Xu Jinglei at the launch of her e-magazine in April this year — thinking about the numbers?

Today Fang Jun on the excellent group blog Mind Meters published an article about the figures currently being tossed about by China's e-magazine industry. The article is roughly translated below.

Note: In China e-magazines (电子杂志) refers to electronic magazines that are downloaded and then viewed by turning virtual pages, sometimes containing multimedia content.



E-magazines often say that they have several million readers. For example, Xu Jinglei's Kai La magazine claims that it has been downloaded 3 million times.

I have always been a bit doubtful; I know that figures for circulation of media in China have always been grossly exaggerated. The usual practice is to divide a newspaper's stated circulation numbers by two or three, and to divide a magazine's stated circulation numbers by five to ten. 

In the bubble of e-magazines, how do we derive the real circulation numbers? Do we divide the stated numbers by ten or by 100?

I am very curious about this so I have asked many people about the circulation numbers for e-magazines, but as you can imagine, this is an industry secret, and it's very difficult to get at the truth. But guesstimating from the information I got hold of, in general you can say that dividing the stated numbers by ten is not nearly enough because e-magazines use what Dilbert calls 'weasel math'.



What is weasel math? Here are two examples:



For example, more than three people visit the California Flannel Pocket Museum every year, but you could say that the museum is in a state that attracts more than 12 million visitors a year. 



And how does a new cable TV station tell people how many viewers it has? Well, they will tell guests to their talk shows that '25 million people can watch you on this show'.

According to data from the end of 2006 and the beginning of 2007, three big e-magazine platforms have released these figures: xplus says they have 11 million users, ZCOM says more than 20 million, and POCO says they have more than 40 million. Without even mentioning that how these figures are padded, the statistics are quite clearly exaggerated: anyone who has downloaded their e-magazine reader software even once is counted as a user even if the person never uses the software again.

To speak specifically of individual e-magazines, generally they use this calculation: e-magazine platforms that offers several different publications for download usually use 'push' technology, so a certain proportion of all the users becomes the figure for an individual magazine's 'readership'. But how many people download the magazine of their own initiative? How many people actually read the magazine? Nobody knows these figures. Of the stated three million readers of 'Kai La', how many people actually read the magazine? 300,000? 100,000? Or 30,000?



iResearch [an Internet research firm] has figures about e-magazines, but they are also calculated using weasel math. For example, one of their reports in 2007 predicted that 30 to 40% of all Internet users in China were e-magazine users based on the 'user numbers', so the figures are huge. You know, there are already more than 140 million Internet users in China, and this figure is set to increase many times in the next few years. 



(Note: Industry insiders are welcome to submit their calculation methods for readership numbers. Thanks.)