The current online reputation systems for online sellers face great challenges from bad-faith behavior such as malicious negative reviews, click farming, mismatch between images and commodities, and forged commodities. To optimize the design of online reputation systems, explore the consumer utilization of credit clues, and describe the law of mutual trust, this paper puts forth three hypotheses about the influencing factors of consumer perception of online seller credibility and integrates various research methods such as an eye-movement experiment, questionnaire survey, econometric analysis, and empirical research. To evaluate the three hypotheses, the display modes of commodities on a current e-commerce platform were optimized, and eye-movement experiments were conducted on original and optimized webpages. Results show that the display of sales growth, the refinement and tagging of review content significantly impacted consumer perception of seller credibility. Further, designers of online reputation systems were advised to display sales trends, provide personalized sales queries, and tag a variety of reviews for consumers to easily ascertain credible sellers. This advice helped curb bad-faith behavior.