63% of your readers don’t finish your book. Here’s why.

book-reviveThis is the second in my current series of articles that examines reader immersion and the importance it plays in the success of a book. In the first article, I explored the psychology of what happens in a reader’s head when she reads a new book—especially one by an unfamiliar author—and encounters small irritants that break the immersive flow. In this article I’ll be exploring the not-so-obvious relationship between immersion and discoverability. And in the final article, I’ll look at the most common reasons for broken immersion, and what you can do to fix them.

A few critics of the first article have pointed out that not every reader is bothered by the sorts of disruptions I described, which is true. And yes, there are even some readers who, having opened the cover, will then read to the end no matter what they find inside. But for most readers, there is some list of sins that they do not suffer gladly. One might bail in response to illogical character behaviors. Another might recoil from awkward prose. And still others might rage against repeated clichés. Whatever editorial sin you can think of, rest assured that there is some population of potential readers who will turn their noses up at it and depart, even if that population is small.

But here’s the crucial bit. In fact, it’s so important that I’ve devoted this entire article to explaining why I think it’s true. Taken together, editorial flaws kill word-of-mouth marketing. And that means that the more errors your book has, the less visible it is to potential new readers.

Consider how word spreads about a popular book. It all begins with a marketing campaign of some description, which reaches a small group of potential readers and hopefully inspires a few of them to go off and try the book. What we’re all hoping for, of course, is that mythical viral response, in which every person who does try it is so energized by your story that they immediately rush out to tell everyone they know about your amazing book. And if we’re really lucky, those friends will try it too, be similarly impressed, and will then amplify the rumor mill by telling their own friends as well. Repeat until rich.

But for most indie authors, that ideal is not how things actually play out. For the great majority of us, we launch our little marketing campaigns with great hopes, and then, after a brief tremor of activity, the waters go silent again. What happened?

Finishing the book

Well, there are hundreds of factors that might explain what happened in your particular case. But if you’re hoping for that viral response, there is one factor that I can guarantee is crucial:

Readers who try your book should at least be able to finish it.

I mean, that only stands to reason, right? If we’re trying to build a club of rave-reviewers, surely “finishing the book” is a basic qualification for membership. The more people who finish it, the more potential ravers you’ll have on your team.

As I mentioned in the previous article, each time your readers’ immersion in the book is interrupted, they go through a series of mental steps, part of which is deciding whether or not to continue reading. So each time they are confronted by an error that distracts them from the story, there is another chance, however small, that they will put the book down and never come back.

So one of the keys to getting more people to finish your book is to minimize the number and severity of these interruptions. And as I said, some interruptions you can’t control, like the number of times their kids will ask for lunch, or the number of phone calls they’ll receive while reading. But you can control the number and severity of the editorial mistakes that might distract them. Remember, no mistake will bother all readers, but it’s a simple fact of mathematics that as the number and/or severity of mistakes increases, the proportion of readers who bail as a result will increase with it. Still, we’re talking small risks, right?

How little risks accumulate

Unfortunately, small risks often have a habit of accumulating into big ones. As a proof of that concept, let’s do a little thought experiment.

100sidedDice-1_largeSuppose I give you a die with 100 sides, numbered from 1 to 100, and I make you the following bet: roll the die once, and if a 7 comes up, you pay me $100. But if some other number turns up, I’ll pay you $10. Would you take the bet? Most people would. The risk of losing is only 1%, so it looks like an easy ten bucks.

But what if we change the rules? Suppose you have to roll the die 100 times, and if you avoid the 7 every time, I’ll pay you the $10, but if a 7 comes up even once, you lose and have to pay me the $100. Now would you take the bet? It’s still only a 1% chance that the 7 will turn up on any roll. But as statistics tells us, the chance of avoiding the 7 in this scenario is only 37%. Suddenly, this is starting to look like a bad bet. And if we increase it to 1000 rolls of the die, statisticians in the room will be able to tell you that you’re doomed.

The point is that, while a risk might be small, if you roll the dice on that risk often enough, it will eventually bite you. So what we need to know now is whether we’re rolling the die just once on our book, or many times.

Accumulated interruptions

In a quick search of research findings, I discovered three interesting articles on the consequences of interrupting readers:

  1. Interruptions impair comprehension
  2. Interruptions increase anxiety, irritation, and perceived difficulty
  3. Lexical ambiguity increases reader confusion and slows down reading

But I wasn’t able to find any clinical research on the simple metric of how often readers get interrupted. This is probably because that number would be entirely dependent on the book itself and the interests of the reader, so no really meaningful generalizations can be made. But one thing these researchers do agree on is that the more interruptions there are, the more extreme the response gets.

Since we don’t have clinical study data about error rates to work with, I’ll use the data I’ve collected myself in over 240 ImmerseOrDie readings. If we consider a typical indie book, my data suggests that we will find approximately 1 error in every 1000 words.

When I’m reading for pleasure, I usually tolerate more interruptions than I do for IOD, but even so, by the time I’ve found 15 or 20 mistakes, even in a leisure book, I’m pretty likely to put it down. Let’s assume, though, that the typical reader is more forgiving than I am, and that he will need to find 100 errors before he’ll abandon the book, which makes for a 1% chance that he will do so on any single error. (This actually seems insanely low to me, but I don’t want any accusations that I’ve rigged the numbers to favor my argument.) So if we take a typical indie fantasy book, which is about 100,000 words long, our typical indie reader will, on average, be interrupted by errors 100 times before he reaches the end.

Do you recognize the situation? It’s the same one I gave you earlier about the 100-side die. And do you remember what the chances were of avoiding the 1% chance in 100 successive rolls? It was only 37%. Translating that back into our book scenario, it means that if we assume a fairly congenial reader who is relatively unfazed by editorial errors, he still has only a 37% chance of finishing the typical indie book, because we confront him with that finish-or-abandon choice 100 times. And that calculation assumes that his level of irritation does not increase as he continues reading. (Remember, the research linked above shows that irritation levels do rise with repeated interruption, so the actual rate of indie books being read to completion might be even lower.)

More importantly, this means that almost 2/3 of the people who start the typical indie book will not finish it. And that should be a frightening number to every author. No matter how wonderful your story is, if your book has frequent low-grade editorial errors, you may be kissing 63% of your readers goodbye. That means 63% of their word-of-mouth recommendations, too. And it’s not just spoken referrals, either. It also means 63% fewer mentions on social media, and 63% fewer reviews posted. Or worse, a higher number of bad reviews posted. Nobody who bails on a book gives it a good review.

By way of contrast, let’s have a look at non-typical indie books. The best indie books that I have data for are the ones I featured in the StoryBundle collection earlier this year. From an editorial perspective, these were much cleaner than the typical IOD books, and the average rate of errors I found in this group was about 1 in every 10,000 words. That means that with a well-edited indie book, the reader will only see 10 errors in its 100,000 words. And if we roll the 1% chance 10 times, fully 90% of readers will make it to the end, which is far better than 37%.

Which of the two books do you think has the better chance of getting word-of-mouth support?


Admittedly, this has been a great deal of loose math and waving hands, but the principles of viral propagation are not yet known with mathematical rigor, so we’ve been working with the data we have. We started with two very reasonable assumptions: that a successful book involves happy readers telling their friends about it; and that they are unlikely to do so if they can’t even finish reading it. We then made some fairly generous assumptions about how many errors a typical reader will tolerate before abandoning a book, and compared that to the typical error rates found in indie books. But even with those generous assumptions, this still suggests that only a third of readers who pick up an indie book will actually finish it.

In practice, however, I abandon far more than 63% of the indie books I try. My own reject rate is about 92%, and most of the other people I talk to estimate that they quit on at least 80%. But one thing is clear. If those books we rejected had been cleanly edited, we would have read an awful lot more of them through to completion. And who knows—we might even have told our friends about a few of them too.

So if you’re out there spending piles of time, energy, and money on advertising, blog tours, pleading with reviewers, and standing on street corners waving free copies of your book in the air, hoping somebody—anybody—will give it a try, doesn’t it frustrate you that 2/3 of the people you do convince to give it a try will bail before they finish? Pretty demoralizing, huh? Feels like so much wasted effort. But the fix is simple. Want to almost triple the number of people who finish your book and then tell others about it?

Get an editor.

Convinced that editing is important? Then maybe you’ll want to join me for the next and final article in this series, where I’ll lay out the most common editorial problems that break immersion, with examples, and tell you how to spot them, and how to fix them. And if you don’t want to have to keep coming back to look for it, you can always register for new article alerts. Either way, see ya next time.

51 things that break reader immersion, with examples
Why readers bail on books. Maybe even yours.

About the author

Jefferson Smith is a Canadian fantasy author, as well as the founder, chief editor and resident proctologist of ImmerseOrDie. With a PhD in Computer Science and Creativity Systems compounded by a life spent exploring most art forms for fun and profit, he is uniquely unqualified in just about everything. That's why he writes.