At this point, the second biggest reason (after Stockholm syndrome) that authors shun Amazon in favor of legacy publishing is a lack of information. Two recent articles at Mad Genius Club have helped inform authors about the risks and rewards of tradpub vs. indie; thus helping them make better decisions.
The first post, by Amanda Green, points out the logistical flaws in the old saw that authors must be anointed by legacy pub gatekeepers as proof that they're good enough to self-publish. The most obvious hole in this theory is the fact that trad publishers lack sufficient slots in their lists to publish all of the books they deem worthy. Then there's the fact that the old gatekeepers haven't had a very strong track record for picking winners.
Next up is Sarah Hoyt's expose on most publishers' archaic, convoluted, and sometimes highly suspicious accounting practices. I can't name many better arguments in favor of indie than legacy publishers' lock-step practice of basing royalties not on how many books an author actually sells, but on third party estimates extrapolated from sales reported by certain retailers in certain regions.
Oh, and those sales are reported by an automated system that's notorious for being grossly inaccurate.
In contrast, Amazon issues crystal clear royalty statements based on the exact number of ebooks sold, and unlike legacy publishers they do this monthly instead of semi-annually.
Back on the subject of gatekeepers, I respectfully took issue with some advice that Brad Torgersen gave in the comments of Amanda Green's post.
I greatly respect Brad. He worked his ass off to get where he is (which is deservedly far above me, so weigh my objections accordingly), and he's proved his integrity a thousand times over in the course of Sad Puppies 3. On the other hand, when the context is helping authors make informed decisions, it's the information that counts.Someone I respect once gave this advice: before going indie, be sure you’ve written at least 500,000 words of fiction (lifetime) and are getting personalized rejections on a regular basis. This was how you knew you were at “entry level” professional craft.
In the interest of charity, the advice from Brad's mentor echoes what I've heard from countless pros who broke into book publishing through the short story market. That's how Brad got started, and it certainly worked for him.
However, my reading of the data says that Brad is probably among the last authors who'll be able to follow that path. If sales of major SF magazines continue their downward trend--and there's no sign of an upturn on the horizon--there simply won't be anyone to get personalized rejections from.
As for the 500,000 word benchmark, it's basically a repackaging of the 10,000 hour rule, which has been thoroughly debunked. Though it's an unpopular opinion in our hyper-egalitarian times, natural talent is real; and those without it will never have the effortless mastery needed to go pro, no matter how much they practice.
So if up-and-coming writers won't have the venerable sci-fi mag editors to act as kingmakers, who will tell them if they're any good? The answer is the only people whose opinion, in the final analysis, really matters--the readers.
If you're crazy and thick-skinned enough to try your hand at this business, by all means keep writing till you've typed END on a first draft. Assemble a team of trusted beta readers and hire a freelance editor to help polish your manuscript. Then put the results in your readers' hands. They're perfectly willing and able to tell you what they like or don't like about your work. If multiple people offer the same criticisms, you'll be wise to take the hint.
After all, it's the readers--not publishers or editors--who a writer works for.