Marketing Strategy Analysis of BYD Using 7P Theory and the Large Language Model

Abstract

As new energy vehicles (NEVs) have become more popular in recent years, many brands have started to focus on this market. With China emerging as the world’s largest market, BYD (Build Your Dreams) has been the top-selling brand for nine years in this highly competitive space. Staying in first place for so long in such a competitive industry is impressive, and the marketing strategies behind this success are worth exploring. At the same time, BYD offers a wide range of vehicle models with different configurations, but why do customers prefer one over the other? How can the business observe its customers to achieve better sales? This paper analyzes BYD’s marketing strategy and target customer group using the 7P marketing theory. The 7P framework expands the traditional 4P model to include seven key elements: product, price, place, promotion, people, process, and physical evidence. I integrate online comments from different websites, including BYD’s official website and other vehicle-related websites. Through theory-driven analysis, the conclusion is drawn: the most important dimension is the product, and price comes second. This means BYD outperforms other electric car manufacturers because it has achieved a very good balance between quality and price. These two dimensions have helped BYD become the top new energy vehicle brand in China. More importantly, to further testify the above conclusion, this paper uses the latest artificial tools. With the use of a large language model, data-driven analysis is designed, including generating surveys and simulating consumers. First, five representative BYD users were identified by inputting data on BYD’s top five best-selling models into the large language model (LLM) using appropriate prompts. Then, the LLM acted as these five users from 34 provincial-level administrative divisions in China to answer the designed survey. The LLM prompts incorporated real consumer reviews from BYD’s official forums to enhance simulation accuracy. The survey answers were collected and analyzed to get the final results. Some interesting conclusions were drawn from the data-driven analysis: some are consistent with the above findings, and some are new, which complement the theory-driven analysis.