Understanding Chinese online users and their visits to websites: Application of Zipf's law

• We collaborated with an online marketing research agency to obtain a bundle of longitudinal sectional web surfing data.
• We examined the Zipf's law of web usage behaviors in the biggest Internet market, China.
• We identified that 94.87% of the entire Internet traffic concentrated in 15.08% of all visited websites.
• The approach of the observed Internet traffic remained salient after controlling the user segmentation.
• Both website operators and normal Internet users can get benefits from the revealed Internet traffic distribution.

Competition for consumers to visit company websites has intensified in recent years. An important indicator of website popularity (and consequent survival) is the extent to which the website can draw consumer visits vis-à-vis other websites. A majority of the current understanding on consumer visits is limited to a single website, and leaves little knowledge on the performance of one website compared with others. In tracking the Internet usage behavior of 200 individuals in Mainland China for 30 consecutive days, we applied Zipf's law to identify the divergence points separating popular websites from non-popular ones. Two measurements were used, namely, visit traffic (number) and visit engagement (time spent). We observed that 94.87% of the entire visit traffic is devoted to 15.08% of all visited websites, whereas 84.63% of engagements are on the top 6.16% visited websites. These findings suggest that few websites accounted for the bulk of online traffic and time. Further, we segmented the dataset based on two key proxy variables of user demographics, which are gender and occupation. The findings on visit traffic remained salient after considering user segments, but the findings on website engagement varied across different user segments. Our further analysis, which categorized the visited websites by their main service, revealed the type of Internet users attracted to popular websites.