technology, in turn, is that of a follower. This involves autonomously executing a predetermined set of actions and making independent decisions within a constrained scope. It also requires sufficient context awareness to accurately detect and follow employee instructions while suggesting next steps, as well as learning to adapt its responses based on prior interactions with employees (Tsai et al., 2022). In such employee-led collaboration, technology often takes one of two roles: Either it functions as a “reflexive agent,” directly responding to a wide range of employee queries with a restricted scope of autonomous decision-making (e.g., virtual assistants, automated systems for rebalancing financial portfolios). Alternatively, it functions as a “supervisory agent,” acting as a control system for employees by monitoring their behavior, reminding them of process steps, and suggesting adjustments to stay on track with human-defined goals (e.g., behavior modification systems or guidance systems) (Baird & Maruping, 2021). Nevertheless, the decision-making authority to accept or ignore these technological suggestions ultimately lies with the employee.

To illustrate, consider the role of AT in recruiting where collaboration between recruiters and technology is becoming the norm (Day, 2024). In B2B recruiting, search firms develop long-term, trust-based relationships with clients. These relationships are more complex than one-time hiring decisions because they involve higher relational and emotional stakes on the client side, requiring ongoing collaboration and continuous refinement of hiring strategies based on evolving client needs. B2B recruiting typically consists of three main stages: a) searching for and screening potential candidates, including deciding which resumes to consider, b) selecting a small set of candidates for interviews and conducting those interviews, and c) making a recommendation to the client on which candidate to hire.

In a purely employee-led collaboration scenario, employees take the lead across all these three stages. Leveraging their deep understanding of client preferences and their organizational culture, employees instruct technology to autonomously perform the initial search and screening of candidates based on a set of criteria. The technology refines these criteria based on observed patterns and employee feedback over time. Employees then review the shortlisted resumes and decide which candidates to consider, ensuring alignment with the client’s culture and needs. Technology follows employee instructions while anticipating and recommending next steps, such as scheduling interviews, supplying relevant candidate data, and offering to conduct parts of the interviews to rank and suggest potential hires. Although the technology offers suggestions, employees retain decision-making authority as they accept or ignore these suggestions. In such an employee-led scenario, HUMTECH synergies hinge on the employees effectively delivering task instructions and technology accurately detecting and executing them. Client trust is established based on the employees’ ability to align their instructions and judgements with the needs of their clients over repeated engagements.

Technology-led HUMTECH synergies. Technology-led HUMTECH synergies refer to synergies that arise in human-technology collaboration when technology assumes responsibility for accomplishing a set of interdependent tasks to achieve organizational goals. Decision-making authority for the execution and coordination of these tasks lies with the technology. For HUMTECH synergies to arise, this mode of human-technology collaboration requires technology to take the lead in delivering task instructions to employees and in adjusting these instructions based on employee performance or customer feedback. This requires a high degree of autonomous decision making and context awareness, as technology must review employee performance, monitor customer reactions, and provide clear, easily understandable feedback to employees while continuously adapting its guidance through ongoing learning (Kadolkar et al., 2024). In such a technology-led collaboration, technology acts either as an “anticipatory agent,” proactively anticipating the needs of employees and customers for a specific (set of) task(s) or as a “prescriptive agent” by prescribing or acting autonomously without the need for human input (Baird & Maruping, 2021). Task execution is largely sequential with technology overseeing the transition of work between technology and employees throughout the service process (O’Neill et al., 2023). The role of employees in such technology-led collaboration is twofold. First, employees take on the role of followers, following the technological directives to perform specific actions (Tsai et al., 2022). Second, just as technology acts as a “supervisory agent” to employees in employee-led collaboration, employees can also maintain a supervisory role in technology-led interactions and provide suggestions to keep with technology-defined goals. However, the decision-making authority to accept or ignore these employee suggestions ultimately lies with the technology.

Consider again the B2B recruiting example. In a purely technology-led collaboration scenario, technology takes the lead throughout the recruiting process—starting with analyzing the job description, past hiring patterns, and client directives to autonomously set search and screening criteria (as seen, for example, in recruiting platforms such as HireVue). From there, it handles the initial search and screening of candidates, decides which candidates to interview, determines the interview process, and conducts the final assessment and recommendation to the client. Given the long-term nature of B2B recruiting, technology must not only optimize efficiency but also ensure consistency and alignment with the client’s evolving needs over time. Employees follow the technology’s prompts, handling specific actions as instructed by the technology. This may include providing feedback to adjust selection criteria or handling parts of the interview to evaluate specific skills and fit with the company. However, the final decision-making authority lies with the technology, which may accept or ignore employee input. In this scenario, HUMTECH synergies hinge on the technology’s ability to execute and coordinate tasks while effectively guiding employees, with employees supporting and monitoring these processes. Rather than relying on employee judgment, client trust is established through the technology’s reliability and adaptability in delivering instructions that align with client needs over repeated engagements.

Co-led HUMTECH synergies. Co-led HUMTECH synergies refer to synergies that arise in human-technology collaboration when an employee and technology share responsibility for accomplishing a set of interdependent tasks to achieve organizational goals. In contrast to employee-led or technology-led collaboration modes, employees and technology act as partners (Tsai et al., 2022). Decision-making authority for task execution and coordination is shared between employees and technology as both parties contribute equally to goal setting, task execution, and task coordination. Crucially, both parties must be aligned at each step of the service process, as neither party can override the other party’s decisions. Rather than being executed sequentially with clear task allocation to either the employee or technology, task interdependence is higher, with tasks often being executed jointly in real-time, involving continuous feedback loops and regular hand-offs of sub-tasks (O’Neill et al., 2023). This increased task interdependence requires ongoing interaction between employees and technology. Both partners need to provide and adjust to mutual feedback as they work towards a shared goal, thus requiring high degrees of technology autonomy, context awareness, and ongoing learning. Instead of focusing on the quality of directives, this collaboration mode relies on the quality of the dyadic interactions between employees and technology (Tsai et al., 2022) where both technology and “employees must embrace, interact with, and integrate their behaviors” (Makarius et al., 2020, pp. 262-263) to realize a deep level of involvement. Compared with employee-led or technology-led collaboration modes, co-led collaboration places greater emphasis on team dynamics, particularly on shared mental models that enhance communication, trust, and coordination between employees and technology (Schelble et al., 2022) that ultimately allow HUMTECH synergies to emerge.

Drawing again on the B2B recruiting example: in a co-led collaboration, both employees and technology interact with each other throughout the process of setting selection criteria, deciding which candidates to consider, determining how the interview should unfold, and identifying the candidates to recommend to the client. Both employees and technology must align at each stage, with neither having overriding authority. Here, HUMTECH synergies crucially depend on the ongoing feedback and mutual adjustments of both parties. Client trust is established through the seamless coordination of tasks and responsibilities between technology and employees in line with evolving client needs over time.

Notably, these three human-collaboration modes are not mutually exclusive. Rather, to achieve superior performance outcomes, employees might need to assume responsibility for some tasks, while technology takes on responsibility for others. For example, technology might be better suited to leading the search and screening of potential candidates and humans to leading the interview stage, while a co-led collaboration effort might be best for deciding on the final candidates to recommend to the client. Equally, customer preferences may dictate whether employees or technology (or both) take responsibility for certain tasks. For example, customers may prefer technology to conduct interviews to ensure more objective procedures, while preferring employees to make the final decisions. The emergence of HUMTECH synergies therefore crucially depends on the selection of the “right” human-technology collaboration mode throughout the service process. As shown in Fig. 1, this choice, in turn, hinges on the a) existence, b) recognition, and c) enactment of complementarities between employees and technology. Key elements include the readiness of employees and technology, their metaknowledge, and fit appraisals. Next, we detail each of these elements.

1. Download: Download high-res image (300KB)
2. Download: Download full-size image

Fig. 1. Creating HUMTECH synergies at the organizational frontlines.

Our HUMTECH synergies emergence model is grounded in cognitive appraisal theory (Lazarus, 1991, Lazarus and Folkman, 1984). Previous research frequently uses cognitive appraisal theory as a valuable lens to understand how humans respond to (the use or adoption of) technology in general (e.g., Beaudry and Pinsonneault, 2005, Bhattacherjee et al., 2017, Fadel and Brown, 2010) and in the context of human-technology collaboration in particular (e.g., Paluch et al., 2022, Pinski et al., 2023). Cognitive appraisal theory explains how humans cope with challenging or unfamiliar stimuli—such as a task—in their environment (Lazarus & Folkman, 1984). When confronted with a task, the way humans choose to cope with this stimulus depends on their appraisal of the situation in relation to the coping resources available to them. Folkman et al. (1986, p. 992) define appraisal as “a process through which the person evaluates whether a particular encounter with the environment is relevant to his or her well-being, and, if so, in what ways.” This primary appraisal is complemented by a secondary appraisal, where the individual evaluates their coping options (i.e., the options they have at their disposal to respond to the demands of the situation), considering the control they have over the situation and the resources available to them (Lazarus & Folkman, 1984). Importantly, these resources might include not only their own resources to handle a given task but also the coping resources of a partner or a subordinate (Pinski et al., 2023).

In the context of human-technology collaboration, both employees and AT may engage in appraisal to evaluate their own and each other’s attributes for a given process or task (Baird & Maruping, 2021). Appraisal thus lies at the heart of HUMTECH synergies as it represents the fundamental mechanism for finding complementarities between employees and technology at the organizational frontlines. Accordingly, we use cognitive appraisal theory to explain the numerous appraisal processes that occur throughout the HUMTECH synergies emergence process. First, we explain how organizations engage in appraisal to assess whether potential complementarities a) exist between employees and technology. We then delineate how employees and technology engage in various appraisals when confronted with a specific task to b) recognize these complementarities. Finally, we illustrate how the outcomes of these appraisals enable c) the enactment of complementarities, which ultimately leads to the realization of HUMTECH synergies at the organizational frontlines.

For HUMTECH synergies to emerge, complementarities need to exist among its most foundational elements, namely employees and technology (Fügener et al., 2022). Both employees and technology are endowed with resources that firms may access to achieve HUMTECH synergies at the organizational frontlines (Ennen & Richter, 2010). The intrinsic potential, or “readiness,” of these resources forms the basis for potential complementarities between them, ultimately shaping the possibility of the emergence of HUMTECH synergies.

Employee readiness. Readiness, a concept rooted in motivational theory, generally signifies an entity’s potential to engage in future behaviors (Atkinson, 1964, Kruglanski et al., 2014). When applied to humans, readiness encapsulates both abilities and motivation to engage in purposeful behaviors. In line with cognitive appraisal theory (Lazarus & Folkman, 1984), readiness can therefore be understood as the potential resources or attributes that humans have at their disposal to cope with a challenging situation (e.g., task). Traditionally, human readiness research has explored the readiness of customers and employees to “function” in dyadic service interactions––whether human-to-human or human-to-technology. This entails the readiness required to perform job-related tasks, utilize self-service technologies, or adhere to firm-prescribed instructions (e.g., Bowen, 1986, Dellande et al., 2004, Meuter et al., 2005).

Recently, Danatzis et al. (2022) introduced a broader readiness concept, termed “actor ecosystem readiness,” offering a valuable theoretical lens for analyzing the abilities and motivation that employees require to effectively navigate today’s interconnected frontline interactions (Bowen, 2024, Marinova et al., 2017). Defined as “the abilities and motivation embedded in a human actor that form his or her intrinsic potential to collaborate with multiple actors and use accessible [technological] resources in a service ecosystem” (Danatzis et al., 2022, p. 261), this human readiness concept delineates the cognitive, emotional, interactional, and motivational factors that enable effective collaboration and use of technology, among other resources, in multi-actor settings such as frontline interactions. This involves the ability to make decisions; clarify, align, and systematically pursue self-set or organizational goals; and undertake adaptive thinking (cognitive readiness). It also includes the ability to effectively manage one’s own feelings and those of others (emotional readiness), while effectively connecting, communicating, and quickly adapting to unfamiliar norms, rules, or beliefs in social settings (interactional readiness)—all driven by one’s desire and confidence to achieve certain outcomes (motivational readiness) (Danatzis et al., 2022).

Technological readiness. Similar to employee readiness, technological readiness refers to a technology’s endowments (i.e., capabilities) and preferences (i.e., decision models) that form its overall potential to pursue self-set or human-set goals. Specifically, preferences refer to decision models that use utility functions to rank choices, which technology can leverage to make goal-directed decisions under uncertainty (Russell & Norvig, 2016). In this way, decision models function similarly to human motivational readiness in that their choices reveal specific preferences for specific outcomes, although technology is often unable to provide explanations behind their choices—an issue known as “the black-box problem” (Hassija et al., 2024). Endowments, in turn, refer to a set of capabilities that allow the technology to sense, analyze, and act (Baird & Maruping, 2021). Like humans, technology needs to be ready to make autonomous decisions, formulate and pursue self-set or human-set goals to achieve particular outcomes, adjust to evolving requirements and situations, and effectively interact and connect with humans—be they employees or customers (Danatzis et al., 2022). In short, technological readiness increasingly mirrors human readiness, underscoring the need to align both readiness factors to foster potential complementarities between employees and technology at the organizational frontlines.

Readiness-to-process appraisal. To achieve potential HUMTECH synergies, organizations need to ensure that the readiness of their employees is not identical or symmetrical, but complementary to the readiness of their technology. Although organizations must ensure there is enough shared readiness between their employees (e.g., willingness to work with the technology; Paluch et al., 2022) and their technology (e.g., ease-of-use, collaborative system features; Blaurock et al., 2024) to allow for human-technology collaboration in the first place, they also need to make sure that both employees and technology possess different strengths that align with the specific demands of the service process. We refer to this organizational screening process as readiness-to-process appraisal, defined as the organization’s assessment of the existence of potential complementarities between the readiness of employees and technology relative to the expected demands of the service process.

In line with cognitive appraisal theory (Lazarus & Folkman, 1984), relevant managers in the organization must engage in two appraisal processes to this end. First, they must assess the demands of the service process and the organizational goals they seek to achieve (primary appraisal). Second, they need to evaluate whether sufficient complementary readiness exists between their employees and technology to meet the demands of a given service process (secondary, employee-technology fit appraisal). Ideally, any readiness shortcomings of one party should be balanced out by the readiness of the other party (Danatzis et al., 2022). This, in turn, requires an in-depth analysis of the respective readiness profiles of both employees and technology vis-a-vis the expected demands of (different tasks of) the service process.

For example, employees (or technology) scoring low on emotional readiness (ability to sense and respond to emotions; Danatzis et al., 2022), could be “matched” with technology (or employees) that is (are) specifically trained to handle difficult emotional encounters with customers (Wieseke et al., 2012). Likewise, employees with the ability to recognize and mitigate potential biases in AI algorithms (e.g., race and income biases in the data used to train algorithms in financial services for determining loan eligibility) can complement technology by compensating for potential limitations in fairness and ethical decision-making (Blaurock et al., 2024, Wirtz et al., 2023). Additionally, specifically trained employees, such as domain experts and specialists, can be paired with AT systems prone to hallucinations to balance out technology shortcomings. For example, Cohere, a B2B firm, employs highly trained professionals with advanced degrees to refine AI-generated outputs and avoid hallucinations (Mukherjee & Tong, 2024).

Alternatively, organizations could pre-assign responsibilities for specific subprocesses, tasks, or touchpoints to employees and technology with complementary readiness profiles (Lemon & Verhoef, 2016). For example, a B2B logistics company could pre-assign routine shipment tracking and status updates to AI-driven chatbots, while reserving more complex customer interactions—such as handling high-value or time-sensitive shipments—for employees with strong interactional readiness, such as relational skills (to build trust with clients) and communication skills (to resolve disputes and provide reassurance in high-stakes situations) (Danatzis et al., 2022). Conversely, if the technology possesses advanced cognitive readiness for specific tasks, such as the ability to detect and respond to unusual patterns or anomalies in shipment data in real-time, it could be assigned these more complex tasks while employees handle routine communication and coordination. This pre-assignment ensures that each party focuses on subprocesses, tasks, or touchpoints best suited to their readiness, thereby enhancing overall service efficiency.

Finally, organizations could strategically invest in the development of specific readiness aspects of their employees or technology to overcome readiness deficiencies that hinder complementarities (Danatzis et al., 2022). For example, B2B firms such as Amazon Web Services invest in automated reasoning techniques to enhance the readiness of their technology to detect and correct AI hallucinations (Amazon Science, 2025). Alternatively, firms could improve employee readiness by providing transparency into how AI algorithms work, enabling them to identify and compensate for biases (Blaurock et al., 2024). Firms could also cultivate a corporate digital responsibility culture to equip employees with the necessary skills and authority to assess and override biased AI recommendations (Wirtz et al., 2023). Regardless of whether organizations choose to match, pre-assign, or develop the readiness of their employees and technology, they need to engage in readiness-to-process appraisal to ensure the existence of complementarities and potential HUMTECH synergies.

Readiness-to-task appraisal. The mere existence of complementary employee and technological readiness does not result in any HUMTECH synergies. Rather, possessing complementary readiness is a necessary but insufficient condition to achieve these synergies (cf. Harrison et al., 2001). For HUMTECH synergies to emerge, these complementarities must also be recognized by either the technology or the employee (or both) when confronted with a specific task. That is, a decision-making “partner needs to recognize that complementarities between the two partners exist and that the tasks should be performed by the better-suited partner” (Fügener et al., 2022, p. 679). We refer to this cognitive appraisal process (Lazarus & Folkman, 1984) as readiness-to-task appraisal, which we define as the employee’s or technology’s assessment of the fit of their own readiness vis-a-vis the readiness of the other party relative to the demands of a given task. Unlike readiness-to-process appraisal in which organizations engage to find potential readiness complementarities between technology and employees relative to the anticipated demands of the service process, these complementarities must also be identified by those directly responsible for the task—the technology and/or the employee themselves. As such, this appraisal does not occur at the organizational level, but at the individual (task) level.

Drawing on cognitive appraisal theory (Lazarus & Folkman, 1984), we theorize that three different readiness-to-task appraisal processes take place that facilitate the recognition of complementarities at the organizational frontlines (cf. Pinski et al., 2023). First, the employee and/or the technology engage in an employee-fit appraisal to evaluate how well the employee's readiness matches the demands of a specific task. Second, the employee and/or technology simultaneously engage in a technology-fit appraisal to assess the degree to which the readiness of the technology aligns with the requirements of that task. Both of these task appraisals can be understood as primary appraisals (Lazarus & Folkman, 1984) through which the employee and/or the technology evaluate the demands of a specific situation (i.e., task) in light of the resources available to them (i.e., employee and technological readiness).

These primary appraisals are followed by a secondary appraisal, which we label comparative-fit appraisal. In this stage, the employee and/or technology evaluate their coping options by assessing the comparative fit of their readiness against that of the other in matching task demands. Unlike the two primary appraisals, this secondary readiness-to-task appraisal is not about determining how “ready” the employee or the technology is for a given task. Instead, it focuses on whether the employee or the technology is “equally ready” or “more ready” to complete the task. It is this ability to evaluate one’s own readiness vis-a-vis the readiness of the other party to perform a task that enables these readiness-to-task appraisals—a higher-order cognitive process, often referred to as metaknowledge (Evans & Foster, 2011).

Metaknowledge. Broadly understood as “knowledge about knowledge” (Evans & Foster, 2011, p. 721), metaknowledge refers to the ability of a decision-maker—whether human or technology—to assess their own capabilities (Fügener et al., 2022). Employees rely on metacognitions for metaknowledge which are second-order cognitive processes that monitor and control human reasoning and control the allocation of mental resources (Ackerman & Thompson, 2017). Metacognitions therefore allow employees to reflect on their own thinking, and to decide whether and how to consider available information—be it from the technology or another employee (Jussupow et al., 2021). Technology, in turn, relies on learning from large training datasets to develop flexible models that enable adaptation to new data, feedback, and unanticipated situations (Lecun et al., 2015). Specific architectural features, such as meta-cognitive loops, enable AT systems to build metaknowledge through endowing these systems with monitoring, self-modeling, and repair abilities that allow the systems to “reason” about their own abilities (Schmill et al., 2008). Previous information systems research stresses the importance of metaknowledge for effective human-technology collaboration (e.g., Fügener et al., 2022, Jussupow et al., 2021, Pinski et al., 2023, Gass, 2023, Pinski et al., 2024). With the rapid development of AT, possessing sufficient metaknowledge becomes increasingly critical as it becomes more ambiguous who—the technology or the employee—is better suited (i.e., more “ready”) to perform a specific task (Pinski et al., 2024).

For example, recent empirical research shows that when physicians collaborate with medical AI in cancer diagnosis, having sufficient metaknowledge is key for successful collaboration with the AI system. Specifically, shortcomings in metaknowledge have been found to lead to incorrect diagnostic decisions because physicians were unable to adequately assess how well their own skills (“self-monitoring”) or that of the AI (“system-monitoring”) matched the demands of a given task, and, as a result, their decisions were primarily guided by beliefs rather than objective data (Jussupow et al., 2021). In short, shortcomings in metaknowledge impede both employee-fit appraisals and technology-fit appraisals. Similarly, metaknowledge is equally critical for comparative-fit appraisals. For instance, Fügener et al. (2022) found that employees (in contrast to technology) struggle to evaluate their own readiness in comparison with the readiness of technology for a given cognitive task. Notably, the authors find that this difficulty is not due to the employees’ general reluctance to use technology, but because of insufficient metaknowledge (Fügener et al., 2022), which, in turn, leads to poor human-technology collaboration performance. Hence, the greater a party's metaknowledge, the better will they be able to engage in readiness-to-task appraisals, which, in turn, is essential for recognizing complementarities and achieving HUMTECH synergies.

Once complementarities are recognized, they must be enacted to allow HUMTECH synergies to emerge. This enactment entails four subprocesses: a) task delegation, b) task execution, c) task coordination, and d) task evaluation. In line with cognitive appraisal theory (Lazarus & Folkman, 1984), these processes represent coping mechanisms chosen in response to a task. Employees (or technology) make(s) these choices based on the outcomes of prior appraisals.

Task delegation. Once one party recognizes that the other party is better suited to perform a particular task, that party must follow through with delegating the tasks to the better suited party (Fügener et al., 2022). In other words, delegation must be undertaken by the party considered “more ready”—whether that is the employee or the technology. Delegation refers to the transfer of rights and responsibilities for the execution and outcomes of a task to another human or non-human agent (Baird & Maruping, 2021). Notably, this transfer can be either complete or partial (Baird & Maruping, 2021). Indeed, in co-led HUMTECH synergies, partial delegation is the norm as employees and technology share these rights and responsibilities for task execution and outcomes. Similarly, in technology-led HUMTECH synergies in high-stakes environments, such as medical diagnosis, a lack of trust in technology may lead employees to retain certain rights and responsibilities despite delegating most of them to technology, even if the technology is deemed “more ready” (Pinski et al., 2023). Beyond the transfer of rights and responsibilities, delegation therefore always involves some degree of negotiation. Negotiation, in turn, entails agreeing on the rules that will govern how both parties will work with one another in order to achieve a desired task outcome. It also includes the setting of (implicit or explicit) expectations and boundaries for how the task should be executed and what the task outcome should look like (Baird & Maruping, 2021).

Regardless of the degree of delegation, recent research on human-AI collaboration shows that AI demonstrates better delegation of tasks to humans than vice versa (Fügener et al., 2022). This poor human delegation performance often stems from insufficient general knowledge about the strengths and weaknesses of technology, which impedes humans in assessing how well the technology’s readiness matches the demands of a task (Pinski et al., 2023). Poor human delegation performance may also stem from insufficient metaknowledge, either in the form of overconfidence in one’s own readiness to handle a given task or one’s inability to evaluate one’s own readiness relative to the readiness of the technology (Fügener et al., 2022, Pinski et al., 2024). Although challenging, increasing technology literacy and metaknowledge through, for example, gamified learning experiences, has been shown to foster readiness-to-task appraisals and lead to better human delegation performance (Pinski et al., 2023, Pinski et al., 2024). Technology delegation performance, in turn, can be improved by increasing the system’s overall prediction accuracy through more training data or the design of specific guidelines for collaboration (Fügener et al., 2022).

Task execution. When a delegation decision has been made, the party to which the task has been delegated must then choose whether or not to accept the decision-making authority for executing the given task. Depending on whether a task is delegated to a single party (as in technology-led or employee-led HUMTECH synergies) or to multiple parties (as in co-led HUMTECH synergies), all parties must accept the delegation decisions for their respective (parts of the) task. Otherwise, the task will not be executed and no HUMTECH synergies will emerge. Prior research suggests that humans would be more willing to accept delegation decisions if they trust and have confidence in the delegation process (Bonaccio and Dalal, 2006, Hancock et al., 2011) and if the benefits of accepting the delegation decision outweigh the risks and costs in carrying out the task (Miller and Parasuraman, 2007, Steffel et al., 2016). For technology, in turn, accepting a delegation decision depends on how well it aligns with optimizing a predefined outcome (Baird & Maruping, 2021). For instance, if the goal is to minimize task execution time, the technology may accept the delegation owing to its superior processing capabilities. Yet, if the focus is on optimizing emotional rapport with customers, the technology might decline the task and re-delegate it to employees instead.

The quality of task execution, in turn, depends on the readiness of employees and technology and how well this readiness aligns with the actual demands of a given task. As discussed earlier, both humans and technology engage in various readiness-to-task appraisals to assess their readiness relative to the expected demands of a task before delegating the task to either an employee or technology. However, task demands are often ambiguous, and their complexity may only become evident during the execution of the task itself (Campbell, 1988, Liu and Li, 2012). Thus, the higher the complexity of a task, the higher the risk of a mismatch between the readiness of the executing agent and the unfolding task requirements, resulting in an inferior demands-ability fit (Danatzis et al., 2022). Referring to “a situation in which one’s knowledge, skills, and abilities are [not] commensurate with what the job [or the task] requires” (Kristof-Brown et al., 2005, p. 284), such an inferior demands-ability fit will likely reduce the quality of task execution, leading to suboptimal task outcomes.

Task coordination. Tasks do not exist in isolation. Rather, service processes at the organizational frontlines consist of multiple bundles of tasks (Sampson & dos Santos, 2023), some of which are executed by technology, others by employees, and some jointly. Importantly, these tasks are often highly interdependent, as the outcome of one task might impact the execution of other tasks (O’Neill et al., 2023). For HUMTECH synergies to emerge, it is therefore crucial not only that each task is delegated and executed well in isolation but also that the tasks are coordinated effectively. Coordination generally refers to the management of task dependencies and is considered key to the success of collaborative efforts among partners within an organization (Baird & Maruping, 2021). In such collaborative contexts, coordination can be defined as “the deliberate and orderly alignment or adjustment of partners’ actions to achieve jointly determined goals” (Gulati et al., 2012, p. 12). Given its clear division of labor (Le et al., 2023, Le et al., 2024), the very nature of human-technology collaboration requires such an alignment and adjustment of partner actions. Notably, while it is important that both partners are aligned in their goals, effective coordination also requires alignment with regard to which partner will execute which task by when and how. In short, coordination requires joint planning that enables compatible timing and the meaningful sequencing of (sub)tasks (Gulati et al., 2012, Palmer, 1983).

Importantly, the extent of task coordination required to achieve HUMTECH synergies can vary substantially depending on the degree of task interdependence and task complexity (Campbell, 1988, Liu and Li, 2012, O’Neill et al., 2023). For example, co-led HUMTECH synergies will generally require higher task coordination given the increased task interdependence and the need to subdivide tasks into a number of smaller subtasks that can each be handled by one of the partners (Lee & Siemsen, 2017). In contrast, employee-led or technology-led HUMTECH synergies will likely require less task coordination owing to the mostly sequential nature of task execution and, subsequently fewer hand-offs of work between technology and employees. Nevertheless, regardless of the degree of coordination, task coordination always necessitates the establishment of some coordination mechanisms that allow the effective sharing of information and provision of feedback between technology and employees (Gulati et al., 2012). These mechanisms, in turn, require ongoing evaluation of the progress of task accomplishment throughout the service process.

Task evaluation. The effective delegation, execution, and coordination of tasks all require continuous evaluation of how tasks are being accomplished in relation to advancing organizational goals (Baird & Maruping, 2021). Importantly, task evaluation entails assessment of not only how well technology and/or employees delegate, execute, and coordinate tasks to progress toward a specific goal, but also the degree to which these actions deviate from these goals and the ability to formulate and implement necessary adjustments to maintain or restore alignment with them. In fact, it is this ability to learn from past outputs and adapt behavior or processes to altered situations during the service process that is often highlighted as essential for effective human-technology collaboration (Huang and Rust, 2018, Huang and Rust, 2022). Importantly, task evaluation does not necessarily need to be carried out by the agent who executes and/or coordinates the given (set of) tasks. For example, as previously discussed, in human-led HUMTECH synergies, technology often acts as a “supervisory agent” that monitors employees' progress toward goals (Baird & Maruping, 2021). Similarly, employees frequently engage in task evaluation, even in technology-led HUMTECH synergies, particularly when ethical concerns are significant (Pinski et al., 2023). In any case, task evaluation can be understood as an essential mechanism that underlies the enactment of complementarities, without which HUMTECH synergies would not emerge.

Overall, our proposed emergence model uncovers the three-stage process that underlies the creation of HUMTECH synergies at the organizational frontlines. Crucially, these synergies evolve iteratively. That is, the existence, recognition, and enactment of complementarities is an ongoing process contingent upon the changing readiness and metaknowledge levels of the technology and employees over time. These changes are driven by employee training and learning on the one hand and technological advancements on the other hand, in response to evolving client needs, regulatory shifts, and broader industry developments.

Next, we showcase the applicability of our HUMTECH synergies model for B2B services. Notably, our collaboration modes and our model equally apply to B2C and B2B services. Although B2B services are typically more relationship-focused—characterized by long-term engagements, trust, and continuity through dedicated point-persons, such as advisors, client-relationship managers, and consultants (Subramony & Holtom, 2012)—these relational dynamics do not change the fundamental requirement for HUMTECH synergies to emerge. Regardless of the context, HUMTECH synergies will only unfold when complementarities between employees and technology a) exist, b) are effectively recognized, and c) are enacted.

To illustrate the applicability of our HUMTECH synergies model for B2B services, we next demonstrate how HUMTECH synergies emerge according to our framework in two key B2B contexts: B2B sales and B2B legal services. We chose these B2B contexts because they differ fundamentally in service timing, task complexity, relationship dynamics, and collaboration requirements, thus demonstrating how our framework applies across diverse B2B service types. B2B sales are initiated during the pre-purchase phase and center on establishing trust and guiding potential clients through complex purchasing decisions. This involves service tasks, such as explaining to potential customers the capabilities of often complex B2B products, addressing client concerns, and preparing tailored proposals. In contrast, B2B legal services take place post-purchase and are distinctly professional services that require ongoing collaboration, deep expertise, and sustained trust to navigate complex legal frameworks. While B2B sales encompass a wide range of products and services, B2B legal services are highly specialized and embedded within long-term business relationships.

AT systems, such as AI and automation tools, are rapidly transforming the B2B sales process, with tools that, for example, automate repetitive tasks (e.g., data entry), optimize lead generation, and personalize communications at scale (Meier, 2023). According to a survey of 648 U.S. sales professionals, sales agents report saving over two work hours per day by using these tools, allowing them to spend more time on selling (HubSpot, 2023). In addition, AT systems appear to make sales agents more effective, for instance through the use of AI to identify high quality prospects (e.g., analyzing data from websites, social media, and contact databases) and build rapport with potential customers (e.g., customer sentiment analysis to generate personalized content based on the prospect’s behavior) (HubSpot, 2023, Meier, 2023). Our HUMTECH synergies model helps explain how these synergies emerge.

First, complementarities must exist between the readiness of the AT system and the readiness of the sales agent. To this end, sales firms must conduct readiness-to-process appraisals to ensure that the readiness of their sales agents and AT system is not symmetrical but complementary in meeting client needs. For example, technological readiness may equip the AT system with the capability to automate administrative tasks, enhance sales teams’ engagement with potential customers, and enforce security measures to prevent data misuse (HubSpot, 2023). Yet, for HUMTECH synergies to emerge, sales firms must ensure that the readiness of their sales agents complements, rather than merely mirrors, the readiness of their technology. This requires equipping sales agents with the cognitive, emotional, interactional, and motivational readiness needed to effectively collaborate with the AT system (Danatzis et al., 2022). For instance, sales agents must be able to interpret and act on AI-driven insights for personalized outreach, adapt to AI-enhanced workflows, and navigate complex client interactions where human judgement, persuasion, and emotional intelligence are required. This also entails breaking down the capabilities of often complex B2B products, addressing client concerns, and preparing tailored proposals to help potential customers make informed decisions. Without this alignment, insufficient or mismatched levels of employee and technological readiness will hinder the emergence of HUMTECH synergies from the outset.

Second, complementary readiness alone is insufficient to achieve HUMTECH synergies. Instead, both the sales agent and the AT system must possess metaknowledge to recognize which party is more ready for a given task through readiness-to-task appraisals. For example, 86 % of sales agents recognize that AI is more efficient in drafting initial prospect messages, while 90 % agree that they themselves are “more ready” for refining these messages before sending them to potential clients (HubSpot, 2023). Failure to recognize these complementarities may result in inefficiencies, with tasks being misallocated to the “less ready” party due to inadequate employee-fit, technology-fit, or comparative-fit appraisals.

Third, to ultimately realize HUMTECH synergies, these complementarities must also be enacted. In the case of prospect messaging, for example, drafting initial outreach messages should be delegated to AT systems, while sales agents should refine and personalize them. Beyond effective delegation, HUMTECH synergies also require the seamless execution and coordination of tasks, along with continuous task evaluation. Notably, this involves not only monitoring task performance against client acquisition goals but also adapting workflows to shifting market conditions, internal sales targets, or client needs. Effectively enacting complementarities enables sales firms therefore to realize long-term HUMTECH synergies, ultimately enhancing sales efficiency, customer trust, and revenue growth in B2B sales.

B2B legal services are similarly developing HUMTECH synergies to provide better, more cost-effective, and even new adjacent services to customers. Take, for example, a law firm specializing in property management law that uses an AT system to analyze a client’s set of leases in their property portfolio (Spring et al., 2022). Given the highly specialized and deeply collaborative nature of B2B legal services, attorneys use these analyses to advise clients on how to manage their portfolios more holistically and improve their processes, thereby adding advisory services to their existing legal services. In so doing, the law firm strengthens its long-term relationships, a defining feature of professional B2B legal services.

For HUMTECH synergies to emerge, first complementarities need to exist between the readiness of the AT system (e.g., its ability to extract and organize relevant insights from a vast number of documents) and the readiness of the attorney (e.g., sensemaking of the AT output for advising clients on regulatory compliance and contractual obligations). Notably, the law firm needs to engage in readiness-to-process appraisals to ensure that the readiness of their attorneys and AT system is not symmetrical but complementary to meet client demands. Failure at this stage will result in insufficient or mismatched employee and technological readiness levels that fail to meet client expectations and associated service process demands, thus preventing the very possibility of the emergence of any HUMTECH synergies.

Second, both the attorneys and the AT system itself need to possess sufficient metaknowledge to engage in readiness-to-task appraisals to recognize whether the attorney or the AT system is more ready to accomplish specific tasks. For example, while an attorney would spend hours or days analyzing a set of leases in a client’s property portfolio, the AT system could complete the same task in minutes, ensuring compliance checks and risk assessments are conducted more efficiently. Conversely, attorneys may be “more ready” for tasks requiring creative or relational work such as client communication and strategy development (Marwaha, 2017). Failure to recognize such complementarities may lead to inefficiencies in leveraging the strengths of both the attorney or the AT system. For example, specific task demands may surpass the readiness of attorneys or the AT system owing to inadequate employee-fit or technology-fit appraisals, leading to downstream inefficiencies in task execution. Poor comparative-fit appraisals may also result in tasks being assigned to the “less ready” party, ultimately leading to diminished client trust and service quality.

Finally, these complementarities also must be enacted effectively to ultimately realize HUMTECH synergies. In this case, data-heavy tasks should be delegated and executed by the AT system while client-facing tasks should be completed by the attorney, so that more work will be done with less effort (Clio, 2024). Failure at this stage might not only lead to inefficiencies in delegation performance, task execution, and task coordination owing to incompatible timing and sequencing of (sub)tasks (Gulati et al., 2012). It will also lead to inefficient task evaluation, reducing the ability of attorneys or the AT system to identify misalignments and implement adjustments, ultimately compromising sustained client satisfaction and trust, thereby hindering the long-term realization of HUMTECH synergies.

Notably, this division of labor could shift depending on the nature of the tasks and the level of readiness or metaknowledge of attorneys or different AT systems. In any case, both B2B examples showcase the real-life applicability of our HUMTECH synergies model while demonstrating the distinct outcomes in terms of failures that may occur at each of its stages.

Advanced technologies such as cloud computing, robotics, and AI are rapidly transforming organizational frontlines. Given its increased agentic abilities, this new generation of technologies allows such AT to assume responsibility for a wide range of tasks that were previously handled by employees. However, rather than merely replacing or augmenting employees’ input, AT increasingly “partners” with employees to co-produce the service and achieve superior outcomes (e.g., Blaurock et al., 2024, Le et al., 2024, van Doorn et al., 2023), bringing the question of HUMTECH synergies to the forefront. This paper contributes to this emergent research stream on human-technology collaboration in two important ways.

First, we contribute to the discussion on how employee and technology roles at the organizational frontlines could be aligned to foster HUMTECH synergies, thus going beyond previous research that focuses on the augmentation, replacement, or the mere co-presence of employees and technology (e.g., Huang and Rust, 2018, Marinova et al., 2017, van Doorn et al., 2023). Synthesizing interdisciplinary literature streams from marketing, management, and information system disciplines (e.g., Danatzis et al., 2022, Ennen and Richter, 2010, Fügener et al., 2022, Raisch and Krakowski, 2021), we are the first to define HUMTECH synergies and conceptualize these synergies as arising from three distinct modes of human-technology collaboration: a) employee-led, b) technology-led, and c) co-led. By delineating the idiosyncratic features of each of these collaboration modes that give rise to HUMTECH synergies—depending on whether employees and technology act as leaders, followers, or partners—we critically advance previous research on human-technology collaboration in service settings (e.g., Blaurock et al., 2024, Le et al., 2024, van Doorn et al., 2023) and, more broadly, human–machine collaboration (e.g., Fügener et al., 2022, Tsai et al., 2022).

Second, previous research on human-technology collaboration typically assumes that synergies naturally arise from the integration of (advanced) technology at the organizational frontlines (e.g., Huang and Rust, 2022, van Doorn et al., 2023). However, these studies often overlook the mechanism underlying how such synergies develop. Our paper addresses this theoretical gap by outlining the process of how HUMTECH synergies emerge. Specifically, we integrate cognitive appraisal theory (Lazarus, 1991, Lazarus and Folkman, 1984) with literature on readiness (Danatzis et al., 2022) and metaknowledge (e.g., Fügener et al., 2022, Pinski et al., 2023) to uncover the three-stage process that underlies the creation of HUMTECH synergies. We argue that, first, complementarities need to exist between the readiness levels of both employees and technology, forming the basis for any potential synergies. Our model clarifies the nature of readiness of both employees and technology; further, it outlines how organizations engage in readiness-to-process appraisals to ensure that their employees and technology are complementary in their readiness in terms of meeting the expected demands of the service process. Second, complementarities need to be recognized by either the technology or the employee (or both) when confronted with a specific task, and the mere existence of complementarities will not, by itself, result in HUMTECH synergies. Facilitated by metaknowledge, our model outlines how both employees and technology engage in various readiness-to-task appraisals to evaluate how well their own readiness and that of the other party matches the demands of a specific task. Third, based on the outcomes of these appraisals, these complementarities also need to be enacted, which, in turn, requires the effective delegation, execution, coordination, and evaluation of tasks throughout the service process. By providing such a holistic, multi-stage theoretical framework of HUMTECH emergence, this paper is the first to address the theoretical “black box” of how complementarities between technology and employees can not only be established but also leveraged to successfully create HUMTECH synergies at the organizational frontlines.

Our HUMTECH model can help managers facilitate synergies between employees and technology by ensuring the selection of the appropriate technology for the firm, as well as enabling a human resource management system that attracts, trains, motivates, and retains employees who can collaborate with this technology. Understanding how employee and technology complementarity drive HUMTECH synergies is crucial, as it can help improve the chances of various technological innovations succeeding. For example, there is evidence that firms often struggle to achieve better customer experiences or efficiency gains at the organizational frontlines (Correani et al., 2020), even with the rapid adoption of AT across several business functions, including sales, marketing, and customer operations (McKinsey & Company, 2024). Often these failures arise because of a lack of complementarity between the readiness of employees and the readiness of technology. For instance, a recent survey reveals that only 28 % of frontline workers feel they are adequately reskilled to use advanced technologies (Boston Consulting Group, 2024). Meanwhile, many implemented technologies are not fit for purpose, often failing to deliver the expected synergy gains (Braun, 2025) and sometimes even reversing productivity improvements (Martin, 2024) as the lack of complementary readiness ultimately increases workload (Burleigh, 2024).

Thus, in echoing industry calls to “break free” of the “common narrative that [AT, such as] AI is a straightforward path to efficiency” (Nawrat, 2024), this paper not only equips managers with a better understanding of different modes of human-technology collaboration (i.e., employee-led, technology-led, and co-led) that can give rise to HUMTECH synergies. We also provide managers with an actionable three-stage process model to help them understand how such HUMTECH synergies can be enabled and why they sometimes fail to materialize. Applying our HUMTECH model thus allows firms to implement more targeted enablement strategies at each stage to ensure the realization of HUMTECH synergies at the organizational frontlines (see Table 2 for an overview of enablement foci, tasks, and strategies).

Table 2. Managerial enablement foci, tasks, and strategies across HUMTECH synergies emergence stages.

First, HUMTECH synergies may not emerge in the first place given insufficient or mismatched employee and technological readiness levels. To enable HUMTECH synergies, managers should therefore ensure that the readiness of their employees and technologies are not symmetrical but complementary in meeting internal needs or client demands. This requires the careful assessment of both the readiness of the employee and technology in relation to the expected demands of the service process (employee-technology fit appraisal). As detailed earlier, it also requires managers to actively compensate for readiness shortcomings by adopting one of the three enablement strategies: a) matching employees and technology with complementary readiness, b) pre-assigning tasks or subprocesses to either party based on respective readiness profiles, or c) strategically investing in the development of specific readiness aspects of their employees or technology.

Furthermore, managers need to ensure that employees and technology also recognize these complementarities for specific tasks. To this end, our model outlines three key managerial responsibilities: a) enabling employees to assess how well their own readiness aligns with task demand (employee-fit appraisal), b) ensuring employees can evaluate the extent to which technology meets those demands (technology-fit appraisal), and c) facilitating comparative-fit appraisals to determine whether the employee or technology is better suited for the task. To enable HUMTECH synergies, managers should therefore foster metaknowledge for each of these fit appraisals. This can include increasing employees’ a) technology literacy, b) reducing overconfidence in handling unsuitable tasks, and c) improving their comparative fit judgments, as well as d) enhancing technology’s predictive accuracy. Together, these enablement strategies ensure that tasks are assigned to the “more ready” party, without which HUMTECH synergies are unlikely to emerge due to inefficient task allocations.

Finally, managers need to ensure that recognized complementarities between employees and technology are also effectively enacted to realize HUMTECH synergies. To enable HUMTECH synergies, managers should therefore optimize organizational processes to facilitate effective human-technology collaboration. Specifically, our model highlights four key enablement strategies: First, managers must establish clear delegation mechanisms that facilitate the transfer of rights and responsibilities for a task to the “more ready” party. Second, managers need to foster trust in delegation decisions to drive acceptance by both employees and technology and ensure that tasks are effectively executed. Third, managers need to implement effective coordination mechanisms to align the timing and sequencing of tasks along the service process. Finally, managers need to incorporate continuous task evaluation procedures along the service process to monitor progress, identify deviations from predefined goals, and adapt accordingly. Failure to enable any of these processes is likely to cause service breakdowns and prevent HUMTECH synergies from emerging.

Our model highlights the importance of readiness and metaknowledge for the emergence of HUMTECH synergies and offers actionable enablement strategies through which firms can foster and leverage these capacities. However, it does not explore the broader organizational systems that create the conditions for readiness and metaknowledge to develop over time. As such, we encourage future research to investigate how organizational systems, such as high-performance work systems (HPWS) and sociotechnical systems can shape the emergence and sustainability of these capacities. For instance, HPWS research indicates that complementary human resource management (HRM) practices such as selection, training, performance appraisal, compensation, and job-design can help organizations develop workforces that have the ability, motivation, and opportunity to drive organizational performance (Appelbaum et al., 2000, Jiang et al., 2012, Subramony, 2009). Similarly, sociotechnical systems research (Appelbaum, 1997, Cooper and Foster, 1971) emphasizes the importance of increasing the fit between employees and the technology with which they interact. For instance, technology that disrupts the social system (e.g., by reducing interactions between coworkers) is less likely to be adopted; and process-standardization also needs to consider employees’ social needs and the capabilities of the technology. In sum, both streams offer promising theoretical venues to explore the organizational conditions for readiness and metaknowledge to emerge.

Second, our paper focuses on the interaction between employees and technology but does not explore the role of customers or the organizational context in shaping HUMTECH synergies. However, customers also require a certain readiness level to effectively collaborate with employees and/or technology in co-producing the service (Bowen, 2024). While HUMTECH synergies may still emerge, customers with low readiness levels will be less likely to balance out readiness shortcomings of both employees and technology throughout the service process, thus leading to suboptimal customer outcomes (Danatzis et al., 2022). Similarly, contextual factors, such as organizational culture, management practices, or market pressures may equally impact collaboration success. Future work should thus focus on how customer readiness and contextual factors shape the emergence of HUMTECH synergies.

Finally, we recommend empirical research to test our framework in organizational settings. Going beyond the discussion of single cases, empirical studies are needed to measure HUMTECH synergies and its components (i.e., employee and technological readiness, metaknowledge) and explore their impact on important organizational, employee, and customer outcomes. Overall, our proposed HUMTECH synergies model provides the necessary theoretical scaffolding for service scholars to advance further conceptual and empirical research on human-technology collaboration at the organizational frontlines.

Ilias Danatzis: Writing – review & editing, Writing – original draft, Supervision, Conceptualization. Joy M. Field: Writing – review & editing, Writing – original draft, Supervision, Conceptualization. Mahesh Subramony: Writing – review & editing, Writing – original draft, Supervision, Conceptualization.

1. Appelbaum et al., 2000
2. Bhattacherjee et al., 2017European Journal of Information Systems, 27 (4) (2017), pp. 395-414
3. Gass, 2023Gass, D. F. (2023). Exploring personality-based heterogeneity in metaknowledge and human-AI collaboration. International Conference on Information Systems, ICIS 2023.

Ilias Danatzis is an Associate Professor of Marketing Analytics at King’s College London, UK. His main research interests revolve around services marketing, with an emphasis on (dys-)functional service encounters, customer and employee readiness, digital platforms, innovation, and value cocreation in service ecosystems. His award-winning research has been published in leading academic outlets such as the Journal of Marketing, MIS Quarterly, Journal of Service Research, International Journal of Research in Marketing, and the Journal of Business Research.

Joy M. Field is an Associate Professor of Operations Management in the Carroll School of Management at Boston College. Her recent research focuses on designing and managing service processes for improved efficiency and effectiveness, with an emphasis on the role of the customer co-producer. She has published her work in a number of leading academic journals, and is the author of the book, “Designing Service Processes to Unlock Value,” now in its 4th edition.

Mahesh Subramony is a Professor of Management at Northern Illinois University in the USA. His research examines frontline service employees and employment with a focus on the changing nature of service work during disruptive times. Mahesh has served as Associate Editor for Journal of Business Research, Journal of Service Management, and the Journal of Service Research.

Service Research in Business-to-Business Marketing