What Book Bloggers Read: A New Analysis with Visualizations and Statistics

Since my first steps as a book blogger, in addition to blogging about literature, I’ve always been motivated to get to know the entire blogosphere around book blogging more closely. In one post I analyzed how book blogs are networked with each other, which also led to the Top List. In another evaluation, I looked at where book bloggers come from. Now I want to take a closer look at the mysterious creature that is the book blogger and take a look at what book bloggers actually read. There are quite a few assumptions and prejudices floating around the blogging community, and since this is about gaining insight into reading behavior, it’s a particularly interesting topic. Some clichés are confirmed, others are surprising. I found this excursion very exciting and, in this post, I want to describe a bit of what I was able to find out about book bloggers and their reading tastes—once again using a few scripts.

When I started this post, I wasn’t aware how much information could be extracted from the collected data. This is probably the most extensive evaluation so far, but also a very interesting one, because some figures sharpen the picture of book bloggers and offer very interesting insights. To make the post easier to read, I tried to group the results and structure them with subheadings. I also considered starting a series of posts on these topics, but somehow it all belongs together, so it shouldn’t be artificially torn apart. I therefore ask for your indulgence and hope I can spark the interest of some readers. If this is all too much, you’ll find a summary at the very bottom in the conclusion.

Some of the evaluations were a bit more complex. I tried to validate and double-check all the figures again, as far as possible. But it is entirely possible that there’s a mistake hiding somewhere. If something seems odd to you, please ask and I’ll check it again. In several steps I tried to exclude errors as much as possible and also checked the data on a random basis.

Data Source

Before you can analyze data, you first have to collect it. The questions are: which books do book bloggers read, which genres interest them, and how do they rate their reading? First, I considered simply crawling all book blogs and parsing the books presented there. That, however, isn’t easy. Some dutifully list the ISBN, which is quite easy to recognize and store. But many (myself included) don’t. Or the ISBN numbers are contained in image files along with other data about the book. In addition, loading and reading all pages is a lengthy undertaking. Discovering books based on titles is similarly difficult and probably only possible with machine learning methods. And then there’s the problem with ratings. Some use owls, many use stars, others use fairy-glitter-dust-balls. A simple and pragmatic solution was therefore to capture the LovelyBooks and Goodreads accounts. There, the information is structured, and loading and parsing can be implemented in just a few lines.

I did this for all the blogs on the Top List and 500 blogs (40% of the blogs listed on the Top List) have a maintained and accessible LovelyBooks or Goodreads account. That’s only a subset, but extrapolating from the 40% to all blogs is certainly a legitimate approach. Of course, it’s possible that there is a correlation between the use of LovelyBooks and preferred genres and books. Within a certain range this is surely the case, but the answer is not a clear yes or no; it lies somewhere in between. This is an error I consciously accepted for feasibility reasons—and when I look at the results, they align in many places with subjective impressions, so the outcomes for all blogs can’t be too far off.

The LovelyBooks and Goodreads users are very diligent, and on average each blogger has reviewed 382 books there. In total, the evaluation includes 64,950 books, which were read a total of 190,913 times. A solid foundation that, in my estimation, ensures a representative evaluation. However, note that not every one of these books was also reviewed on the blog. Some may have only shelved and rated the books there. But this analysis is intended to examine reading behavior rather than blogging behavior.

A small interesting side finding is the very low need for expression among male bloggers. Only 4% of bloggers represented on LovelyBooks and Goodreads are male. Overall, 8% of book bloggers are men. So if male bookworms are looking for a partner, they should give LovelyBooks priority over Parship and similar platforms, because this is primarily a space for women who also happen to love reading.

Genres

For the selection of genres, I followed LovelyBooks and adopted their categories. They’re neither too fine-grained nor too coarse. Books I parsed from Goodreads are not assigned to any category there. Using the Amazon PartnerNet API, I fetched the missing genres and then mapped them by hand to the LovelyBooks genres. Amazon offers an excellently maintained database with very fine categorization and allows 3,600 requests per hour to their data. It’s a source I will definitely keep an eye on, especially since there are a number of ready-to-use programming libraries and access is free.

As a starting point, I first want to show the quantity distribution of genres.

In first place, with a considerable lead, are Young Adult books, the most popular reading among book bloggers. 24% of the books read are Young Adult. This matches the first impression you get when you go on a foray and browse through various book blogs. Novels are at 19% in second place. Fantasy is in third place and accounts for 15% of the literature read. Crime & Thrillers are also very popular at 14%. The boundaries between Young Adult, Novel, Fantasy, and Romance are probably not always sharply and clearly delineated. For example, there are young adult books with fantasy elements and vice versa. How intertwined the genres are becomes clearer further down.

Overall, the result didn’t surprise me much. Classics eke out a marginal existence at 1%. The comparatively high share of non-fiction surprised me, but cookbooks and guides are probably of interest here. Travel guides, travelogues, and coffee table books on faraway countries can also be found. Some academic literature from school and university even shows up. I would have ranked science fiction much higher as well. But since book blogging is a female domain, the cliché of low tech-nerd affinity may be confirmed here.

That’s a reason to look at the distribution of genres by gender. This is borderline, though, because 19 male bloggers are compared to 481 female bloggers. But even among male readers, Young Adult ranks highly. Crime & Thriller, however, are significantly more popular than among women. The gap is quite small for Novels and Fantasy, as well as for Comics and Biographies. Romance, on the other hand, is much more popular with the female audience, which again fulfills a cliché. Classics are significantly more popular among male readers, as is science fiction. What surprises me in the comparison is the high share of Young Adult among male readers and the comparatively low share of Fantasy. But the database for this comparison is quite thin.

A prejudice that came up in my last evaluation is the question of whether bloggers who chose Blogspot as their platform have significantly different reading habits. A comparison with bloggers who use WordPress (whether on wordpress.com or self-hosted) and Blogspot bloggers shows that there’s something to it. With Blogspot bloggers, Young Adult, Romance, Fantasy, and Erotica are significantly more popular than with WordPress bloggers. The latter read considerably more Novels, Crime & Thrillers, Non-Fiction, and also Classics.

Ratings

Another prejudice is the above-average positive rating of books, which makes book bloggers seem uncritical. One argument is the fact that book bloggers naturally choose books they want to read, by authors they like, and that they hand-pick. To get to the bottom of this, I plotted the ratings (in percent) by star count.

86% of books are rated with 4 or more stars. The probability that a book receives a very positive rating is therefore very high. It’s understandable that publishers perceive book bloggers more as a marketing channel than as literary critics.

On average, books are rated 4.2 stars. Comparing genres, there is only a small variance. Young Adult, Romance, Fantasy, Historical Fiction, Erotica, Biographies, and Poetry & Drama are rated at 4.2 stars, like the average. Comics at 4.4 and Children’s Books at 4.3 stars are rated a bit more positively than average. Classics at 4.1 are viewed slightly more critically, but overall there are no truly notable differences between genres.

How does the comparison between Blogspot bloggers and WordPress bloggers look? Bloggers who use WordPress are slightly below average at 4.14 stars. Blogspot bloggers average 4.2 stars. So the idea that Blogspot bloggers rate every book with five stars and WordPress users are much more critical can’t be confirmed here, even if WordPress bloggers are a touch more critical.

What about differences between male and female bloggers? Those are small as well. Male bloggers average 4.1 stars, female bloggers 4.2 stars. Given the small number of male bloggers, I’d conclude that gender has no effect on ratings.

Books

I often have the feeling that the same books are driven through the blogs over and over, and some titles seem to pop up everywhere. So the question is whether everyone is really reading the same things, or whether that impression is misleading. Which books are especially popular, and how homogeneous is the selection of literature among book bloggers?

For this, I first determined the 10 most-read books (of which I have not read a single one):

Me Before You by Jojo Moyes
9783499267031
Novels
read by 168 bloggers

Dream a Little Dream by Kerstin Gier
9783841421050
Young Adult
read by 164 bloggers

Sapphire Blue by Kerstin Gier
9783401063478
Young Adult
read by 162 bloggers

Emerald Green by Kerstin Gier
9783401063485
Young Adult
read by 160 bloggers

Ruby Red by Kerstin Gier
9783401063348
Young Adult
read by 142 bloggers

Fifty Shades of Grey by E. L. James
9783442478958
Erotica
read by 142 bloggers

Dream On by Kerstin Gier
9783841421678
Young Adult
read by 135 bloggers

Mockingjay by Suzanne Collins
9783789132209
Young Adult
read by 125 bloggers

Obsidian by Jennifer L. Armentrout
9783551583314
Young Adult
read by 122 bloggers

Slammed by Colleen Hoover
9783423715621
Young Adult
read by 118 bloggers

So, do all bloggers read the same books? Obviously not, because on average each book is read by only 3 bloggers. But there are differences across genres. Young Adult books are read by an average of 7.3 bloggers. That’s the average, i.e., as the Top 10 list above shows, there are many books that haunt the blogosphere and are read by very many people. For Young Adult, marketing probably works best and is most profitable for publishers. Books from Erotica (5.15 readers per book) and Fantasy (4.26), as well as Romance (3.97), are also read above average. With 2.43 bloggers per book for Novels and 2.47 bloggers per book for Crime & Thrillers, there are probably fewer hotly desired titles. Non-fiction (1.36 bloggers per book), Comics (1.44), Biographies (1.56), Poetry & Drama (1.36), and Non-fiction (1.36) have the least viral effect. For Classics, things look pretty bleak too: a book is read by an average of 1.79 bloggers. However, these genres are also not particularly popular, and the likelihood of overlap is correspondingly lower.

Now, three readers per book is a mean value; many extremes are possible. A closer histogram analysis shows that 60% of books are read by only one blogger. Only 5% of books (just under 3,000 titles) are read by more than 10 bloggers. This figure also indicates a handful of very sought-after books, but also that the vast majority choose a quite individual selection.

However, one must differentiate. Considered here is how many bloggers have read a book. It does not allow conclusions about how the readers of a blog behave. Some of my readers who do not blog themselves pick up books I present and recommend. Visitors who arrive via search engines may only read a single review, then bounce, and perhaps buy a book (or refrain from buying if the review is unfavorable).

I see the prejudice that bloggers only inspire each other as refuted here—with the exception of Young Adult. That’s a nice result, because it shows that book bloggers have minds of their own but still don’t isolate themselves completely and do take tips from other book bloggers. It’s a similarly nice result as the fact that book blogs are very diffusely networked and there is no clique formation.

Another exciting question is what the situation looks like for very popular blogs. If I consider only the books read by bloggers who rank in the top 100 of the Top List, the result is identical. These books are also read by an average of three bloggers. This shows that very well-known book bloggers do not follow the masses and only read what is currently in high demand; they still make their very individual choices. On the other hand, it also shows that other bloggers do not blindly follow the book choices of a few well-known bloggers. Popularity does not seem to depend on choosing the most mass-appeal books.

Another intriguing question involves differences by federal state. Which books are particularly popular with book bloggers in each German state? The data here are not really meaningful, because the origin is known for only a small fraction of bloggers who are also active on LovelyBooks or Goodreads. Daniela Lange from KaufDa conducted a similar evaluation and tried to find out via Google Trends which bestsellers are particularly popular in which federal state (in general, not limited to book bloggers). The result is an infographic that fits well here:

Bloggers

But even more information can be gleaned from these data—about the bloggers and their blogs themselves. Especially in combination with my first evaluation, new insights emerge into the networking and structure of the blogosphere.

I need to explain an important assumption up front; without it the following evaluations make little sense. The question is: when can a blogger be assigned to a genre? When is a blogger a Young Adult blogger? My simple criterion is this: if more than 10% of the books read come from a genre, then this is a preferred genre of the blogger. If there are 100 rated books and at least 10 of them are Young Adult, then the blog is assigned the Young Adult category. So when I mention all bloggers who prefer a particular genre below, it means that more than 10% of the books they’ve read belong to that genre. I believe this 10% threshold is a good one, though it’s chosen purely from a subjective feeling and not established empirically or calculated in any way.

An interesting question is whether the genre influences how often a blog appears in other blogs’ blogrolls. For example, are bloggers who primarily write about Classics extremely popular and appear in many other blogrolls? Variance is low here as well. Bloggers who write about Children’s Books stand out with an average of 24 links to their blog. For all other genres, the average number of other blogs linking to them hovers around 16 links. Science Fiction is rather unpopular with an average of 9 links. Biographies, Non-Fiction, Classics, and Erotica are in the lower range with 12 links. Young Adult with 18 links and Novels with 17 links are in the middle. Choosing a specific genre does not, in itself, attract an above-average number of readers.

Furthermore, the question arises whether the age of blogs is related to genre. Blogs that focus on Non-Fiction and Biographies have the longest staying power, with an average age of over three years. The youngest, with an average of about 1.5 years, are blogs that primarily write about Classics, Erotica, or Science Fiction. All other genres have an average age of two to three years. That’s roughly in the range where the entirety of all blogs lies. So, with the exception of the outliers above, no correlation can be inferred.

How does the choice of genre relate to the position in the Top List? Blogs that write about Children’s Books perform best with an average rank of 470. Blogs that write about Young Adult and Novels average rank 580. Bloggers who write about Erotica (760) or Biographies (710) rank more toward the bottom. The remaining genres lie in the golden mean, between an average position of 550 to 650. Blogs that write about popular genres are thus not necessarily more popular, which is plausible because the selection of blogs is simply larger there. Children’s Books are a bit of an outlier, but they are often read together with Novels, Fantasy, and Young Adult (the three most in-demand genres).

Which brings me to the next evaluation: Which genres do bloggers like to read together? The visualization above shows which pairings of genres are preferred by bloggers. The thicker the connection, the more often you encounter that combination of genres together. Young Adult and Romance is therefore a very common combination. Also Fantasy and Crime & Thriller, or Fantasy and Young Adult, or Novels and Fantasy. Here too, the popularity of the genres is reflected. Genres that are very popular are also often read in combination with each other. The result isn’t particularly surprising and offers few new insights.

Networking

But what about blog reading habits? In other words, which bloggers like to read which blogs? For this I created a chord diagram showing which bloggers read which blogs, based on the genres they themselves focus on in their own blogs.

The diagram thus shows the networking of blogs with one another, grouped and weighted by genre. If a connection is very wide on one side, it means a very large number of those bloggers read other blogs that focus on that genre. It’s immediately apparent that bloggers who write about Young Adult very much like to read blogs that themselves prefer Novels and Fantasy (and very likely blog about them). Likewise, a large proportion of bloggers who like Novels, Fantasy, and Young Adult read blogs that focus on these genres. That’s of course obvious. Bloggers who like Crime & Thrillers, however, seem comparatively less interested in Fantasy blogs and in Non-Fiction. Bloggers dealing with Classics are particularly interested in Novels, Young Adult, and Fantasy—again, the popular genres. Interest in Non-Fiction here is surprisingly low.

Overall, this visualization also reflects the popularity of genres. Bloggers remain particularly loyal to the genre that interests them most or is closely related to it. Young Adult, Fantasy, and Novels are often not exactly separable, and thus bloggers in these genres are particularly strongly networked with each other. At first glance, you’d say they stick to their own. On closer inspection, that’s not entirely true. Even if the share is small, some bloggers of one genre do read some blogs from entirely different genres. That seems quite plausible. I primarily read blogs with a very similar taste to mine, but also a few where I’m interested in the blogger themselves or their very specific selection, even if I don’t pick it up myself.

I find it very pleasing that all genres are nonetheless interconnected and none is completely excluded. So if a blogger writes about Classics or Romance, that doesn’t keep other bloggers from reading them at all. Non-Fiction bloggers are not completely sidelined; for example, there are always some Young Adult bloggers who read them. Considering the nominal distribution, these blogs are similarly well networked as blogs from the well-known genres, as you can see above in the comparison of the number of links per genre. Only the absolute numbers are different, because there are simply fewer bloggers focusing on, say, Classics.

In any case, it can be said that the interest in books in the blogosphere is broad across genre boundaries. However, it finds its focus in the popular and much-read genres.

Conclusion

In summary, Young Adult, Novels, Fantasy, and Crime & Thrillers are by far the most popular genres and account for over 72% of what’s read. Among male bloggers, Crime & Thrillers, Classics, and Science Fiction are much more popular than among their female counterparts. Blogspot bloggers read significantly more Young Adult, Fantasy, and Romance. In comparison, WordPress bloggers prefer Novels, Crime & Thrillers, and Non-Fiction.

Book bloggers rate their literature exceptionally positively, with an average of 4.2 stars, with 86% of all books rated 4 or 5 stars. Genres, blogging platform, or gender play no role here and do not influence rating behavior.

On average, a book is read by three bloggers. This depends heavily on the genre, however. Young Adult books are read by an average of 7.3 bloggers. Novels, by contrast, are read by only 2.43 bloggers. The extent of overlap by book thus depends on genre, but cannot be derived from a genre’s overall popularity. 60% of books are read by only one blogger, and only 5% of books are read by more than 10 bloggers. There are a few very well-known and widely read books, but beyond this small number, book bloggers have very individual tastes and make highly personal choices. Even the books read by very popular bloggers are read by an average of only two other bloggers. Very well-known bloggers do not follow the crowd’s taste, nor does the crowd blindly follow the book tips of a few well-known bloggers.

The choice of genre has only a small impact on how many other blogs link to a blog. An exception is blogs that write about Children’s Books, which are linked well above the average of 16 links, and blogs with very unpopular genres, which are linked considerably less. Genre has a limited impact on the average ranking in the Top List. Some genres perform better, others worse, but again there is no correlation between a genre’s popularity and ranking.

The age of a blog and its genre only marginally depend on each other. On average, blogs are two to three years old, with deviations; in some genres, blogs are on average only 1.5 years old.

Genres that are very popular are also often read in combination. The popularity of genres is clearly reflected here. The more popular a genre is, the more often it appears with other very popular genres. That’s hardly surprising.

Bloggers especially read blogs that cover a genre they themselves have as a primary topic—or one closely related to it. Nevertheless, all genres are interconnected, and it can be said that the interest in books in the blogosphere is broad across genre boundaries. However, it finds its focus in the popular and much-read genres.

As a résumé, I would conclude that book bloggers are exactly what you could wish for in a network. They influence and inspire each other, but still maintain their own style and individuality. Of course, these are averages, and there are blogs that primarily discuss very well-known titles and those that read rather off the beaten path. Based on this evaluation and my previous experience, I would conclude that there are platforms and genres where there is less diversity and others where there is greater diversity. Taken as a whole, however, book bloggers are very broadly positioned, don’t let themselves be told what to read, and read what they feel like. And it’s clear they love a large portion of their books, because if you read a lot, you eventually develop a sure sense of what you like and what you don’t. From the blogosphere, they primarily pick the book blogs that match their own tastes, but they still always take a side glance.