While they make income, they are not your bread-winners, despite any individual bestseller. Your income comes from the teamwork of your entire bench.
Marketing your books has to take this into account.
The first thing to learn is to unlearn the conventional wisdom. If you'll pardon my crass assessment: there's lots of crap floating around in the book-marketing sewers passing off as pure gold.
Some lies:1. Reviews only work on Amazon.
2. Social media doesn't sell books. It's just an invitation to discovery.
Some empirical truths:1. No one distributor has all the possible buyers. Amazon's share has been dwindling for years. Every distributor has different niche-tribes visiting them.
2. Celebrities with huge social media followings can produce "instant" results no one else can duplicate. (While they spent years building that following so it could be exploited.)
3. Most books sell little, if at all, by themselves. Publishing a single book to a single distributor has almost guaranteed failure. Some individual books don't sell at all.
4. Authors with sets of books (series) sell books. Authors who spend their time writing instead of marketing make the majority of the income. (See Taleist survey. And DBW survey.)
These tests have proved out over and over.
Those surveys were backed up by studies of classic bestseller authors. Dickens was prolific, as well as Shakespeare. They had uncommon success, even though individual works sometimes failed. Current generation successful authors such as Louis La'Mour, Stephen King, and Amanda Hocking all use(d) this model: Find a niche genre and produce consistent works with feedback from your audience.
People who wrote and published a single or just a few books rarely made any personal success. Moby Dick became a bestseller almost a generation after Herman Melville died. Similar outcome happened for Jane Austen, Emily Bronte, Henry David Thoreau, and others.
Author success depends on a backbench of works. Anecdotal reports say that it's around the fifth book that the series takes off. Recently Howey's "Wool" series took off just at that point. Others hit sooner, some later. Hocking had already written 17 books before she hit it big.
The caveat is that my work down this line are simply and only a successful implementation of the backbench theory. The test was publishing over a hundred ebooks. Since they started selling immediately, this disproved the Amazon marketing data was commonly being promoted. In fact, they were selling 2-to-1 over the same books on Amazon.
With other limitations on Amazon at the time (their policies on Public Domain [PD] and Public-Licensed-Rights [PLR] books) I simply left that platform and pushed over a hundred books to the other distributors. After all, I was making far more income otherwise.
The only marketing these books had were a decent cover, description, and price (most at 99 cents.) The PLR books were given whatever covers they came with (as long as they weren't awful.) Descriptions were pulled from their introduction, usually. They sold best on GooglePlay, where Internet Marketing can take advantage of keyword-based titles.
PD books sold regularly across all platforms. I used what was estimated to be the all-time 25 top-selling books. Priced at 99-cents, there were a few who sold better than the rest.
One hit an algorithm circuit on Kobo - Jane Austen's Emma - and sold very well for several months, then dropped again.
When one of the supervisors at Kobo contacted me about her interpretation of company policy about public domain books (I couldn't find any policy on the site which backed up what she said - and it was contrary to U.S. copyright law) - I was forced to declare all these books as public domain and take the 20% royalty as a penalty for them. So I raised the prices to 3.99 for each book (which gave me a79-cent payout, which was up from the 43-cents I had been getting.) Interestingly, my income from that distribution platform increased. A couple more months of sales should give me some comparatives in this.
One of my books sold well on all platforms, but it really took off on Amazon once it got sufficient 4 and 5 star reviews. Soon that one book on Amazon (along with some peripheral sales) were bringing me twice the income of all the other distributors combined. When I added an audio version, this doubled that again.
Now that Amazon has clarified their public domain policies, I'll reformat the earlier PD books published on other distributors to align with Amazon's PD policies and post those there. Since they will instantly aggregate the reviews from the title, I should have a jump in sales and overall income, even though they force a 35% royalty on PD republishing. Again, reviews only work that way on Amazon.
Lulu has dropped all distribution of public domain books outside their own platform. This necessitated getting a MAC to publish to iTunes and start publishing to Nook on my own. So these tests are still in progress as well. Meanwhile, publishing PLR books via Lulu were successfully distributed.
You can see that this is more down the line of merchandizing rather than what passes for book marketing. But these tests proved that a person can make a decent living simply republishing ebooks - given enough books and a focused approach.
Other studies were undertaken and more tests worked up. I write this while in the middle of one of these (which I'll tell more about in the next installment.) What came up out of that study of copywriting books was a very successful online marketing company and how they've proved what really works regardless of the urban legends floating around.