Over the past couple of months, I’ve been studying the analytics for my free-to-read fantasy serial on JukePop, Metal Shadow. Looking at analytics isn’t always fun. I’m sure long-time readers of this blog know this. Either the analytics reveal something great, or they reveal something terrifying. I’ve yet to witness a middle ground.

However, analytics remains a very useful tool for testing hypotheses about reader preferences. Today I noticed something both great and terrifying while studying it. I sort of expected it when I set out to write the latest chapters, but it still sort of shocked me.

Exhibit A:

JP Analytics for Metal Shadow
JP Analytics for Metal Shadow

Bleh. I hate looking at those first two dips, the ones on chapters 6 and 10. I wish I could take them back.

But now we’ve got some new dips since before, don’t we? One on chapter 13, 18, and another on 22. Let’s go through each one.

Chapter 13

This dip is sort of enigmatic and minor, and not really one I care to talk about.

Up to this point, the story had been gearing up for a tournament of sorts. Our main character, Lloyde, had become a candidate for a seat on Deep Kuralle’s board of directors, and now had to prove himself out in the field. Chapter 13 was the start of this exam. It was a good chapter overall by my standards, but it did have an accelerated pace compared to previous chapters. I guess readers wanted me to draw things out a bit more, build up the suspense and whatnot. Shoot, if that’s what they’re mad at, then the entire first 15 chapters need to be overhauled. I wrote them like a schizophrenic orphan looking for a fix. (I kid, but then again, I’m so seri’.)

Chapter 18

This one is where the magic starts to happen.

You’ll notice a peak in +Votes on chapter 15, the chapter which marked the end of the exams and also dished out probably the best cliffhanger I’ve written in any of my works to date. Chapter 16 was the necessary follow-up chapter to finish off the cliffhanger, as well as the start of the second story arc (chapters 16-34).  Things seem pretty normal don’t they? Alas.

In the second half of chapter 16 I spontaneously did something I had NOT told myself to do in the outlines: I switched the story’s POV from Lloyde Esmonde to someone else. This marked the first time I did this since chapter 5—the story went a total of 10.5 chapters from Lloyde’s POV, before finally switching to someone else.

Then, in chapter 17? I switched from that character to someone else.

18? Take a guess. I switched POVs to yet another set of characters, then halfway through that chapter, switched again to a fourth set of characters.

BOOM. Dip.

“Okay,” you say, “big deal. Maybe chapter 18 just sucked and people didn’t vote on it because it sucked. What other evidence do you have that switching POVs causes a dip in +Votes?”

Chapter 22

Hey graph, come back in here for a sec, would ya?

JP Analytics for Metal Shadow
JP Analytics for Metal Shadow

See chapters 19-21? Notice the steadiness, then the uptick in votes? Those chapters were where we came back to Lloyde’s POV. Apparently, people like his storyline. They’re invested in him. Then again, I also changed up my writing style, and made those chapters a bit more action-oriented and slower-paced. Maybe all that came together to earn those extra 2 +Votes.

…Except, there’s chapter 22, when I switched back to the POV of the guys from chapter 18. Behold The Dip.

…Then I go back to Lloyde in chapter 23, and presto.

Get it now? I think it’s a pretty obvious pattern. By process of elimination, we can exclude the writing style and pace of the story as the factors that most alter +Votes.  Switching character POVs is the more likely reason readers lost interest.

It makes sense. People invest in one main character or group of characters when they first open a book, and when you force them to reinvest in new people, they lose patience. Perhaps you’ve experienced this before, or perhaps you haven’t. Either way, we now have some modicum of proof.

I have a couple more questions based on this data that I’d like to have answered. Let’s see what the charts say in a few weeks. I should also note, this is a pathetically small sampling size. Send some readers my way, would ya? Tell them it’s For Sciance.

Anyway, it’s good to be back on the grind with my favorite story. Gosh, I really hate my writing style from pre-chapter 17. We’ll have to do something about that before the next story arc. In the meantime, onward! New chapters of Metal Shadow will be coming in every week on JukePop.

How often do you switch POVs in your story?


6 thoughts on “Switching POVs = Bad

  1. I change my POV all the time for In Evening, but I made sure the characters I switched to were as interesting, if not, more so than the main. But I had a dip in chapter 15 as well, which is when I introduced a new anti-villain and had the story told from their point of view. I don’t have enough +votes to conclude anything though. Just the view counts.

  2. I switch POVs for each chapter, between the two main characters. After trying out nearly every other style, I find that I like this one best, and that readers tend to accept it. I think switching POV is fine if it’s consistent and the reader isn’t caught off-guard.

  3. Like Natalie, I also switch POVs alternating for every chapter between the two MCs. My analytics don’t show that the trend has a down side. I have also read many books that switch POV constantly to great effect. I think the problem stems from the POV changes seeming random, difficult to keep up with, or not appearing to add to the story. I have certainly stopped reading books where this has been the case. I haven’t read Metal Shadow far enough in to get to the POV switch, so I don’t know if any of that applies. To make general claims though I would say that POV switches need to be intentional, and well thought out in order to work for the reader.

    To give an added perspective to the analytics I would just add that mine look similar in overall trajectory. They start with a very high number of plus votes on chapter one plummet over the next few chapters, then level out around chapter ten and then have little ups and downs (but overall a downward trend) through to the most recent chapter. I have come to believe that analytics don’t actually tell us all that much about our work except that only a small core group of people will read the novel all the way through to the end.

    1. Virginia, I think you’re right. The first and second chapters should be excluded from your interpretation of the statistics, because the consistent readers will tell you which chapters are good and which need work.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s