I’m interested in what MacGuffin analytics – specifically reader drop-out analytics – have to say about narrative structure.
The example I keep coming back to is Cory Doctorow’s excellent story ‘The Right Book.’ This story’s subdivided into 4 distinct segments (including an afterword), which are each denoted with a subtitle.
Conventional wisdom suggests that short, sharp sections like this keep the reader engaged (I’ve heard editors argue that it gives the reader a sense of accomplishment; that they’re progressing through a story or novel). I’m struggling to find other credible sources for this, or formal studies, but there’s plenty of chatter in writers’ forums and blogs on the merits of short chapters vs long chapters in novels. Here, for example, and here, and here.
But the analytics for ‘The Right Book’ seem to suggest otherwise. This graph from the analytics page shows the points at which readers exited the story. Blue for readers, orange for listeners to the reading (by Neil Gaiman).
The sample size is 132 readers and 38 listeners. Drop-out data for reading comes from scroll progress and page exits; drop-out data for listening comes from audio progress and page exits.
The graph starts at 100% readers (y axis), then decreases as readers stop reading or listening. Let’s ignore the sharp dropout rate right at the start (0-10% on the story progress x axis); we’ve found this is typical of most stories on MacGuffin, and likely reflects readers clicking on a story, then immediately deciding they don’t want to commit to it.
But after that, you can see readers and listeners dropping out at points that correspond approximately with the section breaks in the story, at around 37%, 75%, and 87% of the way through (drop-outs at >95% likely mean the reader got to the end but exited without quite scrolling to the bottom of the page or letting the audio play right to the end).
I’m not sure what kind of psychological factors are at play here, but I think it raises these questions:
Does arriving at the end of a section invite readers to pause and contemplate the story so far, and then become distracted?
Does arriving at the end of a section engender a sense of accomplishment that lets readers disengage?
Does the reader see the next section looming and think: that looks like too much commitment?
What the analytics don’t tell us.
Our analytics only collect data from one session, so if a reader gets to the end of a section, exits the MacGuffin website (or app), then later relaunches and resumes the story, that still counts as an incomplete read on the above graph.
Likewise, for all we know, if the story didn’t have section breaks, the same number of readers and listeners might still have exited the text or audio before the end of the story (though these drop-outs would probably not be clustered).
And of course, there’s no contextual data, so we don’t know why these readers exited. They could have arrived at their bus stop at around the same time as a section break occurs.
These analytics don’t say anything about the quality of a story. Stories with high reader-retention are not necessarily of high literary merit. Check out the stats on stories by Joyce and Chekhov.
Please let me know what you think below.