AI in sales: Laying the foundations for future research

Abstract

Artificial intelligence (AI) tools have seen widespread adoption in the sales function. However, the pace of adoption means that sales researchers are often several steps behind the business world. A way to alleviate some of these concerns is to provide a practical, yet up-to-date understanding of AI in sales. Thus, the purpose of this manuscript is to clarify what AI is, the role of AI in sales, and its implications for multiple stakeholders in the sales organization (i.e. salespeople, sales managers, organizations, and customers). To achieve this objective, the authors provide an up-to-date examination of business practices that are interwoven with practitioner and academic literature. Next, based upon exploratory interviews with eighteen practitioners, the authors discuss challenges and opportunities that can arise due to AI adoption. Finally, the authors provide directions for future exploration of AI in the sales domain.

Keywords:

• Artificial intelligence
• sales process
• sales function
• sales managers
• organizations
• customers

Introduction

Artificial Intelligence (AI) or “a system’s ability to correctly interpret external data, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation” is evolving rapidly (Kaplan and Haenlein 2010, 17). AI has infiltrated all aspects of business, and scholars are beginning to examine how AI will alter business functions, including marketing, service, and information systems (see Collins et al. 2021; Davenport et al. 2020; Enholm et al. 2022; Huang and Rust 2021a). In a business to business (B2B) sales context, Syam and Sharma (2018) suggest that AI will initiate a “sales renaissance.” Today, this sentiment appears correct, as AI technologies are helping to augment every phase of the sales process, especially as it relates to complex B2B sales (Paschen, Wilson, and Ferreira 2020; Rodriguez and Peterson 2024). In fact, according to Gartner (2023), AI is predicted to fulfill 60% of sales tasks by 2028, a stunning increase of 55% from 2023.

Practitioner literature has highlighted applications of AI and the impact on various stakeholders, as shown in Table 1. Accordingly, sales research has begun to consider the broad impact of AI on different stakeholders (Fischer et al. 2022; Paschen, Wilson, and Ferreira 2020; Singh et al. 2019; Syam and Sharma 2018). At the salesperson level, scholars consider how salespeople will respond to feedback (Hall et al. 2022), or how AI will impact customer relationships (Chang 2022). At the sales manager level, AI has been used for coaching purposes and determining how to drive adoption of AI tools (Chen and Zhou 2022; Luo et al. 2021). At the organizational level, AI helps to generate accurate sales forecasts, manage relationships with customers and formulate strategies (Huang and Rust 2021b; Libai et al. 2020; Sohrabpour et al. 2021). At the same time, customers’ perception of AI as sales agents has also been explored (Adam, Roethke, and Benlian 2023; Fiestas Lopez Guido et al. 2024).

Table 1. Practitioner perspective: AI applications and implications.

Steps in the sales process	Types of AI in use	Impact on salesperson	Impact on sales manager	Impact on organization	Impact on customer	Examples of AI tools
Prospecting	Automated email messaging Automated calling Lead scoring Predictive analytics White space identification	Reduced workload Increased focus Increase in analytical skills Uncertainty about accuracy of algorithms	Time management of salespeople Resource allocation Coaching of salespeople to use AI	Lead generation capabilities Upskilling sales managers Content creation Market intelligence capabilities	Adequate information to identify right needs Focused attention from suppliers	Gong helps salespeople to “score” and “rate” their leads and prospects. A benefit of this is that it can remove much of the guesswork and subjectivity that was involved in the process. In addition, Gong also supervises deals and interactions with salespeople to help them understand the customer and where issues may arise. SalesIntel offers AI-generated insights into the customer research process, improving salespeople’s intelligence gathering.
Pre-Approach	Market intelligence Intent data Social listening	Better targeting Reduced workload Increased efficiency	Increased coaching KPI setting Performance management	Coaching of sales managers Sales manager KPI setting Performance management	Adequate information to identify right needs Focused attention from suppliers	LinkedIn Sales Navigator, powered by generative AI can help salespeople to search and curate prospects. It can provide salespeople with ideas on how to open a dialogue.
Approach	Conversation analytics Call transcriptions Sentiment analysis Sales enablement	Improved call effectiveness Self-coaching Peer coaching Analytical skills Accuracy of reports Pushback from customers Fear of incompetence being exposed	Reduced manager workload Individual coaching Team coaching Salesperson training Performance management	Ethical governance concerning use of customer and salesperson data	Effective needs assessment Increased efficiency	ActiveCampaign uses generative AI to automate contact with customers. Eloqa (an Oracle company) helps salespeople to engage with and develop leads at various stages of the buying process. Eloqa uses generative AI to help salespeople target and approach customers in the most efficacious way possible. Generative AI allows Eloqa to understand novel insights (i.e. email fatigue) levels about potential and current customers to time email messages.
Presentation	Sales enablement Content analytics Automated proposal generators Automated simulations Dynamic pricing	Quality sales interactions Increased sales efficiency Shortened sales cycle times	Activity based KPI setting Coaching based on AI and tool usage	Enabling content creation and analytics across various silos Activity based KPI setting for salespeople Joint KPI setting for marketing and sales	Personalized information per stage in the buying process	Leexi summarizes sales meetings, identifies salespeople’s questions, and the customer’s objections. SetSail records, tracks, and rates salespeople’s activities.
Overcoming Objections	Conversation analytics Call transcriptions Sentiment analysis Sales enablement	Increased hit rates Shortened sales cycle times	Salesperson coaching Sales team coaching	Clearer and better understanding of the customer	Risk mitigation	Closem.AI helps to send follow-up emails with responses to objections. Dasha.AI helps provide real time generative responses to customer objections that are adaptive and based upon the characteristics of the interaction.
Close	Conversation analytics Call transcriptions Sentiment analysis Sales enablement Automated scanning of legal documents	Reduced errors Win analytics	Salesperson coaching Sales team coaching	Investment in new capabilities	Customer education Value promise	Salesforce.com offers an AI integrated CRM to analyze customer interactions/conversations/presentations. Showpad with PitchAI offers real-time feedback on salespeople’s verbal and nonverbal communication.
Follow-Up	Customer onboarding Customer engagement	Increased customer satisfaction Increased repurchase intent	Team viability Team performance	Investment in new capabilities	Customer education Value in Use Increased effectiveness Increased efficiency	Sentiment analysis from Amazon Web Services (AWS) allows for organizations to pick up on various aspects of customer sentiment in different channels (i.e. social media, conversations, emails, texts etc.) to understand and predict what customers are “saying”. Docebo, an AI-powered LMS, is used by companies worldwide for education and training. Sciolytix has a virtual role-play system that finds, screens, and utilizes AI-driven role-plays for pre- and post-sales hires.

As AI impacts various stakeholders, a general overview of the challenges and opportunities AI poses to the sales function is necessary. There are still valid questions about what AI means in practice as aspects of AI remain a “black box” (Rai 2020). In addition, two of the most influential conceptual sales articles (Singh et al. 2019; Syam and Sharma 2018) were written prior to the rise of Chat-GPT and other advanced generative AI technologies. Singh et al. (2019) discuss AI and its potential to the sales function, but also devote substantial attention to other technological aspects (i.e. digitization, digitalization, and digital transformation). Syam and Sharma (2018) emphasized on machine learning (ML), which is a subset of AI. Given that both focus more on the potential of AI in sales and substantial time has passed since their publication, it is prudent to provide an up-to-date understanding of what AI means in sales. Because of the general uncertainty associated with AI and given the pace at which AI is evolving, we have devised the following questions: (1) what is the role of AI in sales, (2) how is AI currently being used in the sales function, and (3) what are the opportunities and challenges associated with the impact of AI on the sales function (i.e. salespeople, sales managers, and the sales organization) and the customer?

To help achieve a better understanding of how AI is impacting the sales function, we first examine evidence from the practitioner domain. We then review the extant literature in the sales domain to provide insights into how AI is impacting the sales function. Specifically, we draw on existing academic frameworks to look at how AI is affecting salespeople (Dubinsky 1981), sales managers (Zoltners, Sinha, and Lorimer 2008), the sales organization (Amit and Schoemaker 1993; Freeman 2010; Porter 1991; Weber et al. 2023), and the customer (Allal-Chérif, Simón-Moya, and Cuenca Ballester 2021; Lemon and Verhoef 2016). Given the nascency of the literature stream in AI, we utilized the research method employed by Andzulis, Panagopoulos, and Rapp (2012) in their pioneering work on the impact of social media on salespeople. We provide insights from 18 conversations with practitioners about their experiences with AI and their insights about the future of the technology in sales. We also integrate practitioner and academic literature to provide insights into AI in the sales world.

With a combination of key elements from the different frameworks, a review of the academic/practitioner literature, and finally the observations from the practitioners we make three contributions to the sales literature. Specifically, we discuss that (1) AI is not a single entity but is a constantly evolving technology that has the potential to both automate and transform the tasks and processes being performed by different sales function stakeholders, (2) we highlight unique opportunities and challenges that stem from AI and (3) we identify a set of research questions that could provide inspiration to sales, marketing and other business academics while also informing practitioners about how to better exploit the different forms of AI while avoiding pitfalls.

Overview of digital technologies in sales

Professional selling has undergone sweeping changes in recent years. Customer demands are increasing (Bonney, Beeler, and Chaker 2022; Epler and Leach 2021; Paesbrugghe et al. 2020) and there is a push for digital transformation efforts by sales organizations (Fischer, Seidensticker, and Poeppelbuss 2023). Specifically, sales organizations have responded to contemporary challenges by enabling their salespeople (Rangarajan et al. 2020) with a set of tools referred to now as the tech stack (Agnihotri et al. 2023; Bauer et al. 2023). The sales technologies contained within the tech stack have helped salespeople to gather enormous amounts of information and data about their prospects/customers and the effectiveness of their marketing/sales efforts to win customers, customer interactions (during and after sale). Criticality, the tech stack also as well as information about their customer engagement across social media platforms. In turn, the organization also receives benefits from the technologies.

Owing to faster data processing speeds, the use of ML algorithms helps sales organizations to better utilize the data gathered from the different technologies or the tech stack referred to above. Naturally, this has been the focus of attention in the sales domain (Paschen, Wilson, and Ferreira 2020; Singh et al. 2019; Syam and Sharma 2018). Advances in Natural Language Processing Models (NLP) combined with ML techniques can help sales organizations harness predictive analytics to improve the effectiveness of their sales force (Paschen, Wilson, and Ferreira 2020; Singh et al. 2019). Finally, with the introduction of ChatGPT in late 2022, generative AI, defined by García-Peñalvo and Vázquez-Ingelmo (2023, 8) as “a production of previously unseen synthetic content, in any form and to support any task, through generative modeling” has received substantial media attention. Ultimately, large language models (LLMs) have made it possible for salespeople to leverage their tech stack more fully and have actions prescribed to them - mimicking a coworker or manager.

Despite the clear evolution from algorithm-based descriptive analytics to the more advanced prescriptive analytics driven systems (Habel, Alavi, and Heinitz 2023a), AI applications in the sales function are still in their early stages (see Paschen, Wilson, and Ferreira 2020). While the work by Paschen et al. (2020) provided a useful first step to understand the nuances of AI, more work is needed to help academics and practitioners differentiate between different types of intelligence, alongside their impact on the sales function (salespeople, sales managers, the sales organization, and customers).

What AI is and what AI is not

AI is a complex and transformative set of technologies, but there are many misconceptions as to what AI is, as well as what it can/cannot do. There are also distinct limitations to the technology and aspects of sales that complicate the issue of AI usage. To unpack these complexities, we refer to the definition of AI from Agnihotri et al. (2023, 37) as “technology with human-like thinking capabilities, such as replicating the problem-solving process, making predictions of various outcomes, suggesting alternative strategies, and offering advice to a sales professional.” In addition, AI is typically implemented with complex ML algorithms (Ma and Sun 2020). It is important to differentiate AI from related technology. It is common to conflate ML or NLP with AI, but both are distinct (Bawack, Wamba, and Carillo 2021). ML is considered a “sub-field of AI that gives computers the ability to learn without explicitly being programmed” (Brown 2021). A strength of any ML technique is that it can handle various types of data, such as unstructured (i.e. video or text) or structured, complex network data. ML is often used to target customers with personalization efforts, segment and focus on aspects of the customer journey (Ma and Sun 2020), which provides a unique opportunity for the potential of ML in sales (Glackin and Adivar 2023; Nguyen et al. 2023; Syam and Sharma 2018).

NLP refers to the ability of a computer to comprehend human language. NLP utilizes algorithms to transform unstructured data into computer language and this allows for various chatbots or forms of customer assistance (Olujimi and Ade-Ibijola 2023). In essence, the algorithms can be trained to recognize customer sentiment, intent, and answer customer queries (Beeler, Zablah, and Rapp 2022; Olujimi and Ade-Ibijola 2023). As an example, IBM’s Watson allows for companies to have an additional 24/7/365 worker on the “frontlines” with chatbots and digital assistants. NLP’s have a clear relevance in marketing and recent research provides technical and empirical guidance into current applications (see Shankar and Parsana 2022). In terms of sales, Paschen, Wilson, and Ferreira (2020) discuss the possibility of NLP for salespeople in detail. However, for accuracy’s sake, it is important to refer to each technology in an accurate manner (see Siegel 2023), and in this sense, ML and NLP are essentially the “output” of AI.

Similarly, sales force automation (SFA) and customer relationship management (CRM) were implemented widely to reduce task monotony, build relationships, and monitor salespeople (see Hunter and Perreault 2007; Rapp, Agnihotri, and Forbes 2008; Singh et al. 2019). While AI shares an overlap with these technologies, the key distinction is that AI is being implemented into both sets of tools to drive their efficacy (Chatterjee et al. 2021). Figure 1 provides an overview and definition of different AI tools and applications.

Figure 1. AI in sales.

Within these parameters it is evident that AI has many facets, however, the limitations of AI relate to when its abilities/potentials are misconstrued. Thus, it is important for sales scholars to be realistic about AI and what it can do. Kozinets and Gretzel (2021) caution against overstating the benefits or promises of AI. For example, AI provides the opportunity to generate insights and provide efficiency but overreliance on technology and taking an entirely mechanical approach to customers can mean that the important nuances are missed. In a similar vein, AI derived content (i.e. coaching, training, feedback, or analysis) lacks a human approach and may not be well received (Luo et al. 2021).

Next, AI is certainly powerful and its impacts on the labor market have been the subject of speculation and debate. To illustrate, scholars have measured occupational exposure or suitability for AI (see Brynjolfsson, Rock, and Syverson 2018; Felten, Raj, and Seamans 2021, 2023). Not surprisingly, certain aspects of the sales function, especially roles that can be automated (inside sales/telemarketing) are considered “exposed” to threats of AI. However, economic researchers (Acemoglu et al. 2022, S337) have found that while AI is having impacts on the labor market, aggregate effects are “not yet detectable”. Ultimately, a more realistic view suggests that AI is not soon to “take over” all business functions and supplant the various stakeholders involved in selling (Brock and Von Wangenheim 2019; Pappas et al. 2023).

Evolution of AI and application in sales

Several factors have coalesced to make the current evolution of AI possible. The availability of vast amounts of data, storage cost reductions and a rapid increase in data processing speeds have all played a role. Therefore, the first evolution of AI was in the domain of data analytics. As Davenport and Fitts (2021) explain, AI has transitioned rapidly from (a) being used for providing descriptive analytics based on historical data to aid business decision makers, (b) to predictive analytics which involved using the historical data and applying ML algorithms to develop predictive models that could be used to forecast possible outcomes/trends, and finally (c) prescriptive analytics where AI uses optimization and simulation techniques to not only predict outcomes but also provide actionable recommendations to users. Adding to this, conversation analytics was developed to analyze spoken or written conversation (verbal or sentiments), allowing a deeper understanding of customers (Habel, Alavi, and Heinitz 2023a).

Additional research proposes multiple types of AI. For example, the “multiple type” view (see Huang, Rust, and Maksimovic 2019; Huang and Rust 2018) suggests that mechanical AI learns and adapts in a minimal way and is well-suited for tasks that are easily automated and fit for standardization. Moving a step further, thinking AI is proficient at learning and adapting from customer data and engaging in personalization efforts. The most advanced type, feeling AI can learn and adapt from experiences and allow the AI to behave in a more relational manner. Furthermore, feeling types of AI are the least developed; yet, have the most potential to be transformative in a marketing context (Huang and Rust 2023). To summarize, this viewpoint proposes that the different types of AI have a specific function in understanding, targeting and relating to the customer.

In terms of current practices, a recent Salesforce.com survey (2023) indicates that sales organizations are currently using AI for data analytics, automation and content creation. AI is bolstering data analytics and helping practitioners to understand their data and customers in a more comprehensive manner. Generative AI also allows for the creation of new content (e.g. targeted messaging) (Stokel-Walker and Van Noorden 2023). Combining those functionalities, AI is being increasingly predicted to act as a “co-worker” to salespeople (Davenport et al. 2020; Dickie et al. 2022). We now turn our attention to how analytics, automation and content creation are evolving to make this a reality.

Data analytics in sales

With access to relevant data and ML algorithms, AI can be used to predict business trends and customer behavior, spot opportunities within existing customers, and make revenue forecasts (Habel, Alavi, and Heinitz 2023a, 2023b). These predictions can be used by sales managers to identify high-potential leads and opportunities, allocate resources effectively across leads and opportunities, and develop effective “go to market” strategies. AI insights can also be used to identify underperforming market segments, customers, and even salespeople, allowing sales managers to take proactive remedial actions.

AI-powered algorithms can analyze historical sales performance data compared to market trends and external factors to generate more accurate insights into the performance of their sales team (Guenzi and Habel 2020). This helps them to set more realistic targets, while ensuring proper resource allocation. Tools like conversation analytics can help sales managers experiment with different types of messaging with their sales team members, resulting in increased sales force effectiveness (Luo et al. 2021).

Automation in sales

The use of algorithms to analyze large volumes of data to identify and prioritize leads has been used by B2B marketers and sales as part of their demand generation activities. Similarly, AI can facilitate the analysis of current market trends that are coupled with historical data to improve forecast accuracy. The use of predictive analytics to identify upselling and cross-selling opportunities, churn prediction, while also aiding in inventory management is another way in which AI is contributing to sales automation (Habel, Alavi, and Heinitz 2023b). The use of chatbots and virtual assistants combined with predictive analytics tools are being used to handle simple customer inquiries and schedule appointments (Ramesh and Chawla 2022). In practice, this frees up the time of salespeople to focus on more critical, value-adding activities.

The practitioner literature indicates that AI can also be used by sales organizations by targeting email campaigns to prospects based on a historical analysis of customer data and behavior (Bicancic et al. 2023; Dixon and McKenna 2023). Interestingly, AI can even provide preferred content and properly timed emails. AI can scan the market for information on trends, competitive prices, and the customer’s market situation to inform dynamic pricing optimization allowing the salesperson to be more effective in their role. Thus, by helping automate parts of the sales process with ML algorithms and historical data, AI is increasingly being used to drive the efficiency and effectiveness of salespeople (Rodriguez and Peterson 2024). For example, by helping to automate routine processes like administrative tasks, appointment scheduling, follow up emails with salespeople, customers, and internal departments, AI can help managers save time and energy which they can put to better use to help their sales teams. For instance, sales managers can reallocate their time to coach their salespeople with difficult customers, build effective customer relationships, or shorten sales cycle times that could result in improved sales team satisfaction and performance.

Content creation in sales

Generative forms of AI are at the forefront of the current media attention because of their ability to provide natural sounding answers to questions (Kshetri et al. 2023). With the advent of LLMs, sales organizations can now use input from customer interactions to combine with historical data to produce content. This content can include blogs, white papers, case studies, and product descriptions. AI can now be used to generate personalized emails based on customer interactions and adapt the content based on the personality type of the customer to ensure better reach and email open rates (Deveau, Griffin, and Reis 2023).

Aside from providing market trends and competitive analysis, AI can also help shed light on the customer (Narayandas and Sengupta 2023). Specifically, sentiment analysis enables salespeople to better understand and target their customers. B2B organizations that have AI enabled websites can provide relevant content to visitors based on their browsing behaviors (as well as based on past visits). Voice-activated search and assistance combined with translation and geo-localization options can not only help customers but also salespeople for companies operating in global contexts. Furthermore, use of LLMs in sales playbooks as well as other sales enablement efforts highlights ever evolving benefits of AI to the sales function.

Generative AI is also set to play a vital role in helping sales managers become more effective and efficient in their jobs (Chen and Zhou 2022). By automating routine tasks, providing predictive insights, and helping sales managers make better decisions, AI is making it easier for sales managers to spend more quality time leading their sales teams. AI can help sales managers optimize their team composition, help them in resource allocation across different customers, different sales roles, and across different sales activities - thereby enhancing sales team performance. In the next section, we highlight the opportunities and challenges that AI poses for various stakeholders associated with the sales function.

Implications of AI for stakeholders: opportunities and challenges

AI has spread rapidly into various aspects of the organization and is reshaping aspects of the customer interface (see Ahearne and Rapp 2010; Mick and Fournier 1998). Thus, it is imperative to discuss the implications of AI at different levels: salesperson, sales manager, organization, and customer. To gain a better understanding of the implications and applications of AI in sales for each level, we conducted eighteen informal interviews with sales professionals involved in various sales roles and levels within their firms (e.g. Andzulis, Panagopoulos, and Rapp 2012; Marshall et al. 2012; Rangarajan et al. 2020). As noted in Web Appendix A, the respondents worked in various industries such as technology, cloud, and manufacturing and had an average of 14 years of experience in sales. The interviews ranged from thirty to sixty minutes and were conducted via Zoom or other video conferencing technology. The interview guide is in Web Appendix B. Like previous research (e.g. Andzulis, Panagopoulos, and Rapp 2012; Hartmann and Lussier 2020; Rangarajan et al. 2020), the interviews were designed to be exploratory to gain insights about the current applications of AI in their organizations, the tradeoffs associated with using the technology, and the future of AI in sales for various stakeholders.

Salesperson

The traditional Dubinsky (1981) B2B sales process has been discussed for over forty years in the sales literature. This framework describes seven steps (i.e. prospecting, pre-approach, approach, presentation, overcoming objections, close and follow-up). Given the potential impacts of AI on every aspect of the selling world, we highlight the impact of AI as it relates to current and future applications throughout the sales process in Table 2.

Table 2. Salespeople and AI: implications and applications.

	Types of AI applications for salespeople			Implications
	Machine Learning	Natural Language Processing	Generative AI	Opportunities	Challenges
Prospecting	Identify potential prospects utilizing demographic information Large-scale knowledge gathering and lead scoring Demand prediction	Respond to prospect query information Predict potential customers Assessing risks and opportunities Segmentation; Competitive analysis; Search engine optimization (SEO); Content Marketing; generate awareness	Blogs, social media posts Create unique marketing and sales content Develop specific messaging based on prospect information Develop targeted sales scripts	Reduced workload for salespeople More qualified leads are identified Reduced stress for salespeople Improved productivity	Could reduce time spent with current customers Data security and privacy issues Increased stress for counterparties or other organizational members
Pre-Approach	Examine large datasets for novel customer/marketplace insights Text analytics for qualification	Monitor social media or text mine other types of communication Predict purchase likelihood	Synthesize important customer information for salesperson/sales teams Brainstorming ideas and creative approaches	More sophisticated understanding of customer	Analytical skills become more important Potential loss of tacit knowledge Individual customer variance
Approach	Personalized links and information Task automation	Analyze email responses to set up meetings Chatbot Customize the pre-trained data to the specific customer	“Cold Calling” Email outreach Send the right message at the “right time”	Task delegation	May lose out on rapport development Decreased responsibility/ ownership for salespeople
Presentation	Opportunity scoring Real time optimization Understand the specific aspect of the customer journey	Record and analyze the conversation in real-time Deliver value propositions Rapport development	Create relevant product/service descriptions Dynamic messaging capabilities Personalized content (e.g., custom images)	Streamlined and efficient process of developing sales presentations	Customer may have questions that only a human could answer
Overcoming Objections	Voice analysis Facial recognition Analyzing previous objections from interactions to help tailor a response	Examine customer grammar and granular sales interaction details Provide natural sounding responses	Real-time responses Customized responses to objections/issues Understand the customer persona	Help salespeople better understand what the customer is saying	AI may not “hear” the actual objection
Closing	Recommendations on how to best close a customer Purchase prediction Dynamic pricing	Monitor call and transcript to provide closing suggestions Sentiment analysis Accurate responses to customer issues	Observe the closing process and guide both parties Provide recommendations Streamline the negotiations process	Help salespeople to optimize the “close” Start relationship on the right foot	Customers may feel uneasy that a computer is analyzing their words
Follow Up	Customer Lifetime Value prediction Analyze CRM/Purchase data for targeted outreach Churn prediction	Help with CRM input Chatbots can help to personalize post-sales service and route the appropriate help for issues	Personalized training content for customers/general “care” provisions Coach salespeople and other service personnel with immediate feedback/guidance Content generation and follow up with digital voice assistants	A more holistic view of post-sales service	A mechanical process will preclude anything that is not captured in the data

Note: ◆ bullet points indicate applications that are well established; ○ bullet points indicate applications that are not well established.

According to many of our participants and prior literature, AI-usage is more commonplace on the prospecting step of the sales process. AI has been shown to produce valuable leads (Deveau, Griffin, and Reis 2023), as the technologies are designed to learn on existing data sets (Davenport et al. 2020), but what constitutes a lead may contain contextual details that are specific to an organization or an industry. Instead of searching for prospective customers that may or may not buy, salespeople can use the insights from AI to contact those that are similar to their current clients and those that present a higher likelihood of becoming a customer.

• “Sales reps will be less focused on activities such as prospecting and cold calling.” (George, CEO)

• “AI is clearly helpful for lead generation because it can collect data for an ideal customer profile.” (Benjamin, CEO)

• “AI will be a huge part of lead generation moving forward.” (James, Sales Consultant)

• “Salespeople and sales organizations differ on what leads actually are. So the question is, what is a lead and this will differ vastly for each company.” (Daniel, Sales Consultant)

As the steps of the sales process continue (e.g. the pre-approach, approach, presentation, objection handling, closing, and follow up), AI tools can help to identify which customers to reach out to, recommend what they may need, and possibly follow up with the customer, thus making the salesperson more effective. Moving a step forward, generative AI could even help salespeople and their organizations determine whether a digital or face-to-face meeting is more appropriate (Kshetri et al. 2023). The use of generative AI can provide salespeople with content that is curated to individual customers along with recommendations of approaches to handle objections or present information that resonates with customers (Huang and Rust 2021a). Generative AI can provide recommendations of products/services for the salesperson to upsell or cross sell for different customers, but the salesperson must use their own judgment to determine whether the solution offered is the best fit for their customers. For salespeople, showing customers how they provide value that is “over and above” the AI technology will be imperative, especially during the latter stages of the sales process.

• “You can utilize AI to analyze the trajectory of a relationship. Did the customer agree to a demo or trial process?” (Cory, Head of Growth)

• “AI can be used to help prepare for sales calls and meetings and create programmatic playbooks. It helps the salesperson prepare for the conversation that they want to have.” (Louis, Account Manager)

• “AI will allow for personalization at scale.” (George, CEO)

However, an important distinction is the differences between transactional contexts with shorter sales cycles and in B2B with (typically) deeper engagement, more complexity, and a longer sales cycle. There are certainly exceptions as a participant indicates below, but at this juncture, the current applications for AI appear to be more commonplace in areas where tasks are easily automated, or the interactions are simpler. Importantly, because AI can “learn” (Davenport and Fitts 2021), it is likely that AI will eventually be integrated into all types of selling from simple transactions and short sales cycles to complex offerings with protracted and complicated sales cycles (Dickie et al. 2022).

• “Transactional sales can be done entirely with chatbots.” (Kevin, CEO)

• “There are billion-dollar deals being done with AI. IBM is selling to other large companies (with AI).” (Cory, Head of Growth)

• “No AI [should be used] for large deals, [but] for fast and dirty deals, AI can be very useful.” (James, Sales Consultant).

Much of the uncertainty about AI and the salesperson stems from the notion of whether AI will replace the salesperson (Huang and Rust 2018). Several respondents noted that many salespeople today struggle to showcase value to the customers and struggle to identify solutions that meet the complex needs of their customers. AI may outpace or outperform salespeople, especially if the salesperson is not taking advantage of the vast amount of data that is available to aid their decision making. If this trend continues, customers may prefer to interact with AI instead of salespeople.

• “For the approach AI and a sales agent can have about the same level of proficiency however, in the long run AI may be better at providing the solution.” (Frank, VP Analytics)

• “Salespeople can lack the ability to listen; they tend to talk more and listen less. AI can do a better job of listening to customer needs and wants than salespeople.” (James, Sales Consultant)

• “Customers have changed, 82% do not feel that the salesperson is prepared. 50% of sellers themselves do not feel prepared. [I've observed] that sellers spend hours a week in follow-up emails. AI and tech can make this more efficient and this matters because many companies do not want to see salespeople anymore. They are not providing as much value anymore.” (Peter, Division VP)

• “A good AI (tool) can identify call intent as well as a salesperson. AI may even be better when providing solutions. Human recommendations [alone] may be based on limited information vs. an algorithm or AI that can access more information and then propose an optimized solution.” (Frank, VP Analytics)

For now, it is important for salespeople to understand that AI, like other technologies, is an enabling tool and not a complete solution. This would imply that salespeople need to understand how AI tools can help them manage their daily routines and tasks. While some AI tools can help automate certain mundane sales tasks like leads/opportunity prioritization, handling routine inquiries, scheduling meetings with customers (Dickie et al. 2022), more advanced AI tools can help augment tasks performed by salespeople like personalized communication with customers (Huang and Rust 2021a), providing real time customer insights/intelligence, dynamic pricing, self-coaching, and predicting future sales (Habel, Alavi, and Heinitz 2023b). Furthermore, different types of AI-driven sales dashboards are helping to provide additional clarity to different aspects of the sales process. Yet, the potential does not imply that AI will be a panacea, as salespeople will still need to work hard to be effective, and this is especially true with the added transparency.

• “The sales process and methodology will still be followed but everything is more complicated now….AI will make the sales funnel (process) much more transparent.” (Adam, VP)

• “All of the training methods (e.g., BANT, Challenger, SPIN) are still relevant. However, AI becomes a check on salesperson skill.” (Adam, VP)

Throughout the entirety of the sales process, rejection and failure are common (Bolander et al. 2017; Nguyen et al. 2023). Yet, the fear and stress that comes along with missing quota or failing to win a bid, may be lessened by the utilization of AI technologies. Which in turn, decreases the number of times a salesperson hears “no” and must handle rejection. Participants indicated that AI can likely help them to meet qualified prospects and have the potential to reduce some of the negative aspects of being a salesperson (e.g. deals falling through, anxiety about closing, quota attainment, account prioritization) and therefore, mitigate mental health concerns (Lussier et al. 2023). Ultimately, AI has the potential to reduce rejection-laden, repetitive, or time-consuming tasks, such as data entry, outreach and even sending follow up emails (Dickie et al. 2022; Porter 2017).

• “If AI is able to provide “warm” leads to salespeople, there will be decreased attrition in the sales role. The job becomes less stressful as the AI takes care of the “no’s” in the beginning for the salesperson.” (Elizabeth, Sales Consultant)

• “What is the likelihood of a deal closing? It essentially (generative AI) will be a copilot that is complementary to the sales team/person and can help to get the job done (deal closed).” (Louis, Account Manager)

AI can analyze large amounts of data much faster than humans, but the output still requires interpretation, editing, and dissemination. Salespeople need to engage with these tools consistently to become more familiar with the AI, and to also help “train the AI tool(s)” by providing feedback. Salespeople will also need to work on exposing the AI tools to various customer-facing scenarios to collect critical information. Salespeople may have to correct errors, advise on nuances, and oversee the output to have it work most effectively. For example, with the implementation of AI, the salesperson must adapt to things that the tool cannot do such as, “managing exceptions, tolerating ambiguity, using judgment, shaping the strategies and questions that [AI] machines will help enable and answer, and managing an increasingly complex web of relationships with employees, vendors, partners, and customers” (Baumgartner, Hatami, and Valdivieso 2016). The complexities and nuances that AI cannot consider will need to be complemented by a salesperson’s experience and knowledge. Ultimately, at this juncture, the basket of technologies is far from perfect and for the foreseeable future, will involve training and retraining.

• “There may be missing pieces or things that [AI technology] does not take into account. Intuition, experience, and knowledge are irreplaceable.” (Peter, Division VP)

• “[Salespeople should] be the last mile, and let the technology do the heavy lifting and research and then humanize it.” (George, CEO)

• “When you are an expert and look at the [AI outputs], you can take the material and make it better than the tool can create.” (Isabella, Management Lead)

For salespeople to fully adopt AI into their routines, they will need to see value in it for themselves, otherwise they may be hesitant to buy-in (Kellogg, Sendak, and Balu 2022). Salespeople can be hesitant to change or disrupt how “things are done”. This can insert tension into an organization because there will be different incentives and viewpoints into the adoption of AI-based tools and technology. To illustrate, Pachidi et al. (2021, 20) refer to a “regime of knowing” that “offers socially sanctioned conventions about what counts as valuable knowledge, about what actions are appropriate, and about authority arrangements.” In their study of AI adoption, the authors found that employees would often provide resistance. What this means is that AI implementation is going to require managerial oversight and encouragement.

• “There are a lot of questions around whether or not the salesperson is even using the data [from the AI]. The salesperson may not even use the information or even use their previous knowledge to conduct the sales process.” (Daniel, Sales Consultant)

• “The handoff [of data from AI to the salesperson] is complicated, there needs to be consistency between the chatbots and the sales team. Wrong info means the unselling of your product.” (Matthew, Manager)

Sales managers

Sales managers already play an important role in leading their sales team (Plank et al. 2018). Excitingly, the advancement of AI technology will directly impact how they can effectively drive the performance of their sales teams. Zoltners, Sinha, and Lorimer (2008) and Johnston and Marshall (2020) discuss a variety of sales manager responsibilities including: training and coaching, performance management, account management, and hiring. Sales managers can differ in how they drive efficiency through each of these tasks (Zoltners, Sinha, and Lorimer 2008), but all the tasks have distinct impacts on the sales organization. Drawing from this literature, we discuss the differential impacts and applications of AI for sales managers in the following framework (Table 3) and discussion.

Table 3. Sales managers and AI: implications and applications.

	Types of AI applications for sales managers			Implications
	Machine learning	Natural language processing	Generative AI	Opportunities	Challenges
Training & Coaching	Analyze sales metrics to identify salespeople that may need training Identify top sales performers to act as mentors	Analyze communications and negotiations of sales team members to identify areas of opportunity Identify the success rate of negotiation approaches	Develop personalized coachings and trainings for individual sales team members Recommend additional coachings to sales teams Provide role play scenarios or trainings for sales team	Increased transparency in sales processes/methods for each sales member Reduced training workload Implementation of personalized training to match the unique needs of the sales team members	Additional data may be hard to manage Adoption of AI technology will require ongoing training Some of the skills needed may be difficult to train
Performance Management	Analyze product portfolio sales margins and unit sales Analyze performance trends of sales teams	Summarize sales goals and areas of opportunity for sales team members	Provide feedback throughout sales process on areas to improve	Up to date performance on sales teams Improved productivity Improved metrics from sales teams	Identify optimal points of the customer journey to utilize AI vs salespeople Identify KPIs to match new sales roles
Managing Accounts	Identify opportunities to improve customer satisfaction and retention Analyze current customers purchase behavior for growth opportunities Identify key drivers of customer satisfaction Determine market potential of sales territories Develop forecasting trajectories	Identify customer satisfaction through sentiment analysis Respond to customer inquiries or questions	Recommend products or services to upsell or cross sell to customer Develop customized communication	Improved customer satisfaction Increased efficiency and productivity More accurate understanding of customer needs and buying habits	Inaccuracies in data may lead to incorrect recommendations of products or services Loss of direct interaction with customer
Recruitment & Hiring	Identify potential employees utilizing the specified criterion Predict potential job performance or other measures Score resumes and applications	Respond to questions from applicants Identify qualifications and skills on resumes Conduct job analyses to determine requirements Send job opportunities to qualified individuals through social media or email	Develop recruitment communication based on segmentation information Develop interview questions based on job requirements and necessary skills Draft job descriptions Draft customized messages to candidates	Increased productivity for recruiters and sales managers Improved efficiency in hiring process More qualified applicants are identified	Human bias in systems may exist and lead to inaccurate scoring of resumes and applications Removal of the human element may deter candidates Accurate task text data is necessary for job analysis

Note: ◆ bullet points indicate applications that are well established; ○ bullet points indicate applications that are not well established.

Performance measures and management practices will need to be adapted and changed as the roles of the sales team change. AI will be able to provide sales managers with greater oversight of their salespeople. The technology will allow managers to have access to almost every task done by the sales team to see what is being said/done. This transparency will allow managers to hold their teams more accountable (Guenzi and Habel 2020). As noted by participants, a general idea is that AI will be able to root out inefficiency, scan CRM systems to see if deals are being pulled forward or if salespeople are “sandbagging” deals. Moreover, managers will also have greater insights into what is occurring during the interaction, which has been an ongoing research priority (Arndt et al. 2024; Evans et al. 2012).

• “AI creates transparency within an organization, there is nowhere to hide and people and employers will be held accountable.” (George, CEO)

• “AI will help managers to “read between the lines” with salespeople to uncover issues. This will help sales managers to “weed out” and identify issues that may have been overlooked.” (Adam, VP)

• “Because a manager can record all conversations, they can also tag keywords, so they don’t have to read the whole conversation. For example, they can learn if the call was price focused or product or service focused. [However,] this does mean that predefined tags need to be created.” (Frank, VP Analytics)

• “A company like Gong helps to record sales calls for training. You can ask for feedback on calls and help the salesperson to learn.” (Natalie, BDR)

Sales managers will need to be very clear about how, when, and why AI is to be used throughout the sales process. They will be responsible for continuously training their teams on the AI tools, along with developing new skills that will be needed for their sales team to be successful. AI can analyze data at a rapid pace, but deriving actionable insights from complex data is much more time consuming (Balducci and Marinova 2018), and even more difficult to train. Thus, training for salespeople should prioritize how to manage the personal aspects of relationships along with interpreting the AI outputs to develop solutions (Porter 2017). Indeed, improved training and oversight of salespeople was a common element of our interviews.

• “Another thing to think about is that so many different things can go wrong during the sales process, so when to use the tools is a real question.” (Cory, Head of Growth)

• “Salespeople need to be trained and this is a critical issue. You just cannot expect salespeople to use it effectively and to use it without issue.” (Peter, Division VP)

While AI may be able to replace certain functions, important aspects such as tacit knowledge, thoughtfulness, and other types of extra role behaviors are unlikely to be replaced by technology. However, these skills are more difficult to teach and train, which may mean that managers need to hire individuals who already possess them. Sales managers will need to determine whether a different set of skills or traits are important for their sales team and adjust their hiring processes accordingly. Therefore, along with the changing roles, come differences in hiring and recruiting (Chamorro-Premuzic and Akhtar 2019; Dattner et al. 2019). By clearly understanding the key skills, competencies, or experience that a role may require, hiring managers can utilize AI to find potential applicants, and sort through the applicant pools by scoring resumes based on specified criteria leading to more qualified applicants being considered for sales roles.

• “Empathy, emotional intelligence, curiosity and business acumen need to be taught to humans since these are things that AI cannot do.” (Adam, Vice President)

The managers will also need to be trained on how to optimize their work using AI, but they may be overwhelmed with the amount of structured and unstructured data that is available to them. They will need training on the different AI tools to understand how different processes can be streamlined or automated. ML has the capability to analyze prior data to predict sales trajectories and aid in forecasting, whereas NLP can examine sales calls to identify key words or phrases that provide insight into the wins and areas of opportunity for the sales team (Bawack, Wumba, and Carillo 2021). Generative AI can then recommend or develop personalized training for sales team members based. Sales managers will need to vet the proposed courses of action, but they do not have to develop training from scratch. Sales managers can also encourage their teams to learn and train on their own using the AI feedback throughout the sales process.

• AI will help with each step of the sales process because it can help listen to the call and give feedback. For example, did someone talk too long? It can give a summary of the call and identify the pain points and save a lot of time.” (Louis, Account Manager)

From the interviews, it seems that the technology may also provide a unique opportunity for training for sales managers themselves. It is often noted in academic literature and in speaking with sales managers that they get very little (if any) training and are typically thrust into the roles because of selling ability (Plank et al. 2018). However, AI may change some of this as the co-mingling with other sales technologies (e.g. SFA, CRM or activity dashboards) can supply managers with a starting point for leading their teams by better highlighting the successes and internal pain points. AI can provide recommendations about which salespeople need additional coaching and how to target the coaching. According to Luo et al. (2021), managers need to also be aware that top and bottom performing agents respond differently to AI coaching. Each salesperson may require different combinations of sales manager and AI coaching, which only adds more complexity to the role of the manager.

• “Managers could use some AI to help with their coaching as well. Especially for younger or newer managers with little to no training or other people to learn from” (George, CEO)

• “What type of personality or what will predispose someone to AI and AI led coaching?” (George, CEO)

Another challenge for managers will be to motivate the salespeople to “buy in” to the technology. Many of the positive things that can come from the use of the unique AI technologies are contingent upon the idea that salespeople will readily adopt the training and AI recommendations. This may be difficult for sales managers as they will have to believe in the value of the tools themselves before trying to get their teams to implement them into their routines. In fact, recent qualitative research from Epler et al. (2023) finds that managers often expend substantial amounts of time and energy in helping their salespeople to better understand the benefits of various customer-facing activities. It is likely that many of these same issues may be intensified with such radical technology. Implementation of AI could lead to managers focusing more on training, motivating, and providing support for their teams and once the salespeople are on board, they can use the tools for monitoring and coaching.

• “For AI to work you will need to get buy-in and train everyone a lot. This means there cannot be competing priorities, and you need to focus on the top things. To use the tools properly, employees will need to be trained over and over again. It needs to be built into their routines and if it doesn’t work well, they will fall back into their old habits. (Peter, Division VP).

• Do [managers] give the sales teams time to utilize and explore the new technologies or are they pressured to hit a quota or deal with other tasks? [salespeople] may be more willing to incorporate [AI] if they have the time and tools to do their jobs correctly.” (Natalie, BDR)

• “For transformational changes, there needs to be change management. This may require a lot of hand holding and training in order [for salespeople] to be successful.” (Adam, VP)

Organization

AI is considered a technological and innovative capability for firms, but it can impact other organizational practices, thereby creating challenges and opportunities (Weber et al. 2023). Drawing from previous management literature, we highlight the impact of AI on three different capabilities: strategic management, operational, and external stakeholder (Amit and Schoemaker 1993; Donaldson and Preston 1995; Freeman 2010; Porter 1991) in Table 4. Strategic management capabilities include the objectives and goals of the organization and the alignment of strategy, capital expenditures, and activities (Porter 1991). As AI is implemented across business functions, it is important for firms to reevaluate their strategic objectives accordingly. Knowledge, employees, and processes are examples of operational capabilities that can give organizations a competitive advantage. AI implementation means that certain capabilities may be complemented or even replaced (e.g. information gathering, process optimization and strategic decisions). Relationships with channel partners and other stakeholders can directly impact the resources and information available to organizations (Burchett et al. 2023; Donaldson and Preston 1995), thus, communication about the usage and boundaries of AI must be clear for all parties.

Table 4. Organizations and AI: implications and applications.

	Types of AI applications for organizations			Implications
	Machine Learning	Natural Language Processing	Generative AI	Opportunities	Challenges
Strategic management capabilities	Assess current performance against strategic goals Evaluate the allocation of capital expenditures	Evaluate communications against strategic objectives and goals Summarize information and feedback regarding strategic actions Analyze feedback given by top management across departments	Recommend future strategic objectives and actions Develop messaging for employees and consumers that matches strategic goals	Cohesive messaging received at all levels of the organization Increased efficiency in capital expenditures	Overall strategic goals may not match departmental goals
External stakeholder capabilities	Assess and score performance of channel partners Identify potential channel partners utilizing specified criterion	Summarize strategic goals of relationships Evaluate channel partner offerings	Recommend new channel partners Develop customized messages to channel partners Recommend solutions to channel partners	Increased transparency in channel partner performance Increased value created in relationships	Governance of data Accurate and timely information is necessary from all parties
Operational capabilities	Identify inefficiencies in processes Analyze labor cost expenditures	Develop standard operating procedures from task information Analyze communication across departments	Recommend opportunities for new products or services to fit customer segments Recommend innovations or reconfigurations of business aspects	Increased employee productivity Increased cooperation across departments	Integration of systems and data across departments may be difficult

Note: ◆ bullet points indicate applications that are well established; ○ bullet points indicate applications that are not well established.

Organizations are utilizing cognitive computing to manage data, help with interactions and enterprise cognitive computing (ECC) to streamline organizational processes. To provide an illustration, an AI-trained algorithm will be embedded within repetitive and replicable organizational tasks. This can include answering customer questions at any time and diagnosing issues that are routed to the appropriate team (Tarafdar, Beath, and Ross 2019). In practice, AI is being used to help with decision making, and problem identification in an autonomous manner. This means that organizations must decide the appropriate balance between task automation or task augmentation (Raisch and Krakowski 2021).

However, before integrating AI into the tech stack, it is important for upper-level managers to understand if AI makes sense for their organization and if they are prepared for the adoption (Jöhnk, Weißert, and Wyrtki 2021). For additional complexity, organizations often deal with ill-structured problems that lack easy answers. This necessitates human “checking” of AI decisions and problem solving (Von Krogh 2018). For AI to work well for an organization, data must be consolidated, quality must be checked, and the sources vetted. In many instances, data is compartmentalized which limits the scope of AI as small data sets are not yet useful (Hagendorff and Wezel 2020). Integrating the fragmented data from all departments is the first step and can be time consuming. AI has the capability of linking departments together. The increased transparency and availability of data across departments can lead to more cohesive strategic objectives and actions. Next, organizations also need to look at the quality of their data before training the AI to make decisions. Data is often “messy” and generating predictions and insights may be more difficult than many have proposed, especially in complex situations. It is possible that the AI recommendations will be incorrect if there is poor quality data.

• “[AI] fundamentals matter, you will do things wrong faster if the fundamentals are not in place.” (Daniel, Sales Consultant)

• “There needs to be consistency among departments along the customer journey. Departments should work together at each stage.” (James, Sales Consultant)

If AI is to be added into an organization, it is important for high level managers to be realistic about what the AI can do and how it will help lower-level employees. Kozinets and Gretzel (2021) caution organizations to avoid overstating the benefits or promises of AI. For example, they state the risks of becoming “subordinate” to the technologies (instead of the other way around). The authors further state that the “black box” nature of AI is daunting. This was a sentiment that was shared by our qualitative participants. Our respondents indicated that this challenge occurs because there is a disconnect between the developers of the tools and the salespeople/managers that implement them. The software engineers may develop a complex, multifaceted tool that they believe will improve the work of the sales team, but it is common for them to create the tool without ever talking to someone in that department. This can increase the frustration of the salespeople and sales managers if the technology hinders their ability to complete tasks instead of helping them. It is important for upper management to include different business functions in the meetings with software engineers to avoid these issues from happening and constantly train the technology to match the needs of the business.

• “Many AI platforms have a lot of bugs since they don’t talk to people within the business. They are programmed by people with software knowledge, but there needs to be some conversations between the parties during the setup of these programs.” (Peter, Division VP)

• “[Currently], there is a separation between the software department and organizations. IT lacks business thinking in terms of routines, processes and standard operating procedures so some type of conversation needs to be had. They need to get in the same room.” (James, Sales Consultant)

• “Leadership does not always understand the processes of all parts of the business, yet they want to layer AI on top of it without knowing how it could be used in the lower levels.” (Isabella, Management Lead)

• “(Organizations) need to train the AI, like other employees, in order for it to be useful. The training needs to be based on organizational needs and conducted by higher-level employees.” (James, Sales Consultant)

• “The handoff [of data from AI to the salesperson] is complicated, there needs to be consistency between the chatbots and the sales team. Wrong info means the unselling of your product.” (Matthew, Manager)

AI will likely have great impacts on the structure of the organization. Roles within sales departments will adjust and change, meaning potential restructuring. Data and analytics have become an important focus at every level of the organization, but more data does not necessarily lead to immediate efficiency. If AI can connect data from across departments, it will be easier for the different business functions to communicate and strategize with one another. Even with the help of AI, it will take time and effort to sift through the available data. The next task will be for upper-level managers to determine which roles and/or processes will be replaced by AI, and which will be augmented (Floridi and Cowls 2022; Pappas et al. 2023), but these decisions may be dependent on firm size, available resources, costs, innovative culture, and other capabilities. AI requires a large upfront cost to ensure a cohesive implementation with current technologies. Other costs include data storage and memory as large amounts of information are needed for AI.

• “AI is an interface between different departments, such as sales, marketing and engineering.” (James, Sales Consultant)

• “The predictable revenue model is broken…. The [sales development representative] SDR to account executive [AE] does not work anymore. The SDR might be going away as AI can outdo [them]. AI may replace the SDR.” (George, CEO).

• “(Every organization needs to decide) if AI is helping or hurting…More data means you will be busier and being inundated with data means that there will be information overload. You need storage and memory space which are additional costs.” (Peter, Division VP).

• “Organizations need to ask themselves when is AI going to have an impact in my business? If the cost is less than the benefit, they will have some clues. Other things to consider will be if AI can make the best decision faster or more efficiently.” (Frank, VP Analytics).

From a data governance perspective, organizations have a lot to be concerned about with regards to AI. Data transparency from firms upstream and downstream can allow organizations to delve deep into the spending, offerings, and performance of channel partners. Organizations can employ the technology to assess current relationships and recommend potential partnerships, but their customers or suppliers may be evaluating the organization’s performance as well. For established relationships, holding counterparts accountable becomes easier as they will be able to see how their relational counterparts are utilizing their money and resources. However, there are many difficult aspects, such as issues with customer data, customer privacy, proprietary information and the “black box” nature to consider (Floridi et al. 2021). Costs for data privacy and security will also increase as organizations seek to protect information. Part of the training that all employees will need to receive should clarify what information can or cannot be utilized with AI technologies. An important point is that information being input into applications like ChatGPT becomes property of OpenAI. Thus, it is strongly advised that organizations avoid putting personal/sensitive information into these types of applications. Guidelines and rules will need to be put in place for the tools to ensure data privacy and security for all parties.

• “Channel partners will want access to the data of others. They will want to see if the money and resources are being utilized properly.” (Adam, VP).

• “Companies are being told to stop using ChatGPT. [Organizations] were giving [AI] data to learn with, but some of the information was proprietary, about their competitors, or even incorrect or illegal to share.” (Matthew, Manager)

• “In regard to privacy issues with AI, some companies have a lot to lose if things go wrong.” (Stuart, CEO)

Customer

The adoption of AI impacts not only the sales process, but also the customers’ purchasing process. Our customer framework builds upon current literature on the customer journey focusing on behaviors that occur either pre-purchase, purchase, and post-purchase situations (Lemon and Verhoef 2016). Before purchasing, customers identify their needs and then begin the search for suppliers. Once a supplier has been selected and the purchase has been made, the customer may continue engagement with the supplier and request for further services. Building on the customer journey, our current framework focuses on the customer’s use of AI within their purchasing process and is presented in our Table 5.

Table 5. Customers and AI: implications and applications.

	Types of AI applications for customers			Implications
	Machine learning	Natural language processing	Generative AI	Opportunities	Challenges
Need recognition, consideration, & search	Analyze expenditure Optimize stages of the purchasing process Contract management and KPI monitoring Search for optimized price/value by comparing suppliers’ real-time prices online	Matching systems that analyze all the features of a supplier Quality control of various internal data Identify the supplier’s online WOM	Chatbots to contact, select, and share information with suppliers Summarizing general conditions and areas for improvement Summarizing online reviews/information to have a clear overview of the market Offer relevant product comparisons, customer reviews, and testimonials to help buyers evaluate their options.	Reduced workload Optimized processes New and more competitive suppliers	Risk of too much relying on AI and lose sight of what is important Selecting suppliers on analytical elements and not considering it as a trusted partner
Choice, ordering & payment	Analyze prior purchase decisions Improve decision making and selecting the right supplier	Analyze supplier’s message and emotions Scan through all supplier’s quotations to optimize choice	Provide in-real time advice for improved negotiation Support the production of complex purchasing documents (e.g., calls for tenders, contracts)	Buyer focuses on more complex tasks Optimized purchase decisions Improved efficiency	Buyer’s opinion might be challenged by AI Mechanical process could lead to a “battle of the machines”
Consumption & usage	Send direct updates and information about product/service usage to supplier	Analyze any post-usage social media content to gauge sentiment	Usage assistance and recommendations	Better usage reduces negative customer experience	A firm getting information about in-real time usage can be seen as intrusive for customers
Engagement & service request	Electronic supplier relationship management enhanced by real-time data to stay the supplier’s best client	Evaluate reviews, social media chatter, or communication with salespeople to determine courses of action	Direct feedback to send to suppliers	Sharing direct feedback to firms allow them to better match customer’s preferences and improve their product and services to match customer needs	AI may not be able to build trust with the customer Engagements rely on competitiveness

Note: ◆ bullet points indicate applications that are well established; ○ bullet points indicate applications that are not well established.

Like salespeople, customers are benefiting from AI early in the purchasing process for simpler tasks such as information searching or identifying potential sellers. AI can learn from a vast amount of publicly available data (e.g. social media, financial reporting, company websites) prior to any interactions with salespeople, saving time and money for buyers (Salminen, Ruohomaa, and Kantola 2016). It may be more beneficial for a customer to interact with a chatbot to receive quotes and responses in real time, rather than waiting for assistance from a salesperson (Ahearne and Rapp 2010). AI chatbots are identifying potential suppliers and providing customers with more specialized, targeted information (Chung et al. 2020), which is then entered into a supplier matching system (Allal-Chérif, Simón-Moya, and Cuenca Ballester 2021). These systems can analyze the nuances between sales proposals to compare and assess suitable channel partners (Lu and Hong 2019) more objectively and remove potential biases in negotiations (e.g. Hada, Grewal, and Lilien 2013; McClure et al. 2023). Customers may supplement their purchasing decisions with insights from generative AI to make sense of the knowledge from multidimensional data (Baryannis et al. 2019). Nevertheless, at this stage the understanding of AI in purchasing remains relatively limited (see Allal-Chérif, Simón-Moya, and Cuenca Ballester 2021, Karttunen, Lintukangas, and Hallikas 2023). However, the inevitable advancement will have direct implications on customers as they may have fewer and more limited interactions with salespeople.

• “As [AI] becomes buyers, it will change the way that sales have been done. You may not need relationships or salespeople, but instead their role will be more like customer success managers.” (George, CEO)

As such, the integration and implications of AI on customer-seller relations needs to be understood from both sides (see Chang 2022; Huang and Rust 2023). Many firms jumped on the bandwagon of AI usage in customer interactions without fully considering whether customers would accept this new technology (Liu-Thompkins, Okazaki, and Li 2022), or the impact on the customer experience (Trawnih et al. 2022). Customers have already been forced to embrace new ways of doing business and accept technology at an accelerated rate due to COVID-19 (Collins et al. 2021), yet many are still not on board or ready to work with AI. Like building relationships with salespeople, customers may need to develop a relationship of sorts with the AI to understand the value. This may be difficult, as there is a fair amount of risk involved for customers to interact with these new technologies and many are suspicious (Burns et al. 2023). Ultimately, trust is essential in any relationship and the introduction of AI into the mix can create tensions if it is unclear what the AI is doing or what data it is accessing, but these concerns can be alleviated with transparent communication.

• “For integration with customers there needs to be trust. Are both parties comfortable with the information? You need to validate what you are doing is working and the AI is doing what it says it will do.” (Peter, Division VP)

• “Salespeople need to identify and understand the customer’s threshold for AI acceptance before sending emails and using other tools.” (Peter, Division VP).

In general, it will be important to focus on the intersection between the customer’s interests and the potential financial benefits. Because AI “listens” so well, an exorbitant amount of information is available, yet it is important to note that there are many questions from customers related to the type and amount of data that is being collected about them from sources like CRM databases and social media. Customers may feel uncomfortable or as though their privacy has been violated if they feel the AI has gone too far. Additional uncertainties may develop as AI learns and develops as it is unclear whether they are dealing with a real person or a chatbot during sales interactions (Gnewuch et al. 2023).

• “People don’t like to be tricked and [AI] should try not to be creepy. Customers will not like it if the [AI] goes too far.” (Isabella, Management Lead)

• “Most companies just assume that people are ok with dealing with [AI], but there are still questions about the suspicious nature and risk of using it.” (Olivia, CEO)

Further usage and adoption of AI by salespeople may also impact the different stages of the customer journey/purchasing process. Customer interactions with salespeople could involve less cold calling as earlier stages become more automated. This will speed up an existing trend, as customers may not interact with the salesperson until even later in their decision cycle (Ahearne et al. 2022). It is also likely that AI will help to identify the seller that best can serve the needs of the customer. In addition, AI can learn at a much faster rate, thus providing more customized recommendations for salespeople to present to customers. This will ideally lead to a streamlining of the latter aspects of the purchasing process. By utilizing the AI tools, the customers can capture additional value and build a sense of trust and commitment with the tools (Trawnih et al. 2022). This will likely become more prominent as the tools are trained to be more humanlike (Liu-Thompkins, Okazaki, and Li 2022). Ultimately, the customer-AI relationship adds another layer of complexity to B2B relationships.

• “AI will help a sales organization to better understand the customer journey. On a basic level “how do they buy?”. The beginning rates a flow for the “right” customer engagement throughout the process. The end result will be converting the sale and referring others. Throughout (if used appropriately) it will help to remove customer fear or change fear into a different feeling.” (James, Sales Consultant).

• “AI can help with the customer journey…give trust and reduce fear, align oneself with customer journey, this is the key.” (Daniel, Sales Consultant)

Given the insights that our interviews and prior literature have generated, we now turn to future research opportunities to further understand the implications of AI in sales.

Future research

We believe our efforts at clarification of the current state of AI in sales will allow sales researchers to better engage in AI-related sales research. As Raisch and Krakowski (2021) have noted, AI can introduce tensions between automation and augmentation. We build upon their work by identifying additional tensions that can emerge during the different aspects of the sales process (Figure 2). These tensions can create tradeoffs between the stakeholders. Because of this, we have created a future research agenda which will help to unpack these complexities, with key questions at the different levels of the sales function (e.g. salesperson, sales manager, organizational, and customer) in Table 6.

Figure 2. The implementation of AI in the sales process: evolution and tradeoffs.

Table 6. Future Research Questions.

Stakeholder	Area of impact	Research questions
Salesperson	Sales Process: Augmentation & Automation	What sales tasks will become AI tasks, AI + human tasks, or remain human tasks (see first insights in Chang 2022; Pappas et al. 2023). what specific sales role will AI support or replace (e.g. purely prospecting roles, like SDR versus purely relational roles like KAM)? As repetitive tasks are being done by the AI, what new skills and competencies should salespeople learn to be ‘irreplaceable’, or in other words, what human characteristics can or cannot be taught to AI? To what extent do salespeople need to understand the technical complexities vs a basic understanding of AI to utilize the technology? To what extent will salespeople be responsible for training and supervising the AI? What are the tradeoffs between efficiency and quality with AI in the sales process? How can/when do salespeople utilize sentiment analysis during negotiations to adapt their approach or closing tactics?
	Relationship Development	How do service expectations from customers differ between AI-customer and salesperson-customer interactions? How will the development of customer-salesperson relationships change due to the addition of AI? What are the implications of strong customer-AI relationships and how will they differ from customer-salesperson relationships? To what extent can salespeople develop a relationship with AI technology? To what degree should the AI share information with salespeople to enable enough transparency and explanation so that the salesperson trusts the AI, but not too much information to avoid information overload? Does AI impact the salesperson’s feeling of relational ownership or responsibility?
	Mental Health & Well-Being	What impact will AI have on a salesperson’s well-being as the traditional role of the salesperson changes? While AI can unburden salespeople from monotonous tasks, it forces them to deal with more complex issues on a regular basis. How will the removal of the simple tasks impact salesperson well-being? What impact does AI adoption have on salesperson role stress? Will AI have a positive or negative impact on salesperson burnout/attrition?
Sales manager	Training & coaching	What new skills and competencies will be necessary for salespeople to be successful? Following salespeople’s need for new skills and competencies in the AI era, to what extent is it possible to train salespeople on such new skills (e.g. soft skills)? Should sales managers have new recruitment criteria? To what extent can skills such as business acumen, emotional intelligence, and interpretation of data be taught? Will the AI generate better feedback than the manager, and to what extent will salespeople accept and follow recommendations of AI feedback versus the manager’s feedback? How can managers encourage the adoption or “buy-in” of AI for their sales teams?
	Performance management	How will performance measures and KPIs adjust as AI is integrated into sales? To what extent should sales managers redefine the sales compensation plan considering AI’s integration into the sales process? How can managers utilize real-time feedback from AI to encourage their team and increase engagement? How can sales managers navigate the potential divide between salespeople and sales engineers? How can sales managers organize and facilitate the optimum level of involvement amongst different departments?
	Account management	To what extent will the sales manager’s role be transformed when AI takes over the role of forecasting, sales monitoring, and providing instant feedback to salespeople? To what extent will AI be able to integrate DEI aspects into the system knowing this information remains confidential and personal (e.g. allowing more flexibility for someone disabled, granting higher work-life balance after pregnancy), and how will managers balance AI’s recommendations considering this perspective?
	Recruitment & hiring	As AI automates much of the hiring and recruitment process, how does the role of the manager change? Are there tradeoffs in terms of applicant quality with the usage of AI? What traits and/or competencies can or cannot be captured by AI during the hiring process?
Organization	Strategic management	To what extent can organizations implement their culture, values, or strategic objectives into AI-generated content? What are the short- and long-term implications of AI on firm resources? As AI requires a colossal amount of computing power and data, increasing the energy footprint and computing costs, how do organizations integrate this in cost/benefits analysis and ensure their sustainable goals? How can organizations that use private LLMs to overcome and resolve ethical concerns (e.g. transparency, data privacy) among their internal stakeholders (i.e. salespeople, sales managers) and external stakeholders (i.e. customers, third parties) when AI is used as a customer-interaction tool?
	External stakeholders	How can organizations prevent AI from leaking proprietary information to external stakeholders? What impact will transparency between channel partners have on the risks associated with resource exchange and outsourcing (i.e. opportunism)? How can organizations use AI to source external stakeholders, synthesize KPIs and evaluate external stakeholder performance?
	Operations	To what extent should organizations involve business functions in early stages of AI adoption to avoid disconnects between the technology and departmental tasks? What department (e.g. IT, sales, marketing) or function (e.g. the user, the manager) within the organization is responsible for training employees on AI technology? What elements need to be considered to measure the cost/benefits of implementing AI in the sales process? While organizations use mass lead generation allowed by AI, how do they ensure they don’t lose more of what they get? Should organizations adjust their pricing strategies considering AI, a cost-effective tool, might partially replace some salespeople?
Customer	Purchasing process	At what stage of the purchasing process will the customer need a human, and what is the core reason why customer’s need humans (e.g. trust, more creative solution, someone who “understands”)? How will interactions and negotiations change as AI is incorporated in the purchasing process?
	Customer journey	At what stage(s) of the customer journey is AI most impactful or harmful to negotiations or customer-seller relationships? How do customers perceive the sales message generated by an AI versus a salesperson especially if it replicates human characteristics (e.g. empathy)? In what ways can customers identify if the content/message is generated from an AI or a salesperson and should there be a mention of the fact that it is AI generated? How will customer’s pricing perceptions and negotiations change when dealing with an AI salesperson?

In general, we encourage future research into the tradeoffs of AI implementation. Indeed, as Rapp and Habel (2024) have recently called for scholars to emphasize tension to advance knowledge and the expansion of AI into sales provides ample opportunity. For salespeople, it is evident that AI will have profound implications on various aspects of their role and additional downstream impacts. It is impossible to predict what will occur, but we propose relevant questions about fundamental changes to the sales process, relationship development and well-being. Specifically, we encourage future research into how the tradeoffs that AI will have on relationship intricacies with customers and if there are unintended consequences for the potential wellbeing benefits. The impact of AI for managers regarding training, performance management, recruitment, and account management also warrant further study. AI presents sales managers with a more granular look into the work of their sales teams, and specific ways that managers can utilize real-time feedback to train their teams and promote engagement. However, this brings up issues of salesperson “buy-in” and potential tensions that may arise with the inclusion of newer KPIs and increased workload.

For organizations, we present research avenues about strategic management, operations, and relationships with external stakeholders. One of the key future research questions for organizations is how to mitigate potential disconnects between AI engineers, who design the technology, and those who will be utilizing the technology. Furthermore, as customers utilize AI throughout the purchasing process, customer-salesperson interactions will continue to evolve. For customers, it is important to understand when AI utilization will be most impactful during the buying process. In a similar manner, it is imperative for salespeople to understand the benefits and harms of AI utilization during the customer journey. Because of the scope, a detailed discussion of the various research agendas can be found in Web Appendix C.

Conclusion

To summarize, a key benefit of this manuscript is to clearly distinguish between the different types of intelligence available for sales organization and identify how they help drive the sales process and help aide sales managers do their job effectively. We build upon existing sales research by (a) providing a classification of different intelligence systems available, (b) suggesting how these different intelligence systems can help drive automation in routine and more complex sales tasks (c) identifying how intelligence systems can accentuate the efficacy of the existing tech stack, and d) provide some initial insights into how practitioners approach different type of intelligence systems.

In closing, for decades now, practitioners and academics alike have talked about the potential of AI to disrupt normal routines and tasks, thereby playing a transformative role in organizations. However, the past decade has witnessed a rapid evolution of AI from engaging in simple automation to more complex tasks that can augment the work of employees in organizations. In this study we looked at how AI is continuing to transform the sales function across all levels. With a grounding of literature in the practitioner and academic domain in addition to complementing the findings with discussions with 18 professionals, we identify specific research questions that can further our understanding of how AI is impacting the sales function.

Disclosure statement

No potential conflict of interest was reported by the author(s).