Ebook Sales Conversion Rates (A Case Study)

Ebook Sales Conversion Rates (A Case Study)

A common tactic for today’s self-published author is to make one or more written works available for free. This tactic is recommended by many reputable authors, from Hugh Howey (long may he reign) on down. The major benefits to doing so are usually listed thus:

  • Free books get more downloads. You’re putting your name and your writing in front of more readers. If you did your job well with the book, you can acquire fans this way.
  • If you include a mailing list CTA in the back of the book, free ebooks can be a good way to sign people up for an email list.
  • If the book is the first in the series, a free first book is a GREAT way to sell later books in the series. In this case, you’re sacrificing your first book like a pawn, to reap the benefits of paid sales later in the series.

But on this last point—what can one actually expect to gain from employing this tactic? We don’t mean the general theory, here—that’s sound. We mean numbers.

My name is Garrett Robinson, and I’m a full-time independent author. I write in multiple genres, but by far my most popular series is the Realm Keepers series, co-written with fellow indie author Z.C. Bolger. In this article, I want to share the data I have on my own conversion rates of the books in the Realm Keepers series, which uses the “first book for free” tactic. My hope is that it may be useful to other independent authors  considering this approach.

For a first point of reference—I will only be presenting and contrasting data for two platforms in this article: Amazon and Kobo. Nook and Apple get less than ten sales per month, in my opinion far too few sales to draw any sort of meaningful conclusions.


Let’s take a look at how the series works, from both a story and a marketing perspective.

Realm Keepers is serialized YA fantasy. We have published two Books so far, entitled Midrealm and Wyrmspire. Each Book contains six Episodes, each about 25,000 words long, and a Finale that’s about 100,000 words long. The total length of each book is therefore about 250,000 words.

Here’s a graphic representation in case your head’s spinning:

Realm Keepers

The first episode is free. Episodes 2-6 are $0.99 each. The Book One Finale is $2.99. Midrealm, the first Book, is $4.99.

In Book Two, Wyrmspire, we want people to buy the full volume. So the episodes go up in price. Episodes 8-13 are $2.99 each, and the Book Two Finale is $4.99. Again, we do not want people to buy these episodes after the first book—we want them to get the complete volume to take advantage of the 70% royalty rate.

Episode One is the entrance to the funnel and our star player. It’s permanently free. It gets thousands of downloads per month.

That other important ingredient? “Do a good job with the free book?” Well, Episode One has 91 Amazon reviews, a 4.4-star average, and not a single 1 or 2-star review. The quantity of reviews is lower than industry average, based on how many times it’s been downloaded. But the quality of reviews is there. Clearly most of those who read it, like it.

A quick note: The funnel as described is almost the entirety of the marketing that’s been done for this series. We haven’t run BookBub ads or any other kind of paid advertising for the volumes or other episodes. ALL we do is promote Episode One. So this data is free from the influence of paid ads on the Books. It is, as near as we can tell, pure conversion data.


The first and most important question, the one I’m sure most of you are interested in, is this:

What percentage of readers download the free book, then immediately convert to a paid purchase afterward?

It would also be great to know what percentage of people who download Episode One actually read it, but there’s no way to measure it. We have to make assumptions based on straight download-to-purchase ratio.

This answer comes in two parts. When readers finish Episode One, the CTA in the back gives them two options: go buy Episode Two for $0.99, or go and buy Midrealm for $4.99 and save money over buying the episodes individually.




What about readers who take door number 2—purchasing Midrealm to save money instead of buying the episodes individually?




However, on both of these numbers, it’s important to remember that the CTA at the back of Episode Two sends people to purchase Episode Three, OR to buy Midrealm and save money on the episodes.

In other words, people could be doing the following:

  • Purchase Episode One
  • Read it
  • Like it, but not be invested enough to buy Midrealm
  • Purchase Episode Two to keep trying it out, because it’s only a buck
  • Reach the end of Episode Two, decide they’re in whole hog, and buy Midrealm

This hypothesis would seem to be supported by the fact that Midrealm’s number one “also-bought” on Amazon is Realm Keepers: Episode Two. However, there’s no way to be absolutely sure. It’s merely a caveat to keep in mind as you study this data.

05 copyWhen we add up these figures, we see that approximately 6% of readers IMMEDIATELY convert from a free download of Episode One, to a paid purchase.

The conversion rate to Episode Two ($0.99 purchase) is about twice as good as the conversion rate to Midrealm ($4.99 purchase). However, since the royalty on Episode Two is 35% of $0.99 and the royalty on Midrealm is 70% of $4.99, Midrealm conversion are half as numerous but return about 5X as much royalty income as Episode Two conversions.


What conclusions, however tentative, can we draw from this data?

  • A free first book is a good tactic to acquire paid conversions, albeit at a low percentage.
  • In all cases, the lower the conversion sales price, the more customers you are likely to get. However, Kobo customers seem to be much more likely to do a highly-paid conversion than Amazon customers.
  • Whether you will do a $0.99 conversion or a $4.99 must depend on the author’s strategy. Are you looking for more paid sales, period, or a higher return of royalties? In my case, I am shotgunning to get both, but it is by no means set in stone that this is the best method.


Immediate conversion is sexy. We all want people to be driven to buy that next book. But what good is it if they don’t stick around for the long haul? You want to build long-term readers, not flash-in-the-pan overnight rankings boosters.

That’s where long-term conversion comes in. Some have posited that a free download can “devalue” a series in the reader’s mind. Having gained the first book for free, they’re less likely to be invested in the series as a whole, even after purchasing something later.

There are few different ways we can measure long-term reader conversion with this series. The first and most obvious is how many readers continue buying episodes. After all, if they buy Episode Two but stop reading at that point, you haven’t got a long-term reader. You have someone who impulse-bought, then forgot about you.


Unfortunately, long-term conversion for Episodes 2-7 isn’t that impressive on Amazon, with only 34% as many sales of Episode Seven as there are of Episode Two—barely more than a third. As you can see in the graphic, there’s a rather significant drop off episode to episode.


However, we look almost twice as good on Kobo. The most significant drop off is between Episode Two and Episode Three. But once people get to Episode Four, there’s very little reader bleed.

Therefore it’s hard to say whether or not a free first book devalues the series in the reader’s mind. Again, this is also influenced by how many people may have gone on to purchase Midrealm after reading Episode Two. The number could be significant, but unfortunately there’s no way to tell for sure.


  • As with any series, drop-off gets larger the longer the series goes on.
  • Kobo readers seem to have more long-term loyalty than Amazon readers
  • It would be great to have visibility on Episode Two readers who go on to buy Midrealm instead of Episode Three. This could be accomplished with special forwarding links in the Episode Two CTA—something that’s relatively easy to implement, and would provide AWESOME visibility for getting a clearer picture of the data.


Perhaps an even better method of determining long-term conversion is seeing how many people stick around with us for the second Book, Wyrmspire.

Now remember, once people get to the second book, we want them to buy the volume. We’re no longer interested in selling the episodes. Therefore we’ve increased the price of episodes 8-14 rather significantly to drive people to the volume instead.

What we want to measure is those who FINISHED BOOK ONE. That means anyone who bought the Book One Finale, as well as anyone who bought Midrealm.



But what’s interesting is when we look at who went on to buy the Episodes, despite our attempts to turn them away.



So a higher number of people on Kobo would rather convert to Book Two than the Episodes, but not as many people as we’d like. We want people to buy Wyrmspire.

We’re clearly not pushing Wyrmspire strongly enough. People must not understand that they’re saving about 60% by buying Wyrmspire instead of the individual episodes. So that’s something we need to take a look at, see how we can make it more clear to people.

15 copy

When we add up Wyrmspire and Episode Eight sales, on Amazon we can see that we get approximately 41.3% as many buy-ins to Book Two as Book One. This is a much, much better number.

16 copy

On Kobo, it’s comparable at 40.1%.

This is significant. It tells us that, on both platforms, about two-fifths of readers who buy Book One (in either format) want to go on and read Book Two. We wish that number was higher, and we’ll work on it, but it’s something.

Let’s check one last table before we go—episode-to-episode drop off for Episodes 8-13. Here’s Amazon:


We have a MUCH lower drop off rate than we had for the first seven episodes—except for the last one, the Book Two Finale. Why is this?

Well, if someone’s invested in the story up to this point, they’re much less likely to drop off unless the price becomes prohibitive. And that’s probably what’s happening on the Book Two Finale. At $4.99, it’s as much as most other ebooks on the market (it’s also longer than most other ebooks on the market, but whatever). Again, this price was set because we WANT people to go buy Book Two, not the episodes.


When we look at Kobo’s episode drop off, we notice something bizarre. There are MORE purchases of Episodes 9-11 than there are of Episode Eight. I have no explanation for this phenomenon. Drop off after that is very low, until the Book Two Finale where it drops off even more than Amazon did.


  • The deeper people get in a series, the more invested they are. Drop off rate in the first few episodes was merciless. Dropoff rate by the end is tiny. What does this mean? If for some reason you have major drop off late in a series (i.e. the Book Two Finale) something is wrong! Fix it!
  • People definitely aren’t being pushed hard enough toward Wyrmspire. I need to increase the CTA strength in Episodes 8-13, but in Episode 8 in particular. In fact, I may simply remove the link to Episode Nine from the CTA altogether. The message should be along the lines of, Stop! You’ve just finished Episode Eight, and Episode Nine awaits. But you’ll save yourself a ton of cash—about fifteen bucks—if you go and buy Wyrmspire instead! Obviously it needs to be prettier than that, but you get the idea. These people are already invested in the story, as clearly demonstrated by their low drop off rate. They should have no problem getting the full volume if its value is clearly communicated.

I hope this has been informative to you. If it has, let’s make this interactive. Are there any major metrics you can see that I’ve missed? Is there data I haven’t covered that would be helpful to you in your own publishing journey? Let me know in the comments. If it’s data I actually have access to, I’ll revise the post to include it.

Until next time, I hope I’ve been of service. See you in the marketplace.

Garrett Robinson

Over 100,000 readers have read and loved Garrett's books, like the fantasy hits Nightblade and Midrealm. He's also a film festival favorite with movies like Unsaid, and a tech guru who posts lots of helpful how-tos for writers and filmmakers over at garrettbrobinson.com.


First, thanks for sharing your numbers so openly. It's useful and interesting information. Second, I'd like to address something as a reader...I read Midrealm -- I can't remember if I got it on sale, or what -- I think maybe Wade Finnegan raved about it and so I went and bought it. Anyway, I read it all, then went to see what came next and got confused. 

The way it is set up, with both episodes and a name for the collected volume is confusing. I lost track of the episodes while reading Midrealm -- didn't even realize I was reading episodes. So, when I went to see what was next, I couldn't tell. Did I read episodes 1-7? Was episode 8 part of Midrealm or did it stand alone? I know that I could have figured it all out without too much trouble, but like most people, I'm pretty lazy. 

I still haven't purchased Wyrmspire...for some reason I don't think it was out yet when I read Midrealm. Or maybe it was and I lazily gave up when I couldn't figure out what to buy next. I will eventually buy it, read it, and review it (though the review will likely be on GoodReads rather than Amazon). 

Do you think that your conversion rates might be suffering due to lazy and confused readers who don't know what to buy next? Of course, it could just be me. 

Garrett Robinson
Garrett Robinson moderator

@davastewart I suppose that could be the case. We title the volumes "Realm Keepers Volume One" and "Realm Keepers Volume Two," and we have a CTA in the back of all the books directing readers to the next one. But still, there's always that chance people won't get it. We're definitely not planning on doing serialization again in the near future.


Garrett and Simon, what are your thoughts on how the conversions would change if the reader began with a purchase? Even at a low entry point (.99) I believe people feel more invested from the get go. I could be wrong but getting them to open up their wallet is critical in my mind. 


@writerwade I think that people paying at the start would make them much more likely to buy later books. However, and it's a big however, only a tiny fraction of people will buy the 99 cent book.

I've had a 99 cent novelette out since December called "Hard Vacuum" and it's sold probably 10% of the second book in the fantasy series.

Permafree is now ubiquitous for a reason. It works really well. The readers get to try something with no risk and the author reaches more readers. It's a win-win for both sides, as long as you can write fairly quickly.

If you're writing one book a year, then a different approach may be better. Joanna Penn has had a lot of success without going permafree, for instance.


I also do the first book free in my Bytarend fantasy series. At the moment, the first book is free, the second book is 99 cents and the third book is $3.99. I've tried the second book at $4.99, $3.99 and $2.99 and the royalties were much higher but the buy rate was lower. I brought the third book out last month and dropped the price on the second to get more readers.

Since then, my conversion rate has been:


1st book to 2nd book: 4.4%

2nd book to 3rd book: 30.7%


1st book to 2nd book: 4.7%

2nd book to 3rd book: 47%

However, the first book gets much fewer downloads in the US (About a quarter), so the sales in the UK are much more significant for me. Kobo, B&N and the rest are practically non-existent. Smashwords is second to Amazon, but only with a handful of sales.

I had a BookSends promotion in November than boosted my visibility significantly (Sales went from a few per month to 1-2 per day) and then had an ENT promotion in March that boosted me again (Sales went from 1-2 per day to about 5 per day).

If you haven't done any promotion, Garrett, I'd consider it. However, if you're seeing that big a dropoff on Wyrmspire, it might be worth having a look at that before you do any promotion.

If you're interested, I can give you my two cents. I'm about to read the last two episodes of Wyrmspire now. Hit me up on email, if you want feedback :)

Share This