Table 4 Coding process Phase 1:... | Emerald Publishing

Table 4

Coding process

	Phase 1: Questioning and analysis of current activity systems and customer care practices	Phase 2: Modeling new chatbot-based solution	Phase 3: Examining and testing multiple chatbot-based solutions	Phase 4: Implementing and reflecting on AI-augmented, chatbot-based solutions	Phase 5: Consolidating new human–robot activity system and new customer care practices
Selective coding	Service activity tensions lead to the use of new resources in customer care activities	Resources are applied and integrated to model a new solution, but additional tensions emerge	Improvements are performed while implementing the new model, resulting in the emergence of further tensions	Refinements are executed through testing and implementing multiple chatbot-based systems	Results and performance allow for enacting new work practices
Axial coding	Tensions in considering additional resources (personnel or automation) to manage workloads efficiently and handle customer inquiries appropriately	Tensions in improving chatbot training and performance to meet the diverse nature of all customer inquiries	Tensions in performing interconnected actions among all involved actors (i.e. chatbots, employees, customers and third parties)	Tensions in collecting and automatically analyzing huge volume of customer data insights to improve company’s strategies	Use of chatbots as collaborators with employees to foster improved customer care service
Open coding	Managing customer needs Repetitive inquiries Time wastage Regulatory challenges in health products	Chatbot inefficiency Training challenges Customer frustration Complexity	Distinct scenarios Autonomy of chatbots Outdated information Limited employee involvement Third-party connection	Generative AI enhancement Workflow improvement	Enhancement of work practices Collaborative work context
Raw data (employee interviews and observations)	Employee 3: “Managing the multitude of customer inquiries becomes an insurmountable task” Employee 12: “Customers always ask the same thing, and we have to sit there wasting time to answer” Employee 4: “The institutions are not on our side. Since these are health products, you need to pay attention to every word that is said”	Employee 9: “The newly implemented chatbot promptly addresses inquiries around the clock with a set of answers that we have specially prepared. This is a small step forward!” Employee 2”: The chatbot is still in its early stages …. Its training is not complete.” Employee 5: “Customers were often annoyed because they had not received a response from the chatbot.”	Employee 11: “We manage entirely distinct scenarios due to the existence of seven different chatbots.” Employee 6: “Their training is not interconnected. When information changes, it needs to be updated everywhere.” Employee 3: “Chatbots are not always comprehensive; a lot depends on their history.” Employee 8: “Only a limited number of employees are directly involved in chatbot training.”	Employee 1: “I don’t believe there was any concern about it replacing work. On the contrary, from our perspective, it helped us to reorganize and enhance it.” Employee 10: “Generative AI has revolutionized our approach to customer interactions, liberating us from the challenges of data overload. It has streamlined our workflows.”	Employee 14: “Nowadays, the chatbot is utilized to identify potential areas for enhancement”. Employee 13: “Through our new systems Customer data can be used not merely as a tool for analysis and response improvement but as a strategic asset that can be seamlessly integrated into our marketing strategies.”

	Phase 1: Questioning and analysis of current activity systems and customer care practices	Phase 2: Modeling new chatbot-based solution	Phase 3: Examining and testing multiple chatbot-based solutions	Phase 4: Implementing and reflecting on AI-augmented, chatbot-based solutions	Phase 5: Consolidating new human–robot activity system and new customer care practices
Selective coding	Service activity tensions lead to the use of new resources in customer care activities	Resources are applied and integrated to model a new solution, but additional tensions emerge	Improvements are performed while implementing the new model, resulting in the emergence of further tensions	Refinements are executed through testing and implementing multiple chatbot-based systems	Results and performance allow for enacting new work practices
Axial coding	Tensions in considering additional resources (personnel or automation) to manage workloads efficiently and handle customer inquiries appropriately	Tensions in improving chatbot training and performance to meet the diverse nature of all customer inquiries	Tensions in performing interconnected actions among all involved actors (i.e. chatbots, employees, customers and third parties)	Tensions in collecting and automatically analyzing huge volume of customer data insights to improve company’s strategies	Use of chatbots as collaborators with employees to foster improved customer care service
Open coding	Managing customer needsRepetitive inquiriesTime wastageRegulatory challenges in health products	Chatbot inefficiencyTraining challengesCustomer frustrationComplexity	Distinct scenariosAutonomy of chatbotsOutdated informationLimited employee involvementThird-party connection	Generative AI enhancementWorkflow improvement	Enhancement of work practicesCollaborative work context
Raw data (employee interviews and observations)	Employee 3: “Managing the multitude of customer inquiries becomes an insurmountable task”Employee 12: “Customers always ask the same thing, and we have to sit there wasting time to answer”Employee 4: “The institutions are not on our side. Since these are health products, you need to pay attention to every word that is said”	Employee 9: “The newly implemented chatbot promptly addresses inquiries around the clock with a set of answers that we have specially prepared. This is a small step forward!”Employee 2”: The chatbot is still in its early stages …. Its training is not complete.”Employee 5: “Customers were often annoyed because they had not received a response from the chatbot.”	Employee 11: “We manage entirely distinct scenarios due to the existence of seven different chatbots.”Employee 6: “Their training is not interconnected. When information changes, it needs to be updated everywhere.”Employee 3: “Chatbots are not always comprehensive; a lot depends on their history.”Employee 8: “Only a limited number of employees are directly involved in chatbot training.”	Employee 1: “I don’t believe there was any concern about it replacing work. On the contrary, from our perspective, it helped us to reorganize and enhance it.”Employee 10: “Generative AI has revolutionized our approach to customer interactions, liberating us from the challenges of data overload. It has streamlined our workflows.”	Employee 14: “Nowadays, the chatbot is utilized to identify potential areas for enhancement”. Employee 13: “Through our new systems Customer data can be used not merely as a tool for analysis and response improvement but as a strategic asset that can be seamlessly integrated into our marketing strategies.”

Source(s): The above table was created by the authors