Men Are from Mars, Women Are from Venus, But What About Readers?

Expert publishing blog opinions are solely those of the blogger and not necessarily endorsed by DBW.

Audience + InsightThe recent decline in ebook sales has led many pundits to pronounce that the digital revolution is over. Admittedly, this is indeed the end, but merely the end of the beginning of how ebooks will affect publishing. We are now in the next wave of publishing’s digital transformation—one that is based on the use of data that is collected digitally and allows us to develop unique insights into audiences.

This is the fifth post in my exploration of data-smart publishing, and it explores whether gender and other demographic factors actually affect reading.

Over the past couple months, Jellybooks has tested hundreds of books by embedding a piece of Javascript software called candy.js into ebook files. Readers received the ebooks free of charge in exchange for sharing their reading data with us and the publisher. The software recorded their reading data both online and off, and when the user clicked a button at the end of the chapter the data was uploaded.

When users claimed their free ebook as part of the test reading trial, we asked them for their age and gender; this allowed us to examine whether these traits influenced their reading behavior.

When it came to participating in the trials, far more women signed up than men. This is not a surprise: women account for more book purchases and books read than men do. In general, we recorded 20/80 male/female splits across test groups, though some books were noticeably more likely to be picked by men than others (up to five times more likely, in fact).

What was more interesting to us, though, was whether the sub-group of men that read a book had the same completion rate as women. If a man decides to read a book, is he less likely or more likely than a woman to finish it? In other words, is the completion rate of a book at all gender-specific?

In general, the result was a firm no. In most cases, the likelihood that a reader will finish a book is not correlated with gender; both sexes have an equal probability of finishing a book. Issues such as writing style, strength of characters, topic and other factors have a bigger influence on the completion rate.

This holds true across non-fiction and literary fiction, including genre fiction like fantasy, science fiction and crime. Below is an example of a book by a Canadian author (Jellybooks test title #1048) that shows how men and women complete the book at near-equal rates (27 percent for male readers and 28 percent for female readers, which is not a statistically significant difference for sample of 400 test readers).

Men and Women

There is one noticeable gender-specific difference in reading across most books, however, which is well-illustrated in the above example: men decide much faster than women do if they like a book or not. The initial decline during which most readers are lost is much sharper and earlier for men than it is for women, and this is a behavior that we observe for the majority of books (the above title also loses readers in the middle of the book, which is a rather rare occurrence).

So put another way, men give up on a book much sooner than women do. Given the identical completion rates, we take this to mean that men either have more foresight in this regard or that women continue reading even if they already know that the book is not to their liking. We suspect the latter, but cannot prove it at this point.

In an earlier post, I highlighted that authors need to capture the attention of readers quickly. What our demographic reading analysis shows is that, when it comes to men, an author has only 20-50 pages to capture their attention. No room for rambling introductions. The author needs to get to the point quickly, build suspense or otherwise capture the male reader, or he is gone, gone, gone.

Now, there is in fact a noticeable exception to the rule of identical completion rates for men and women, as there is a type of book for which we observe significant differences between the sexes: books that deal with feelings. This includes books about emotions like grief, loss and love, but also books about relationships in general and romance in particular.

Books that predominantly deal with these categories show noticeable differences in completion rates, which can vary from relatively small differences (5-10 percent fewer men finish the book) to very large difference, in which the completion rate for men is half the value or less than that for women. Not only do fewer men start reading these books, but those who do start reading them are more likely to give up on them than women are, irrespective of the quality of the content or the narrative.

In such cases, it doesn’t matter if the author is male or female. Even male authors dealing with the above emotions will witness male readers being less engaged than women.

The fact that emotion is such a dominant influence should not come as a huge surprise, and anybody who has read John Gray’s Men Are from Mars, Women Are from Venus will be familiar with the arguments.

So in a lot of cases (e.g., a romance novel), we can pretty much tell a priori what the book’s audience will be. But where the tool comes into its own are the special cases for which it can help immensely to understand if a book has a strong or weak male appeal. Some young adult titles, for example, fall into this category, as do some historical fiction ones.

In an upcoming post, we will examine the influence of age on completion rates. Without giving too much of the story away, age has a much, much stronger impact on completion rates than gender does. More on that in a few weeks, and of course you will hear a lot more about this topic at our DBW workshop on data-smart publishing next month.

Note: All the data reported in this post was collected in pilot projects financed by Innovate UK. EPUB3 files were modified with candy.js by Jellybooks so we could record, store and extract the user’s reading behavior when using iBooks, Adobe Digital Editions (ADE) and selected Android reading applications. The data stored within the ebook file was extracted when the user clicked a “sync” button in the book. All users were informed about the presence of the analytics software.

We will also be holding a workshop on data-smart book publishing at the upcoming Digital Book World Conference in New York City. The workshop takes place on Monday, March 7th from 2pm to 5pm, just prior to the main DBW conference. We will look at the challenges publishers face in collecting data, making sense of data and applying it so they can publish smarter, more efficiently and more profitably. The workshop will include speakers from Elsevier, Piper (Bonnier Germany) and others sharing their experience of turning themselves into data-smart publishers.

To get all the ebook and digital publishing news you need every day in your inbox at 8:00 AM, sign up for the DBW Daily today!

6 thoughts on “Men Are from Mars, Women Are from Venus, But What About Readers?

  1. J. Hunt

    “So put another way, men give up on a book much sooner than women do. Given the identical completion rates, we take this to mean that men either have more foresight in this regard or that women continue reading even if they already know that the book is not to their liking. We suspect the latter, but cannot prove it at this point.”

    Another interpretation could be that women are more likely to give a book time to develop. Some books are “slow” reading but incredibly deep – Marilynne Robinson’s Gilead, for example.

    1. Andrew Rhomberg

      “Another interpretation could be that women are more likely to give a book time to develop”

      If that is the case, and the book gets better in the reader’s view, then you would expect the completion rate to improve as a result of \it developing\.

      Instead, what we are observing in many cases, is that despite identical completion rates, the initial abandonment rate is different. Woman hang in it for longer. Motivation aside, for why they do so, this observation holds even when the completion rates are identical, as they are in fact for many, many of the books we studied.

      Some book, as you point out, are indeed slow to develop, but in the 21st century that is – in my view – a risky strategy for an author. Some of those slow-to-develop books do go on to be very successful, but often because they come with a recommendation and a health warning \you need to stick it through the first 100 pages, it gets better after that…”. That of course increases the discoverability challenge for a book and I am not trying to state anything else.

      Reader analytics does not judge a book. Instead reader analytics explains how an audience is interacting with a book.

      An author should write the book he wants to write. Authors whose primary desire is to be read, though, may benefit from understanding something about readers…

  2. Michael W. Perry

    Interesting, very interesting. The results also fits well with my perceptions based on my own reading habits. I either give up early or find some reason to plow through to the end.

    The television show JAG illustrated the differing tastes of men and women. Built around a team of lawyers with the Navy’s Judge Advocate Group, the show successfully blended military stuff, i.e. carrier pilots, with romantic interests between the lawyers. Men’s attention heightened during the former, women’s during the latter, making the show attractive to both.

    There was, however, one episode that totally bombed in my mind because it was a male-like scene written from the perspective of very clueless females. Set in Hong Kong, I believe, it had the show’s female lawyers standing in the open firing their pistols (yes, pistols) at Chinese soldiers with automatic weapons. I’m sure some exceptionally stupid script writer thought that’d have its women being as bold as its men and thus appealing to women. In practice it didn’t. Anyone, male or female, who would stand in the open fighting automatic weapons with a pistol is an idiot. That weakened the shows credibility in my mind.

    Those doing this research might want to look for patterns in where men and women give up on a book. If there’s a pattern there, writers might want to take note and take care to shift between the two appeals often.

    I know I gave up on a series of novels when the author quit advancing its broader plot in a military conflict and focused instead on the lead character’s conflicting romantic interest in two women. \Choose one, and get on with the interesting stuff,\ was my very guyish response. When he didn’t do that, I dropped the series after the third book. He’d taken a military action thriller and turned it into a romance novel.

    –Michael W. Perry, editor at Inkling Books

    1. Andrew Rhomberg

      different audiences, different expectations

      one of the more common reasons for a book bombing is that readers expect, based on the cover, title or synopsis Book A, but instead get book B. It is one of the more common causes for a book’s completion rate to plummet steeply…

      my recommendation to authors and publishers: never mislead potential readers with regards to what they should expect from a book or they will turn from promoters into detractors…

  3. Anne

    I agree with Buzz Worthy, your analysis is at best one-sided. Men may just be more likely to be lazy ass quitters than women. So, yeah if you want to appeal to only that specific sub-category of readers, you may want to write differently. But that sub-category’s preferences shouldn’t run things for the rest of us. Because lowest-common-denominator art or entertainment is often pretty… low.

    Also worth noting, as Alyssa Cole’s international research has revealed romance category books are NOT limited to feminine appeal. In some countries, men read romances just as happily and openly as women do. It’s cultural, not sexual.

    Lastly, as paid subscription marketing data has proven online for two decades now, user data you gather on free content is not always applicable to paid content. People behave differently with things they pay for vs get for free. (In fact, men are culturally infamous for this.) So, go ahead and draw all the conclusions you want from your data – as long as you serve them up with a big bucket of salt and a bunch of caveats.

  4. Andrew Rhomberg

    (a) you seem to be very judgmental about men calling them lazy ass quitters. Maybe they are, in which case the data accurately records their fast quitting behavior. Maybe it is indeed laziness or maybe it is just a desire to read something more interesting from another author (most common reason given for stopping to read a book is I decided to read something else instead) and they reach that decision faster than woman. I can’t tell you with any certainty. I can only tell you that men – on average – quit a book sooner than woman even when they completion is identical for the sexes.

    (b) Nowhere in the post did I imply that an author should write differently, especially not in terms of appealing to one gender over the other. What I wrote in an earlier post and referred to in this post, is that it is a risky strategy for an author to take his or her sweet time to pull the reader in. The author risks losing a large share of their audience, but the author may not care. That is the author’s choice to make.

    (c) The observation we made was that for most books there was no gender difference in completion rates (completion rates already control for different rates of starting, etc.). The one notable exception where this was often not the case was books that deal with feelings (and that includes feeling other than romance, such as feelings of grief, loneliness, despair and similar). A book in that category was more likely to show a gender difference (in the UK, US and Germany), though sometimes there was none. There are always the exceptions that prove the rule.

    (d) Books are cheap and one-off purchase and as such not very comparable with a subscription commitment. When a reader commits to reading a book, they are making a major time investment. We worry more about people sticking with a book because they feel obliged to do so in return for receiving it for free. I note that Kobo has reported that 40% of books purchased are never even opened. Though, we would be able to put the tracking survey into retail copies, there are all sorts of issues, one of which is that we want to have data available before people go on sale ( I also note that advance reader review copies are made available for free and not for a fee, as such we are not doing anything unusual).

    (e) One should also add the caveat that people who don’t like a certain set of data are more likely than others to challenge that data. All the data reported in these posts is based on samples, because data on every book ever read is not available. Sometimes our initial conclusions may be wrong as more data becomes available, but the journey we are undertaking is to use data to better understand audiences and along the way we are sharing these insights with others.



Your email address will not be published. Required fields are marked *