E-commerce provides an excellent opportunity to increase the company revenues, especially during the COVID-19. Many features that appear in the E-commerce platform can influence the customer purchase intention. Therefore, most companies strive to achieve the highest maximum Star-rating that serves as the electronic version of word of mouth. CV XYZ expects to increase the 5-Star-rating count by creating the after-sales service program. However, does it pay to be perfect? By collecting Star-rating data from six online stores across three different online marketplaces organized by CV XYZ, we develop simple Statistics to a more sophisticated Markov Chain and Regression to analyze the data. The result shows that the distribution movement of Star rating before and after implement the program is not significantly different. However, the after-sales service program greatly improved the average daily Sales around IDR 31M (p-value for the T-test = 0.0039). By simply using Pearson Correlation analysis, we found a negative correlation between Sales and Star-Rating counts that encounter us in the “too good to be true” phenomenon. Therefore, it is more critical for the company to figure out an optimal number of 5-Star ratings achieved to maximize Sales than only focus to achieve the highest 5-Star ratings.