Sophisticated Customers Research Articles

GPT wars: Creating and detecting fake ophthalmology data sets with artificial intelligence

Aims/Purpose: Recently, our group demonstrated how the Large Language Model GPT‐4 can fabricate fake ophthalmology data sets, designed to support false scientific evidence. The objectives of this study were: (1) to identify peculiar statistical patterns of AI‐generated data sets, (2) to attempt at enhancing the quality of fabricated data for eliminating any discernible mark of non‐authenticity, and (3) to develop a custom GPT trained to detect signs of fabrication.Methods: Three fictional studies were designed to compare treatment outcomes for a specific ocular disease. First, prompts were submitted to the custom GPT “Data Analyst” to produce 12 “unrefined” fake data sets for the three studies. Then, the custom GPT “Fake Data Creator” was developed to generate 12 “refined” fake data sets, specifically designed to evade authenticity checks. Forensic analysis was performed on all data sets using IBM SPSS. Finally, the custom GPT “Fake Data Detector” (FDD) was developed to replicate the forensic analysis and assign 1 point for each sign of fabrication.Results: Forensic analysis of the unrefined fake data sets revealed numerous flaws, most notably: name/gender mismatch (n = 12), non‐uniform distribution (n = 11) and repetitive patterns of last digits (n = 7), absence of correlation between study variables (n = 12), and distribution shape anomalies (n = 11). In refined fake data sets, the forensic analysis did not reveal statistical flaws, except for distribution shape anomalies (n = 7). Overall, 5 refined fake data sets (41.7%) passed forensic analysis as authentic. The results of the statistical analyses performed by GPT FDD were identical to those obtained with IBM SPSS, with a mean of 9.67±2.13 (95% IC 8.46‐10.87) and 1.00±1.00 (95% IC 0.43‐1.57) points respectively allocated to unrefined and refined data sets.Conclusions: Sufficiently sophisticated custom GPTs can perform complex statistical tasks and may be abused to fabricate seemingly authentic ophthalmology data sets, passing forensic analysis.References Taloni A, Scorcia V, Giannaccare G. Large Language Model Advanced Data Analysis Abuse to Create a Fake Data Set in Medical Research. JAMA Ophthalmol. 2023; 141(12):1174–1175. doi:10.1001/jamaophthalmol.2023.5162

GPT wars: Creating and detecting fake ophthalmology data sets with artificial intelligence

Aims/Purpose: Recently, our group demonstrated how the Large Language Model GPT‐4 can fabricate fake ophthalmology data sets, designed to support false scientific evidence. The objectives of this study were: (1) to identify peculiar statistical patterns of AI‐generated data sets, (2) to attempt at enhancing the quality of fabricated data for eliminating any discernible mark of non‐authenticity, and (3) to develop a custom GPT trained to detect signs of fabrication.Methods: Three fictional studies were designed to compare treatment outcomes for a specific ocular disease. First, prompts were submitted to the custom GPT “Data Analyst” to produce 12 “unrefined” fake data sets for the three studies. Then, the custom GPT “Fake Data Creator” was developed to generate 12 “refined” fake data sets, specifically designed to evade authenticity checks. Forensic analysis was performed on all data sets using IBM SPSS. Finally, the custom GPT “Fake Data Detector” (FDD) was developed to replicate the forensic analysis and assign 1 point for each sign of fabrication.Results: Forensic analysis of the unrefined fake data sets revealed numerous flaws, most notably: name/gender mismatch (n = 12), non‐uniform distribution (n = 11) and repetitive patterns of last digits (n = 7), absence of correlation between study variables (n = 12), and distribution shape anomalies (n = 11). In refined fake data sets, the forensic analysis did not reveal statistical flaws, except for distribution shape anomalies (n = 7). Overall, 5 refined fake data sets (41.7%) passed forensic analysis as authentic. The results of the statistical analyses performed by GPT FDD were identical to those obtained with IBM SPSS, with a mean of 9.67 ± 2.13 (95% IC 8.46‐10.87) and 1.00 ± 1.00 (95% IC 0.43‐1.57) points respectively allocated to unrefined and refined data sets.Conclusions: Sufficiently sophisticated custom GPTs can perform complex statistical tasks and may be abused to fabricate seemingly authentic ophthalmology data sets, passing forensic analysis.References Taloni A, Scorcia V, Giannaccare G. Large Language Model Advanced Data Analysis Abuse to Create a Fake Data Set in Medical Research. JAMA Ophthalmol. 2023; 141(12): 1174–1175. doi: 10.1001/jamaophthalmol.2023.5162

Engineering Premium Customer Acquisition: A Technical Framework for Value-Based Bidding Implementation

This technical article presents a comprehensive analysis of Value-Based Bidding (VBB) implementation frameworks and their impact on digital advertising performance. Through examination of data from 1,000 e-commerce campaigns encompassing 5 million customer transactions and $250 million in ad spend during 2023-2024, we demonstrate the effectiveness of VBB in premium customer acquisition. The article establishes a novel technical framework for VBB implementation, incorporating real-time data processing of over 1 million customer interaction points per second. Key findings reveal significant improvements across critical metrics: 92% attribution accuracy (a 14% increase), 85% cross-device matching success rate, and 73% improvement in offline conversion capture. The implementation of sophisticated customer segmentation strategies resulted in ROI improvements of 185% for high-value segments and 125% for medium-value segments. Furthermore, the article introduces advanced optimization algorithms that achieved a 23% reduction in customer acquisition costs while improving bid efficiency scores to 0.92. These findings provide a technical blueprint for organizations seeking to enhance their digital advertising effectiveness through value-based methodologies Value-Based Bidding (VBB) represents an advanced approach to digital marketing that focuses on maximizing return on investment. The system analyzes customer behavior and value to make smart decisions about advertising spend. Key performance indicators like Return on Ad Spend (ROAS – how much revenue you generate for each advertising dollar spent) and Customer Acquisition Cost (CAC – the cost to acquire a new customer) help measure success. The data shows impressive improvements, with ROAS increasing by 156% and CAC reducing by 23%, meaning more efficient spending and better results.

Demand for financial advice: Evidence from a randomized choice experiment

Demand for financial advice: Evidence from a randomized choice experiment

Advising Sophisticated Customers: Evidence from Health Insurance Brokers

Advising Sophisticated Customers: Evidence from Health Insurance Brokers

From months to minutes: Creating Hyperion, a novel data management system expediting data insights for oncology research and patient care

Here we describe the design and implementation of a novel data management platform for an academic cancer center which meets the needs of multiple stakeholders. A small, cross-functional technical team identified key challenges to creating a broad data management and access software solution: lowering the technical skill floor, reducing cost, enhancing user autonomy, optimizing data governance, and reimagining technical team structures in academia. The Hyperion data management platform was designed to meet these challenges in addition to usual considerations of data quality, security, access, stability, and scalability. Implemented between May 2019 and December 2020 at the Wilmot Cancer Institute, Hyperion includes a sophisticated custom validation and interface engine to process data from multiple sources, storing it in a database. Graphical user interfaces and custom wizards permit users to directly interact with data across operational, clinical, research, and administrative contexts. The use of multi-threaded processing, open-source programming languages, and automated system tasks (normally requiring technical expertise) minimizes costs. An integrated ticketing system and active stakeholder committee support data governance and project management. A co-directed, cross-functional team with flattened hierarchy and integration of industry software management practices enhances problem solving and responsiveness to user needs. Access to validated, organized, and current data is critical to the functioning of multiple domains in medicine. Although there are downsides to developing in-house customized software, we describe a successful implementation of custom data management software in an academic cancer center.

Application and Importance of Big Data in Insurance Industry

One of the most potent buzzwords in the insurance sector today is big data. Insurance brokers can use it as a beneficial tool to analyses trends and patterns to inform policy, evaluate risk, and spot fraud. Although the majority of insurance companies already utilize big data to some extent, the applications of big data continue to expand and change annually. Any insurer's playbook should start with an understanding of big data and how it can be used to provide sophisticated customers with effective, intelligent insurance products.

Application and Importance of Big Data in Insurance Industry

One of the most potent buzzwords in the insurance sector today is big data. Insurance brokers can use it as a beneficial tool to analyses trends and patterns to inform policy, evaluate risk, and spot fraud. Although the majority of insurance companies already utilize big data to some extent, the applications of big data continue to expand and change annually. Any insurer's playbook should start with an understanding of big data and how it can be used to provide sophisticated customers with effective, intelligent insurance products.

The impact of customer characteristics on exploitation and exploration capabilities: An empirical study of outsourcing service companies

There is a long history of examining exploitation and exploration capabilities as the key drivers of organizational learning, but the research that pursues the synergistic impact of various customer characteristics on these two capabilities of firms is yet to emerge. This article utilizes fuzzy set theory to develop a method that accounts for the combined effect of the three key characteristics of customers, namely customer involvement, customer sophistication, and customer geographical location on the exploration and exploitation capabilities of firms. Applying the proposed method to a sample of 86 Outsourcing Service Companies (OSCs) around the world, our findings highlight the disproportionate impact of customers' influence on the learning capabilities of four different types of OSCs: technology-based, engineering, knowledge-based, and business service firms. This empirical study makes a key contribution by way of developing a quantitative picture of the interplay between customer characteristics and the exploration and exploitation capabilities of OSCs, which in turn provides a measure of the extent to which interactions with customers enhance these two capabilities of firms. • An empirical study of the impact of customers on learning capabilities of firms • A diverse set of service firms around the world is studied • Fuzzy logic is employed to evaluate the combined effect of customers characteristics • The new method prioritizes customers based on their influence on learning capabilities

Development of an authentic concept of engineering activities based on product redesign

AbstractToday's job market is seeking engineers with competencies to design innovative solutions that meet sophisticated customer needs. Engineering education is then challenged to equip future engineers with holistic engineering design skills, especially functional ones. A powerful means to strengthen these skills is the use of reverse‐engineering‐based activities, which consist of examining, extracting information, and redesigning existing products. However, most current education endeavours, based on reverse engineering, consist only of practicing simple teardowns that have circumscribed impact on the acquisition of skills. Therefore, there is a need for more elaborated authentic hands‐on activities to gain a broad set of design skills. This study addresses this gap by the development of a concept of wide‐ranging engineering activities that start with the study of an existing product and ends with an improved redesigned three‐dimensional (3D) printed product. This concept of activities was developed to strengthen a conventional course on product design. Thus, a tailored comprehensive redesign process is proposed first, and expanded as a concept of a set of experiential activities, with associated measures for skills acquisition. This concept encompasses teardown, 3D digitizing and rapid prototyping, and aims mainly at facilitating the understanding of components' functionalities, the numerical reconstruction by 3D digitizing, the mechanical modelling and engineering analysis of parts and finally the 3D printing of the redesign output. To understand, experience, and weigh up the relevance of the proposed concept of activities, a preliminary implementation, and a case study are illustrated. Particularly, the relevance of the concept is demonstrated through the assessment of the activities' measures. In short, this study provides educators with an authentic education tool that leverages on a broader reverse engineering vision to boost the job's sought‐after design skills.

Matching Supply and Demand in Collaborative Additive Manufacturing.

Due to an increasing individualization of products, additive manufacturing is often seen as a solution to cater for more sophisticated customer requirements. In order to fulfill customer needs, manufacturers have to rely on collaboration to distribute risk and improve the utilization of their resources. In this paper, we used qualitative interviews to define requirements for a marketplace that allows the automatic exchange of additive manufacturing capacities. From these requirements, we derived a conceptual model that matches orders to sales offers while taking specific product requirements, such as quality, into account. Additionally, we implemented a demonstrator to evaluate the model with potential buyers and sellers of additive manufacturing capacities. Our research showed that most requirements could be implemented in a marketplace. However, we could show specific limitations for particular requirements.

Sustaining User Experience in a Smart System in the Retail Industry

Retail enterprises are embracing new technologies to provide innovative services to customers and engage them. Unmanned retail stores offer a completely new seamless shopping experience for customers. Moreover, artificial intelligence (AI) technology and sophisticated customer behavior should be explored in depth to develop a smart system to serve customers. This study focused on achieving a sustainable user experience with a smart system installed in an autonomous store. The development of core functions and a rule-based knowledge set are regarded as the most important tasks in designing autonomous services. In the case study, the core functions were developed on the basis of the design science concept and the rule-based knowledge set was constructed using the action research approach. The developed smart system was optimized to understand in-store customer behavior by continuously observing and refining the user experience. The practical experience of this study provides insights into AI technology use in retailing and can guide enterprises in developing smart systems.

Role of Innovation Capability on firm performance in the context of Commercial Banks in Nairobi City County, Kenya

The Kenyan banking industry is characterized by increased environmental dynamism brought about by rapid technological development, customer sophistication and regulations. The achievement and sustenance of competitive advantage is important as any organization that does not gain and sustain this may not survive. In order for an organization to remain relevant, it must be able to leverage on innovation capability for effective adaptation to changes in the environment. This study therefore sought to assess the influence of innovation capability on performance of Commercial Banks in Nairobi City County, Kenya.The study was anchored on dynamic capabilities theory and resource-based view, Resource. Descriptive research design was used for this study. The target population was Commercial Banks in Nairobi City County. The unit of analysis was the commercial banks while the unit of observation was management. Data was gathered using self-administered questionnaire then analyzed using Statistical Package for Social Sciences (SPSS Version 25.0) software. Descriptive statistics such as frequencies, percentages and mean and inferential statistics such as correlation coefficient was used to analyze quantitative data. Charts, tables and graphs will be used to present the data. The study established that innovations capability positively and significantly affects performance. The study revealed that innovation capability is an important requirement for the effective management of inventions and creativity and the introduction of transformative technologies. The study recommends that in an era of faster competition and digitization, commercial banks in Nairobi City County, Kenya must increase their innovation capability. New innovations easily adopted in organizations enables banks to adjust to meet new customer needs and market changes. Innovation also enables creation of new products and services from time to time .Innovation capability of an organization affects organizational performance as it determines organization’s productivity and effectiveness.

EMPIRICAL ANALYSIS OF DEPOSIT MONEY BANKS CUSTOMERS� SOPHISTICATION IN YOLA, ADAMAWA STATE, NIGERIA.

Purpose � the purpose of this paper to investigate customer sophistication of deposit money banks in Yola, Adamawa state, Nigeria.Design/methodology/approach � this study adopts the quantitative survey method with a sample of 375 deposit money banks customers from Yola metropolis using purposive sampling technique. Similarly, PLS-SEM approach was used in data analysis to confirm validity of the model and subsequently analyze the structural model for obtaining the path analysis.Findings � Relationship between search orientation, and market experience are approximately equal in terms of positive strength, while innovativeness of customers emerged as the least in terms of strength in positive relationship. Customers are therefore less innovative when it comes to general knowledge and experience in evaluating the prospects of different options. Brand value orientation is also a strong determinant of customer sophistication because respondents are more attached to brands they are more familiar with than taking brand of services that are new and becoming more risk averse. Customer sophistication in general to the respondents indicates moderate satisfaction with services and low self-image and esteem.Practical implications � customers are less sophisticated because they lack innovativeness and stick to brand value orientation, they don�t realize full satisfaction, self-image and attainment of self-esteem as an experienced customer. Therefore, deposit money banks need to consider the findings of this study to improve customer sophistication.Originality/value � this study is the first of its kind in the contest of the study.

Examining guanxi as an environmental-dependency-coping strategy in China: A 3PL provider’s perspective

ABSTRACT As firms seek to meet sophisticated customer needs, they typically become more dependent on outside partners. Therefore, managing environmental dependence has become a key competency for business success. This observation is particularly true for third party logistics (3PL) providers in China, as most of these providers are lacking in resources. Drawing on resource dependence theory, this study examines whether and when guanxi – a prevalent relationship management approach in China – is necessary for 3PL providers as a strategic option to manage environmental dependence. Based on survey data collected from 132 3PL providers in China, the empirical results show that it is necessary for 3PL providers in China to develop and rely on guanxi to successfully cope with environmental uncertainty and dependence. However, the necessity of using guanxi to deal with these challenges differs between business relationships. The necessity of relying on guanxi to deal with environmental uncertainty is alleviated by the presence of formal, customised legal contracts and by the frequency of contact in 3PL relationships. However, the need of using guanxi to cope with dependence tends to increase with the duration of each relationship. The managerial and theoretical implications of these findings are discussed.

Evidence of customer sophistication behaviour in deposit markets: the case of Qatar

PurposeThe paper investigates the interest rate policy transmission mechanism and the role of market structure of the banking industry in Qatar.Design/methodology/approachCompetitiveness indexes are used to measure the degree of market power in the banking industry in Qatar. The momentum threshold autoregressive model is applied over the monthly period from January 2005 to June 2018 to examine the magnitude of intermediation and adjustment to disequilibria in the deposit market. In addition, to model interest rate volatility and overcome the problem of heteroscedastic errors in the error correction standard models, an asymmetric EC-EGARCH-M model is applied.FindingsThe findings suggest incomplete pass-through and asymmetric response to monetary shocks. The asymmetric adjustment mechanism is found to be downward rigid which suggests a high degree of customer sophistication and an elastic supply of deposits. The results of the EC-EGARCH-M show that the impact of monetary policy shocks has a significant positive impact on deposit interest rates and that negative monetary shocks trigger more conditional interest rate volatility in the next period than positive monetary shocks for a short maturity rate.Originality/valueThe paper is the first to highlight the behaviour of the interest rate pass through channel and measures the degree of competitiveness of the banking industry for the case of a small, rich country. In addition, using recent data, the paper applies different econometric methodologies and overcomes the problem of heteroskedastic errors by modelling the interest rate volatility using the EC-EGARCH-M model.

The European Union’s GDPR and Its Effect on Data-Driven Marketing Strategies

This research paper analyzes the developing effect that the European Union’s (EU) recently developed General Data Protection Regulation (GDPR) will have on the marketing strategies of firms that rely on big data. Big data is identified as consisting of data and data analytics involving a huge volume of data, a diverse variety of data, and a high velocity of data capture and collection. This analysis begins with some discussion of the concept of big data and follows this up with overviews of both the GDPR and big data use in the marketplace. The EU replaced its older Data Protection Directive or DPD with the GDPR. The GDPR consists of a series of chapters and articles that require, among other things, consent to collect and store data, the anonymization of data, announcement in 72 hours of a data breach, provision of encryption and the identification of a Data Protection Officer. Marketing and the marketing function can implement emergent technologies that augment big data and its analysis while simultaneously achieving compliance with regulatory frameworks like the GDPR. These marketing related solutions are those such as blockchain marketing applications like Brave Browser and Blockstack among others. The report also examines the way in which enterprises use big data in their marketing strategies and how they are affected by it now that it has come into effect. Some of the more marketing-oriented uses and applications of big data are found in sophisticated loyalty programs, demand forecasting and customization either of experience or product/service. This study also offers some final recommendations related to GDPR compliant marketing strategies. These include the development of a comprehensive program to purchase consumer data directly from consumers and the introduction of blockchain as a means to facilitate a smoother transition to GDPR compliance.

Managing arrival variability in healthcare services: Case of an out-patient department

Similar to other services, healthcare is also experiencing an evolution of patients into sophisticated customers with high expectations from the service providers. This has led to the application of operations research tools such as discrete event simulation modelling in delivering improved healthcare services. The present study is conducted in an obstetrics and gynaecology out-patient department of a multi-specialty hospital. Process related data is collected from the department and is used to build a discrete event simulation model. Scenarios are developed considering strategies suggested in literature to manage arrival variability, the most crucial form of customer-induced variability. Based on the results obtained, the most optimal solution is highlighted. This study shows that with recent research advancements in simulation modelling, scepticism towards methods adopted from other sectors can be reduced which is beneficial for both, the modern-day healthcare service provider and the patients.

Customer Service Quality Practice in District Co-Operative Bank - with Reference to Thiruvananthapuram District Cooperative Bank

As the entry of sophisticated technology has paved way for greater developments in every field including the banking sector, the banking sector too is characterized by rapid change and increasingly sophisticated customers. It has become very important that banks in India determine the service quality factors, which are pertinent to the customer’s selection process. With the advent of international banking, the trend towards larger bank holding companies, and innovations in the marketplace, the customers have greater and greater difficulty in selecting one institution from another (Choudhury, K. (2018)). Therefore the current problem for the banking industry in India is to determine the dimensionality of customer- perceived service quality. If service quality dimensions are identified, the service managers should be able to improve the delivery of customer perceived quality during the service process and have greater control over the overall outcome. A better understanding of the effects of service quality and customer satisfaction on service loyalty can help academics in the development of a model of service marketing. It can also provide practitioners with indications as to where best to devote marketing attention and scarce corporate resources.

Using Jovanovich’s Learning Effects Model to Predict Advantageous Product among Micro and Small agro-food manufacturing Enterprises in Kenya

Advantageous product is a product with a competitive edge over others in the market. It attracts more revenue for the enterprise, fulfils sophisticated customer demands and is highly differentiated. In this paper, enterprise’s size and age influence on advantageous food products for local and export markets are examined. Formality is considered as a moderating factor. The study sampled 132 food-manufacturing Micro and Small Enterprises (MSEs) in Busia and Nairobi Counties. The researcher used Jovanovich’s Learning Effect Model to inform the conceptual framework. A seven Likert Scale questionnaire was used to measure the MSEs’ gradation in opinion, attitude and behaviour on size, age, registration and how they influenced their advantageous product. Research instruments’ reliability was tested by chronbach’s alpha and realized 0.97 which was greater than 0.70 showing excellent internal consistency . Due to weak information management system of agro-food processors in Busia county, snowballing sampling techniques was used and fisher sampling techniques formula at standard normal deviate of 1.96 on Nairobi County Government given its numerous food manufacturing enterprises. The data was analyzed using moderated Logit model at a corresponding 5% level of significance using Statistical Package for the Social Sciences (SPSS) version 21.0. The study found out that enterprise size, age and registration status (as a whole) significantly influenced agro-food processors to produce advantageous product ( wald (1) = 41.297 , p=.000, sig<.05, 2-tailed) but not as independent stand alone factor. The study recommends census of all agro-food manufacturing enterprises in Kenya. It also suggests combination of size and age factors in the registration and development of the enterprises. Finally the study recommends establishment of a food and beverage administration authority to support MSEs for export market.