How KDP Select Made Me a Bestselling Author

It's no secret I'm a big fan of KDP Select. The first time Idid a free promotional day, I was blown away at how many downloads the bookreceived.  I was also pleased with theresulting sales a week later.
But that was nothing comparedwith my second experiment with free promotional days. Instead of building thesuspense, let's go straight to the numbers—and then I'll explain why I had somany more sales the second time versus the first. Below is a chart of paid booksales for the first seven days after my free promotional event.
As you can see, the free promotion in February vastlyoutpaced the one in January. In January, I sold 73 books in a single day, atthe time a record for A Soul to Steal.In February, I sold 351 books in a single day. I had two other days where salestopped 220. All told, I sold 246 books in the seven days after the freepromotion in January—a number I was very pleased with at the time. In contrast,I sold 1,250 in the week after the February promotion. (Side point: I priced the book at 99 cents for two days following the January promotion, but I did not lower the price in February, leaving it at its regular $2.99.)
So, what changed? Why was it so much more successful thesecond time than the first? While I can't say anything with certainty, I have afew educated guesses as to what happened.
1)                 The book gota new, more professional coverOf all the factors, I believe this is the most significant.I liked my old cover, which you can still see here, but it screamed self-published. Accordingly, while folks had no hesitation indownloading it for free in January, they didn't rush to buy it afterwards.Sure, it got a boost, but not nearly what it would have likely received had Ichanged my cover earlier. Please take note of this, indie authors. If salesaren't where you would like them to be, your first thought should be to look atyour cover. If it doesn't look like a Big Six cover, it's time to get a newone.
2)                 I ran thepromotion for two days instead of oneFollowing the advice of several other authors, I ran mypromotion across two days (a Tuesday and Wednesday), rather than just one. I'ma little unclear on exactly how much this helped, mostly because more than10,000 people downloaded the book on the first day and "only" 5,000 downloadedit the next day. Still, given how much weight Amazon's algorithms put onpopularity, the combined downloads were enough to keep A Soul to Steal front and center in the Kindle store over theweekend. Had I just done a single day, I'm not sure that would have happened.
3)                 I engaged inmarket layeringIn January, I stopped all advertising prior to the freepromotion. I had been using Facebook ads successfully, but was hoping the KDPSelect experience would free me from being dependent on them. In February, Iwanted to follow Jeff Bennington's advice and "layer" marketing following myfree day promotions. (If you want more details on this strategy, check outJeff's excellent how-to guide "The Indie Author's Guide to the Universe" here).As a result, I turned on the Facebook ads again in a big way, spending $50 aday and using my two most successful ads that I had. I also used World LiteraryCafé's Social Media Mania to promote the book. It's difficult to know how much impact this hadindividually, but it's not a coincidence that the same day the promotion ranwas also its most successful. A free excerpt received more than 2,000 hits onthat day. Was it worth the $40? Easily.  Overall,Jeff is right. If you want to ensure your book's success after using a freepromotional day, your best chance is to layer on marketing as heavily aspossible.
4)                 I changed mygenre and hit several bestseller lists If you look at the January numbers, you can see the boost insales fell off rather quickly. After Day Five, they largely fell back into"normal" range. But after getting such large numbers in the initial days afterthe February promotion, my sales continued to be far stronger than normal for amuch longer period. Even now, sales are decidedly better than prior to the freedays. Why? Well, for starters, the novel hit #1 on Amazon's Ghost bestsellerlist, and #4 on all of Horror.        Over that critical weekend, it also charted highon the Suspense, Thriller and even the very competitive Mystery and Thrillerbestseller lists. This actually brings me to another point. In January, my bookwas slotted under Ghost-Horror and Occult-Horror. But in looking at the Kindlestore, I felt like I was narrowly limiting the novel's exposure. Yes, the novelqualified under Occult, but I was essentially doubling down on Horror. Honestly,I've never been entirely comfortable slotting A Soul to Steal there anyway. It's a mixture of several genres,including mystery and thriller. Looking at similar novels by Blake Crouch andJoe Konrath, I decided to switch genres, abandoning Occult and moving toSuspense. This can be risky, and you must make sure you aren't changing genresarbitrarily, but the novel truly fits under Suspense. As a result, I likelyreceived more downloads on the free days from people looking for thrillers, andnot necessarily horror books. And I definitely received more exposure byappearing on those bestseller lists. Once I was part of those lists, the bookcontinued to receive higher than normal sales even after the initial boost fromKDP Select was gone.
Nothing, of course, lasts forever. Although these numbersare very exciting and I continue to have higher-than-average sales, if you lookat the bestseller lists, A Soul to Stealhas largely tapered off. I also have no idea if this kind of success is repeatable.It's possible that every free day brings less of a bump because so many peoplehave already downloaded the book. I may look back at this as the novel's high point.
But all in all, this experiment has been a gigantic success.I've had more positive reviews in the past week since the promotion than at anytime since its initial publication. I'm receiving regular Facebook postscomplimenting the novel. Beyond sales alone, the novel has clearly attractedmore readers—many of whom are excited about a forthcoming sequel.
I can't guarantee others will have the same experience. J.D.Currie has a nice round-up here of what others have reported,and it's clear there is variance. But I will say this: Unless you are sellinggreat numbers on Nook, iBook or Smashwords, there is no good reason not to trythis route. It might not pan out, but if it does, you could find your novel issuddenly a bestseller.