A Systems Analysis Approach for Business Optimization: Integrating Technology Development with Data Analytics and Marketing for GolfCask; Analysis on the Ethical Issues Surrounding Netflix’s Recommender System

GolfCask, a newly established technology-based start-up in Charlottesville, Virginia, is dedicated to cultivating a vibrant online community centered around the shared passions of golf, travel, and whiskey. Employing a systems analysis framework, this project leverages data-driven insights to refine performance indicators and enhance system efficiency within the online community. The strategic objectives, initially structured within an objective tree, prioritize establishing a sustainable and profitable business model, fostering strong community engagement, and attracting and retaining members. A key component involves designing, validating, and deploying an innovative recommendation system to match user profiles with customized whiskey suggestions. Additional tactics in marketing and data processing are implemented and guided by data analytics and visualization tools to strengthen the technology development and enhance user engagement. Research on user acceptance testing and data integration supplies further insight into developing a user-centric design. Incorporating feedback loops and continuous data analytics refines system outputs, ensuring the recommendations and marketing strategies align with user preferences and business objectives. Integrating a comprehensive system analysis with personalized recommendations and marketing strategies, this project seeks to evaluate the effectiveness of these approaches and provide actionable recommendations for GolfCask’s future technical and business developments.
Because understanding recommender systems was such an integral protion of the technical aspect of this project, the STS portion analyzes many different aspects of these systems. Recommender systems are disgusined within daily life in different platforms like Netflix, Youtube, and Tik Tok. These recoemmendr systems are customized to fit user preerences and profiles in order to suggest the optimal viewing content for said user. There have been lots of ethical concerns raised regarding these recommender systems and a lot of these problems stem from data privacy, algorithmic bias, user manipulation, and corporate transparency. Throughout these sections, we will explore the evolution of recommender systems from the early stages of manuel heuristics to the today in which recommender systems incorporate AI and self iteration. Utilizing methods like ethnography and analysis of previous literature and publication I was able to uncover what active users of Netflix knew and thought about their recommender system as well as background research on recommender systems as a whole. After utilizing these methods I decided to use Actor Network Theory to deep divulge into the relationships between Netfix Users, Netfllix Developers, Netflix Executives, and Netflix’s actual system itself. There was focus on the ethical implications and worries about where these systems get their numbers, data, and statistics from and how much their users actually know about how the company is analyziing them. Looking at both of technical aspects and social aspects it is concluded that users as whole should know what is behind recommender systems and how they work as well as the fact as to the issues that surround them.
The two projects relate to eachother in many different ways. For starters, my STS portion of my project was to analyze all different aspects of recommender systems. This went from how to build one to the history of them to the ethics behind them and how they shape society. The technical portion of my project was essentially a consulting position for a recent Charlottesville startup named GolfCask that focuses on building a community of both whiskey and golf enthusiants. As I said, the company is a startup and the goal is to build a community between people that appreciate whiskey. The entire company was yet to have things like user profiles yet an actual recommender system. We were taked with designing aspects that we thought would be key in order to desing a user profile. From these user profiles, we took ratings from the founder of the company to develop a data base of rankings of whiskey. With this data base, we were able to develop and create a recommender system from the resrarch that we had done as to how to create one. We compared different similariti4rs across different user profiles and even iterated and adjusted the system as a whole to improve our recommendations to make sure the vector similarities were most matched with the data base that we were given. Without the initial research that we had done and all of the information that we had composed regarding recommender systems, we would not have been able to develop a successful recommender system for GolfCask regarding Whiskey suggestions.

School of Engineering and Applied Science

Bachelor of Science in Systems Engineering

Technical Advisor: Matthew Burkett

STS Advisor: Joshua Earle

Technical Team Members: Christian Hooper, Elena Johnson, Shreya Malani, Alexander McCall, Matthew Tan