Three months ago, we released our first full report on Amazon e-book sales and author earnings. Our goal was to look at unit sales and earnings by various publishing paths in order to help authors make informed decisions in this rapidly changing publishing environment. The results were eye-opening, but it was merely our first data point. Our long term goal has been to pull data every quarter to see if we can spot developing trends.
A quick recap on our methodology: Using a custom software spider, we can crawl every Amazon bestseller list and pull info from each book’s product page html. This data goes into a spreadsheet, which gives us the price, ranking, average review, and much more for every ranked e-book on Amazon. Using established ranking-to-sales data from numerous bestselling authors (including our own works), we are able to present author earnings by title and publishing type. As with our past reports, all the data has been anonymized and is available for download at the end of this report. And just like with past reports, any reasonable numbers entered for the power curve of the product rank-to-sales ratio reveals the same overall picture. That is, our conclusions are not dependent on our estimates but are borne out of the freely available data.
The exciting thing about pulling this data is that we have no idea what we’re going to find. Our conclusions since the last report might need rethinking. Our advice on what an author might want to do with a manuscript today could very well change as the publishing industry takes another swerve. My partner and I debated what we expected to see from this second round of data. We both predicted no more than a 2%-3% swing from any one publishing path to the other over such a short period of time. I wagered we’d see a 2% drop in self-publishing titles, offset by an increase in Amazon imprints, as the latter continues to snatch up high performing e-books and put more marketing muscle behind their own authors. My partner thought we’d see a 2% hike in self-publishing at the expense of traditional publishing. We bet a dollar on the outcome.
Over the next four charts, we present our findings from the last report and compare them to our current data set. To see how the breakdown has changed over the past three months, hover your mouse over each image (or tap it). You’ll see the February report jump ahead nearly three months to the April report.
This first graph shows the number of titles on Amazon’s e-book bestseller lists by publishing type:
Of note with this chart: The vast number of small and medium publisher e-books reflects their dominance in obscure categories. The overall ranking of most of these books is extremely low, which is why the number of titles on Amazon bestseller charts does not correlate to daily unit sales volume.
Next we have the estimated daily unit sales, based on the overall ranking of each e-book:
This is what my partner (affectionately known as Data Guy) predicted. Yes, I mailed him a dollar. What we see here is that self-published e-books and those from small/medium publishers have captured sales lost by the Big 5 and Amazon’s own imprints.
Using list price, we get the daily gross sales revenue:
Despite the lower number of daily unit sales, Big 5 revenue has gone up 1%. Investigating this, we found the average price of a Big 5 e-book went up nearly 3% between reports, while self-published e-book prices went up 1%.
Finally, the estimated daily author earnings report:
Here we see self-published authors earning 2% more while Big 5 authors hold steady. Again, with fewer daily sales but a higher price, the Big 5 publishers increased their own revenue while their authors stayed pat. Amazon Imprint authors saw a 3% decline, which might be due to the volatility they have as a small number of titles can see a large number of sales. What we can’t tell with these two data sets is if we are seeing real trends or just random fluctuations, but we can say that our findings from February have now been corroborated by this second data set. Self-published authors are clearly earning as much as traditionally published authors on the largest e-book sales platform in the world. A few months ago, this seemed impossible. It is already beginning to feel like old hat.
Once we have seven or eight of these quarterly reports, we’ll be able to look at yearly trends. But even with two data sets, we can now ask questions that were impossible with our first report. One thing our first report couldn’t show was the volatility of these lists. How long were e-books holding on to their rankings? Was this volatility dependent on how one chose to publish? The first thing we wanted to explore was churn, or the rate at which books fall off the list as new books rise up to take their place. By pulling the publication date of each title, we are able to compare this report to our last report to look for duplicate titles and original titles.
We limited our analysis to books earning at least $10 a day. This is a very low threshold, but it prevents the data from being swamped by those books only selling one copy a day or every other day, which makes up a good number of some obscure product categories. In the graphs below, green represents the number of titles where revenue increased 25% or more. Red shows a decrease of 25% or more. In the middle, you have books that stayed within 25%:
This graph shows higher volatility for self-published titles, with more books moving up and far more moving down. Traditionally published titles, on average, have more stability. But a deeper look at the data shows something even more interesting:
Titles by long-established Big-5 authors (those who debuted prior to 2009) are the ones which have all the stability. New releases by tenured authors, as well as backlist titles from those tenured authors, make up the vast majority of the Big-5′s market share and the of Big-5′s stable sellers. Traditionally-published titles by newer authors, with Kindle debuts after 2009, show far less stability, and exhibit very similar volatility characteristics as titles by self-published authors.
The next thing we wanted to look at was the staying power, or link between daily author earnings and yearly author earnings. More data points will make this clearer, but we can already see something very interesting. We took the top 1,000 earning authors for each publishing type and ranked them by combined earnings from both of our data sets. This allows us to look at earnings falloff over the one quarter for which we have data. If a book had dropped substantially, it would drag the average for that publishing type down. This is another way to look at churn, and this time we are using a higher threshold than the $10/day cutoff from above:
What the data shows is that the top 1,000 self-published and traditionally published authors are equally affected by the churn of the bestseller lists. The falloff toward the long tail normally seen with demand curves is also less than might be expected. While the extreme outliers from both camps earn most of the income–similar to what we see in most entertainment industries–there is health and wealth down the long tail for self-published and Big 5 published authors alike. In fact, they almost perfectly map onto one another.
Next we looked at author revenue by publication date. Self-publishing gurus have long discussed the diminished sales outside of the first month of publication, as e-books fall off the “hot new” lists. For that reason, we broke daily sales down by month of publication, giving us 30-day snapshots. We were also able to capture preorder data for books not yet released.
We should start by pointing out how much influence a handful of titles can have on this data. This is due to the blockbuster model that plagues Hollywood and has come to dominate traditional publishing. The big purple bump on September of 2012 is 80% due to George RR Martin’s GAME OF THRONES Box Set. May of 2012 owes 50% of its bump to Veronica Roth’s INSURGENT. The following month sees 60% of its increase due to Gillian Flynn’s GONE GIRL. And a whopping 70% of the leap in January of 2012 is from John Green’s THE FAULT IN OUR STARS. 80% of the huge October 2013 bump comes from Veronica Roth’s ALLEGIANT and the DIVERGENT Box Set, Donna Tartt’s THE GOLDFINCH, and John Grisham’s SYCAMORE ROW.
Major publishers aren’t just reliant on these rare blockbuster bestsellers, they are also heavily reliant on their backlist, which the above chart also captures. Look at that massive bar on the far left of the graph, which shows all books published before 2011. Ignoring pre-orders, the author revenue for Big 5 books published prior to this date is equal to 31% of the earnings on all books published since then. This backlist dates to before the self-publishing revolution gained steam, which suggests that the share of pie earned by self-published authors has a lot of room to grow as these writers establish backlists of their own. It also means that the parity we see in our author earnings charts between self-publishing and Big 5 publishing has a lot to do with the latter’s existing titles and not their new releases. How you decide to publish your manuscript today means looking at the difference in earnings due to recent works. Self-published authors are not just holding their ground with Big 5 authors when it comes to releases after 2011, they are out-earning Big 5 authors by a 27% margin.
In fact, we can see this if we look at author earnings just for books released since January of 2011:
Does this chart indicate where we are heading with author income? As self-published authors build their own backlists, and as more of these authors employ freelance editors and cover artists and pay attention to quality, could there be a future where self-published authors as a cohort are earning a good deal more than traditionally published authors? Our advice to an aspiring author today might be to do one of two things: either build a time machine and travel ten years into the past to query their work — or self-publish today.
And while this analysis is just for e-books, revenue from e-books is already 29% at Simon & Schuster and 34% of overall trade revenue for Hachette (40% in the UK for Hachette). Also, various estimates place 50% of print revenue from online retailers, a market self-published authors have equal access to with their print-on-demand books. As the market for audiobooks continues to increase, the publishing world tilts further and further toward the digital. We have said before in past reports that there has never been a better time to be a writer. It could be that the best of times are yet to come.
Next Monday, we’re releasing a follow-up report that looks at this same data from the perspective of an author getting their start today. We’re going to analyze the difference between tenured authors and newly published authors. The decision on how best to publish will never be an easy one, but our next report might make it a whole lot simpler. Stay tuned. We’re just getting started.