Is your product description having any impact on the customer’s decision to buy a product? A Study done in University of Standford by Prof.Reid Pryzant and team examines the relationship between the language used in product descriptions and their sales performance. The study uses machine learning algorithms to analyze text data from the product descriptions of over 50,000 products on an e-commerce website.
The study found that the language used in product descriptions was significantly related to sales performance, with certain linguistic features being more predictive of sales than others. The authors identified several linguistic features, such as sentence length, use of adjectives and adverbs, and the number of technical terms, that were predictive of sales.
One of the key findings of the study was that the use of emotional language in product descriptions was strongly associated with higher sales. Specifically, the use of positive emotional language, such as words indicating joy, trust, and anticipation, was positively correlated with sales, while the use of negative emotional language, such as words indicating sadness, anger, and fear, was negatively correlated with sales.
Another interesting finding of the study was that the length of product descriptions was also predictive of sales. Specifically, product descriptions that were neither too short nor too long, but rather fell within a certain range of sentence length, were associated with higher sales. This suggests that there may be an optimal length for product descriptions that strikes a balance between providing enough information to potential customers while not overwhelming them with too much detail.
The study provides valuable insights into the language of product descriptions and its impact on sales performance. By analyzing the linguistic features that are most predictive of sales, e-commerce businesses can optimize their product descriptions to increase sales and improve customer engagement.
Additionally, the study highlights the potential of machine learning algorithms for analyzing large amounts of text data and extracting valuable insights. As e-commerce continues to grow and evolve, these insights will become increasingly important for businesses looking to stay ahead of the competition.
For a detailed account read the original paper.
Pryzant, R., Chung, Y., & Jurafsky, D. (2017). Predicting Sales from the Language of Product Descriptions. eCOM@ SIGIR, 2311.
