Why the ‘Author Earnings Report’ is Misleading


I’d seen these Author Earnings Reports flying around Facebook.  At a glance they seemed harmless — I shared with friends, all of whom were eager to pound on the Big 5 and how indie publishing is awesome, the way of the future and all.

This disturbed me.

I’m a data guy and a scientist.  And I don’t need to pimp my resume to point out three things about this report:  it has an agenda and it is biased.  And it could be a GREAT THING.

Sure, I could say that about any report.  I could point out flaws and the limitations; I generally don’t have the time or care enough.  But this report affects my friends and their career decisions.

So, first I’ll do a quick roundabout of the problems with the May’s report.  Then I’ll give my own analysis of the (limited) data.  And I’ll finish with a few suggestions to AuthorEarnings.com

Follow along if you like:  Link


  1. Have a Merry Christmas:  We’re comparing January 2015 to May 2015 for MOST of the data.  Anyone who’s taken a basic stats class will tell you this is stupid (seasonal variation).  Think about it — what do January’s numbers represent?  Has anybody ever heard of a Christmas break?  The holiday creates a huge spike in sales that MUST be normalized.  This detail alone invalidates most of the data.
  2. Lots of free cheap books:  January numbers are for 120,000 books and May is for 200,000 books.  That’s nearly DOUBLING the sample size.  And it says that we’re only looking at about half of the total sales.  Where are the rest?  Based on multiple references to “best seller lists”, I’d infer that they’re throwing out almost half of the data on purpose (because it’s difficult / impossible to collect) .  So if the number of big 5 titles decreased and the sample size doubles, where are all of these extra books coming from?  Indie published bundles and low-price books are the only answer.
  3. Are Big 5 sales really decreasing?  So, they’re only looking at half of the sales.  They admit that their sample size of Big 5 books decreased (by how much?).  So if your sample size decreases and your earnings decrease, you can’t assume that there IS ACTUALLY a decrease.  You can’t quantify the data unless you’re comparing apples to apples.  This is an unfounded correlation!  Truth is, we have no idea what the impact to Big 5 is…
  4. What about PHYSICAL books?  This is the elephant in the room.  We’re subjected to a bunch of hypothesis that Big 5 price increases are hurting authors.  Dumb.  Assume you’re the consumer.  Ebook prices increase and paper-books stay the same.  What do you do?  Basic economics people:  ebook sales decrease, print sales increase.
  5. Pie charts:  We like ’em.  They tell a story.  But they implicitly remove cost totals, making them unbased numbers.  We already discussed the huge disparity in sample size.  Of course your numbers move in favor of indie books — you’ve outright stated that your sample size doubled and your Big 5 population is smaller.


  1. More indies are entering the market.  This is obvious, since revenue is increasing and so is indie sample size.  But I’d wager that the increase is being spread out among A LOT of authors, and individual earnings aren’t growing (that much).
  2. Big 5 are disappearing from bestseller lists AT LEAST in part due to an increase in print sales.  Again, basic economics.
  3. The overall market is getting bigger.  May sales should have decreased overall, as ALL sales spike at Christmas.  But the indie authors aren’t cutting into the pie (market cannibalization).  They’re making it bigger.  Long-term, this means more readers.  FANTASTIC.


  1. Analysis is great.  Keep it up.  But the entire report seems based around two points — you want to be right and you want to promote indie.  That’s fine.  But can we have the data first and the analysis after?  With the current report layout, it’s all tied together to (seemingly) prove your point.  We get it.  You’re smart.  But can’t you give us a little wiggle-room to draw our own conclusions?
  2. Add seasonal variation.  You should have enough data, right?
  3. Give us more raw numbers.
  4. If you’re comparing data, try to normalize the samples.  That is, you can’t just compare oranges to fruitcakes and draw conclusions.  Maybe you’re drawing them from other data we’re not seeing — why don’t you show it to us?
  5. Figure out a way to show the impact of market forces to the top and bottom percentage of earners.  All of this data is useless unless we can see the impact to different author segments.  Total sales can be increasing, but if we get increase the population by 50 times and everybody makes less, your data makes it sound like a good thing.

I love this idea.  I love that we’re trying to help each other by sharing the data.  But I can’t help feeling in my gut that this private blog is  oversimplifying some things and potentially causing great authors to make poor career decisions.  Indie is great, and it’s growing the market, but the loss of the Big 5 would destroy book publishing today.




Add a Comment

Your email address will not be published. Required fields are marked *