China keeps on breaking records. The latest count, according to the 25th report of the CNNIC (China Internet Network Information Center), shows that the country had 384 million internet users at the end of December 2009, a user rate of some 28.9 percent of the Chinese population.
Over the last few years, the value of the ecommerce market in China has been estimated at over 3 trillion RMB (around 325 billion Euros) with a yearly increase of 41.7 percent.
The present study focuses on computer sales sites in China. This business sector has witnessed particularly aggressive competition, paving the way for a whole new ballgame.
Do the Chinese websites of computer manufacturers operate like clockwork, or can they cause a decrease in turnover?
By way of an answer, ip-label.newtest measured the performance of the websites of the six leading computer manufacturers established in China. Measurements were made in an end-user environment --the way an internet surfer would perceive it -- from 12 of the largest Chinese cities*.
Sharp contrast in the performances of online showrooms
The table below is built on a 100-point index on the basis of two indicators: the rate of successful connection, and how fast the homepage of each site loaded.

With 82.5 points, the website of Dell China shows the best performance. The site of this American giant's Chinese subsidiary has Lenovo following close on its heels.
The differences in performance are very pronounced. 360buy, coming in fourth, has a score barely higher than 70 points. TaoBao and HP, with 59 and 41 points respectively, lag even farther behind.
What is the reason for such poor scores? Even Dell and Lenovo, whose performance outdistances that of their competitors, have been awarded rather mediocre scores.
Complex exchanges

Analysis of each of the five phases required for full downloading of web pages for the main Chinese online computer shops indicates that practically the entire time is spent retrieving objects that make up the page. The phases of connection and first data received barely count at all.
Only Dell China, and especially HP China, need more "prep time". For Dell, the interval between connection confirmation and first data received is relatively long. This phenomenon can be explained by the probable creation of a specific environment designed to save time subsequently, when the visitor browses the site.
For HP, on the other hand, the first three phases take a long time. The slowness of the resolution of its domain name (DNS) might be due to poor configuration of data communicated to the domain name resolution servers.
As for the connection's sluggishness, this could be caused by an unfavorable geographical location, or by the choice of connection to a network operator which interconnects poorly with other networks, generating slowdowns in exchanges between the browser and the website. The great number of objects to load amplifies the bottleneck phenomenon.
*Measurements made from Beijing, Chengdu, Fuzhou, Guangzhou (Canton), Hangzhou, Nanjing, Qingdao, Shanghai, Shenyang, Shenzhen, Zhengzhou, and Xi'an
Next page >>