Gutenberg 3.5 eBook Warez Report: Persisting Questions and Inaccurate Estimates

For several years now, the so-called “Gutenberg” report has been released to evaluate the situation in the eBook warez scene. The topic received special attention last year after the now-defunct platform boox.to became well-known to a broad audience through its public relations.

In this context, one of the authors of the Gutenberg report questioned the traffic data provided by the portal and attempted to discredit it using Alexa and IVW data. Why this method was doomed to fail from the start, I explained in a separate article. The subsequent Gutenberg Report 3.4 wisely left Alexa out as a source of comparison and replaced it with another service called “Similarweb.” A few days ago, the new Gutenberg 3.5 report was released, and it also raises a few questions.

More Accurate, but Not Exact

Unlike Alexa, where data is collected only via a browser plugin, Similarweb uses a variety of sources to estimate website traffic, according to their own statements. And here we reach the same crucial point as last time: These data are also only estimates and not an actual statistical image. Similarweb’s data are largely calculated from Google rankings, meaning the total traffic volume can be estimated based on the search engine’s ranking for different keywords.

This method is at least more accurate for Germany than Alexa’s measurements. However, there are also a number of unknown factors that the report should make clearer, as even a brief comparison of Similarweb data with Google Analytics data from ALLESebook.de shows there are (unsurprisingly) inaccuracies. This starts with the traffic volume, extends to the traffic sources, and ends with the countries of origin of the visitors. Similarweb best estimates these things, although the ratio of various traffic sources (at least for ALLESebook.de) is simply incorrect. The traffic data is consistently off by about a third each week. There is also the question of how significant the measurement fluctuations are across different websites.

Since the Gutenberg report also claims to be a study, these inaccuracies should be clearly identified, and the data should not be presented as supposed facts. Especially considering that illegal offerings may be accessed via proxy servers (or in the case of boox.to, via the Tor network) or the browser’s private mode, the traffic data could deviate even more than for our homepage. There are no empirical values here, so the numbers should be taken with a grain of caution.

Forum Clicks for Trend Measurement

Another point that needs closer inspection is the estimates based on clicks on a forum post in a well-known warez forum. The (understandably) constantly growing hit statistics of an eBook thread in the relevant forum are used as the basis for evaluating the increasing popularity of the illegal eBook offering. However, if you take a closer look at the annual development of the numbers, the “central role” attributed to the warez forum in the Gutenberg report, among others, based on the mentioned forum thread, doesn’t seem to be as significant.

How is the demand in a large warez forum?

In January 2011, the thread’s hit statistics were first recorded in the Gutenberg report. Growth accelerated in 2012, but then weakened somewhat by 2014.

• 01/2011: 1,049,950 hits
• 01/2012: 2,150,055 hits (+1,100,105; average +91,675 per month)
• 01/2013: 3,824,116 hits (+1,674,061; average +139,505 per month)
• 01/2014: 5,353,516 hits (+1,529,400; average +127,450 per month)

A similar picture emerges when looking at the number of posts and threads in the eBook subforum.

So at least here, the horror scenario of exploding eBook warez demand cannot be proven. Rather surprisingly, the thread’s opening rate declined in 2013, even though the eBook market grew by 60 percent according to GfK, and Thalia and eBook.de reported even 80 percent growth. It seems a bit different with book pirates. Here, based on the warez forum, a decrease in demand is indicated—if you want to take these (reliable) statistics as indicators for that—as is done in the Gutenberg report. Against this backdrop, the estimated traffic volume of the warez forum, with a reported 9 million visits per month, appears relatively insignificant. This is also reflected in the number of posts: The eBook section of the forum currently has only 41,155 threads and 95,009 posts. The entire forum has 1,004,763 threads and 5,135,356 posts. So, only about 4.1 percent (measured by topics) or about 1.85 percent (measured by posts) are attributable to eBooks.

It’s also worth mentioning that one of the previously mentioned “hits” by no means equates to a download. The hit counter increases with a simple click on the thread, also through various search engine crawlers. As it is an ongoing offering, numerous repeat visitors are undoubtedly present, who are counted multiple times. Therefore, it is impossible to estimate how many illegal downloads ultimately occurred.

Doubtful Estimates

All in all, these estimates seem pointless to me. But the biggest faux pas, in my opinion, is the assumption that for every purchased eBook, ten are downloaded illegally. Especially when you consider that, according to GfK, about 20.4 million eBooks were sold in 2013 (revenue of 160 million euros with an average price of 7.85 euros).

Gutenberg 3.5: For every purchased eBook, there are ten downloaded? Hardly…

If for every purchased eBook there are ten illegal eBooks, that would be a total of an (incredible) 224.4 million eBooks. Not even legally downloaded free eBooks are considered, which undoubtedly make up a significant part due to the large public domain offer. Annually, however, “only” around 380 to 390 million books are sold in the German book market. In my opinion, these numbers don’t really stand in a credible relationship, because if that were truly the case, the German book market would have significantly collapsed in recent years. In reality, industry revenue has been between 9.5 billion and 10 billion euros annually for years. If illegal eBook consumption now constituted more than half of the regular offerings of the entire book market, things would look quite different. But, if one insists on these numbers, then one can only conclude that eBook piracy isn’t causing any real harm.

Either way: In my opinion, this is fear-mongering. This does not mean, of course, that you should ignore the development of eBook piracy, but the situation is probably not nearly as dramatic as depicted in the Gutenberg reports. In fact, the Similarweb estimates and warez forum statistics don’t indicate any significant change.

Primarily, the book market needs to continue creating great offers for users so that they don’t even consider looking into illegal forums in the first place.