the performance of employees [4]. Studies suggest that performance appraisal based on fuzzy logic helps in drawing definite results from ambiguous information [25].
Furthermore, various chatbots are used in the process of performance appraisal. Example: Engazify is used for performance appraisal as it gives real-time feedback and appreciation to its workforce [2]. Besides, data analytics and big data are also used to evaluate the performance of the employees, wherein the grades and the ratings are assigned on a scientific basis. The process begins by feeding the integrated performance metrics into the analytics software to determine the ranks of the employees. Such automated performance appraisal is free from biases that may occur by human beings while assigning grades or while ranking and brings transparency in the appraisal system.
1.3.5 Training
In this era of disruptions, the concept of one size fits all (same course content) for the learners cannot be applicable. Through AI, the learning material can be personalized in accordance with the learner’s requirements (skill gaps). With the help of AI, suitable content can be recommended to the learners, based on their past behavior. Besides, several content creation algorithms can be used to auto-generate content. AI gives the flexibility to the employees to learn at their own pace. Studies indicate that the robot training instructor can track the daily learning status of the learners and can even compute the average value of the learners’ attention [17]. Based on the learning objectives entered by the employees, the robot training instructor can automatically complete the course. Thus, AI facilitates personalized learning.
Furthermore, a qualitative study conducted by IBM Smarter Workforce Institute on senior HR executives of IBM revealed that their organization is enhancing skills inference technology internally. Consequently, employees of IBM have access to their real-time skill insights through an expertise management interface, which is more accurate. AI skill inference technology also helps IBM to analyze the skills of its employees relative to business needs and can also compare the skill profile of its employees with its competitors. This helps IBM to bridge the skill gaps of its employees [15]. Another example of an AI system as used in the military is intelligent tutoring systems [21, 30].
1.3.6 Compensation
Some of the prominent organizations like Google and Tesla use techniques like big data, predictive analytics, and ML techniques to monitor the talent of their employees, and thus, based on their performance, they remunerate their employees. These companies are following the recommendations of the AI-backed software and thus ensure that their employees are not under or overpaid. Example: IBM uses an AI-powered decision support system which helps in the compensation planning of front-line employees and thereby overcomes the issue of underweighting or overweighting the critical data points [15].
1.3.7 Employee Retention
AI helps in employee retention by satisfying the employees through ensuring unbiased performance appraisal. Algorithms can predict as to which employee is likely to leave the organization. AI software can predict the likelihood of employee turnover by tracking their browsing history and emails. Organizations are also using AI-based mood meters which help in tracking the sentiments of their employees and assist the organization in identifying the causes of employee turnover. The organizations are also using predictive statistical models that help in forecasting the employees’ intention to quit the organization and thus help in preventing employee turnover [13].
1.4 Artificial Intelligence in Marketing
The marketing concept comprises of 4Ps, namely, Product, Price, Place, and Promotion. This concept is all about making customer the king, i.e., satisfying the needs, wants, and desires of the customers. The customer satisfaction, over the period, graduated to customer delight. The organizations can delight the customers only when the tastes, preferences, and behavioral aspects regarding their purchase can be traced or known. With the advent of technological advancement during recent times (AI), a wealth of information about the consumers, their consumption patterns, and purchase behavior can be traced to a large extent. The database can be created for the information collected about consumers. The pattern analysis using data mining techniques can reveal homogeneity and heterogeneity. This shall reveal the basis for segmenting the markets into a precise group of consumers/customers. The latest technological innovations especially AI have opened an opportunity for the marketers to enhance the effectiveness of marketing campaigns which can be measured as a return on investment (ROI). The application of AI in marketing and sales has the highest potential value, with estimates up to 2.6 trillion dollars [7]. Earlier marketing was a one-way communication or push the product to inform, persuade, and remind with catchy slogans/jingles. With the advent of AI, consumers can buy/sell anything from any part of the planet anytime. The studies show that “consumers today search much less on brand names than they did 10 years ago. If someone wants to buy shoes on Amazon, they are five to six times more likely to search by category name than by brand name and follow the recommendations suggested by the Amazon algorithms” [32]. The industries where the customers are large in numbers and need to interact frequently with customers have a high potential for using AI. Since the interaction between the two generates a huge amount of data [10], customers perceive AI at a very high level [12]. Potential message from an AI application is convincing when it pertains to the usage of the product or service, instead of why that product or service should be used [19].
AI algorithms can perform pre-defined tasks, for example, automated email replies, blocking of debit/credit cards, etc., and also these AI algorithms can analyze the customer data which can be in various forms, viz., text, voice, and facial expression.
Marketing has become a two-way communication which means consumer searches/transacts with the seller. Some of the ways in which AI can be used for marketing are as follows.
1.4.1 Creation of Customer Profiles/Market Segmentation
The customers’ needs and wants are of immense importance for marketers. Traditional marketing generally used the feedback from consumers’ and also the marketers had to rely on the data provided by the market research firms. With the advent of AI and more people inclined to use the digital platform to search for their requirements, the marketers can now precisely segregate the customers for their product/service requirements. The technological advancement has let the marketers collect the customer’s data such as customer’s name, mobile, email, gender, search pattern, and so on. With this data, marketers can create customer profiles. Therefore, the customers can be segmented and targeted for personalized promotions. It can also help in retaining the customers. Studies indicate that VPSAs (Virtual Personal Shopping Assistants) can predict and optimize the tastes and needs of customers [11]. Lucy: it is created by Equals3 and is named after the granddaughter of IBM’s founder Thomas Watson. It can analyze structured and unstructured data. It helps in segmentation, planning, and interaction with humans in an easy way. SOFMs (self-organizing feature maps) are used for market segmentation, i.e., portioning of a large market into small homogeneous groups of consumers.
Hidden Layer
Figure 1.2 ANN for market segmentation.
The market segmentation for an organization provides translate the opportunity for not only optimally utilizing the resources but also, at the same time, ensuring high profitability. But it remains a big challenge to translate the market’s needs in a precise manner. The ANN provides the solution with several methods developed over a period of time. The SOM (self-organized feature maps), GKA (genetic K-means algorithm), and ART (adaptive resonance theory) are some of the methods used for clustering/segmentation.
An ANN can be constructed for segmenting the market, suppose the parameters for the customer