This Amazon review business (oh, and advertising)

And yes, reviews are definitely a business! Though not necessarily for me! {8′>

As I’ve mentioned before, I really didn’t expect anything at all when I (finally!) published The Secret: A Thriller back in September. But it sold fairly well for a while (especially in the UK!), and it was probably borrowed many times by Kindle Unlimited users. My other novels got a minor boost too.

I don’t really know about the number of borrows. Amazon no longer gives out that information, though I think it’s probably used to calculate “sales rank.” But I do know that many thousands of pages were read by KU or Amazon Prime members, because that info is available. Given that the sales rank never got into stratospheric territory (and is now pathetic!) I’m going to go ahead and guess that the number of borrows was in fact not too high, but those who did borrow the book mostly read it all the way through. This is not bad news! In fact I’m so encouraged that I’ve been inspired to make lots of progress on a sequel to Shiver on the Sky–mostly because it’s going to turn into a trilogy at least, and the likely-useful forms of book marketing of which I’m aware typically seem to work best when a full-on series is available.

Still, there’s this information gap. I’d given up completely on ever selling Shiver again, and mostly for that reason hadn’t worked on its sequel(s) for a while. But it did somewhat better after The Secret came out…so I decided to try running some cheap ads on Amazon via Amazon Marketing Services (AMS) to see what I could discover.

Guess what? Both books got about the same click-through rate (CTR) at a given price (roughly 1.1% to .6% depending on price listed). But a much higher percentage of folks who clicked through to the books’ detail pages actually bought Shiver (this percentage varies by price point…anywhere from 20% for Shiver at $.99 to slightly over 1% for The Secret at $3.99). Now, that could be due to any number of things. Book cover? I’m guessing that’s not it, because after all the CTR was about the same, and the ads showed a thumbnail of the cover. Price point? Nah, because they were mostly identical and I’m controlling for that. Price I bid per click? Nah, that didn’t seem to make much difference to anything–except the number of impressions, of course, but even then it wasn’t as straightforward as I’d expected. (FWIW, anything between $.09 and $.25 seemed to get impressions, but it varied so much day to day that I don’t yet have enough data to draw any conclusions.)

So that leaves two obvious candidates to explain the difference I saw in the data: book description and number/quality of reviews. The book descriptions are in fact very different, in both content and structure. But so are the review-counts. The Secret has three (3) reviews. All favorable, but hardly impressive in number. Shiver has 23 reviews–gained over a longer span of time, and with the first listed being posted by an Amazon Top 500 reviewer (thanks again, Dianne!) (See how I suck up?) (Oh, and the relevance of the ad’s headline to the book’s description might be a third factor, but if anything I’d say Shiver‘s the loser on that front, so if it’s a factor at all it’s strongly outweighed here.)

Incidentally, if you’d like to check out the current book description pages and see if you come up with another theory? Here’s Shiver. And here’s The Secret. Normally, for Amazon US at least, I use affiliate as well as Smile-compatible links…but I can’t do it here because Amazon says it’s a violation of their Terms of Service to send ’em out via email (most of you) or RSS (some of you). You’re welcome.

I mentioned before that I’d love to get information from Amazon about conversion rates on detail pages. It would let me do some testing of book covers and descriptions, yes? And I could modify them? Well, I’ve done that. Or some of that–The Secret might still benefit (or not!) from a more conventional description (I based the idea for the one I’m using on something Blake Crouch once did for Run–though I note that his current description is very different). (Oh, and guess what? Headlines help!) But I decided to focus on reviews for a while.

What did I find? Well, lots of stuff. The Author Marketing Club has a tool that lets users search for a book on Amazon and then automatically extract the reviewers’ contact information if they chose to provide it. The idea is that you could then export the data into a spreadsheet and email the reviewers to ask if they’d be interested in reviewing your book. This sounds to me like spamming people, though if the tool also provides a link to each reviewer’s profile you could go and read some of their stuff. It’d result in better targeting, more-informed review requests, and probably a better experience for the reviewer…but you know what? Maybe all that is a waste of time. Maybe it simply pays off better to spam people. Maybe those who respond are more desperate for freebies, or maybe they actually are intrigued by your book description, or maybe it’s worthwhile from a strictly-business perspective to take this approach even if it offends people because it saves so much of your time. Which you could then use by spamming many more people! After all, what’s the downside? Few potential reviewers will actually go through with an evil plot to get a freebie and then trash your stuff via a one-star review…right? After all, they could do that without even reading the damn thing. I don’t think it happens very often.

I also found some online services that sell fake reviews. Let’s just assume that’s a bad idea, and move on.

Yet other sites offer a review service–you pay them an administrative fee, and they find reviewers for you. The reviewers are unpaid, and the site keeps looking for more (either via sending freebies to book bloggers or by growing their own stable of regular reviewers) until you get the number of reviews you’ve paid for. This actually seems semi-okay…but are the reviews truly objective? A book blogger, it seems to me, is not as likely as a casual reader to post a negative review. First, they want to get more books to read. Second, if they care about Amazon reviewer rankings, those are apparently more strongly affected by “negs,” or “not useful” votes, than by the reverse–and negative reviews attract down-votes from an author’s fans. So who needs the hassle? And a stable of book reviewers used by a review-hunting service have to be affected, at least subconsciously, by the fact that the service will get a better reputation among authors if the reviews are generally positive. I don’t mean to imply that either group has any sort of unpleasant ulterior motive–I’m just looking at their incentives and assuming they affect human behavior.

So that deal might be okay, I guess. It’s probably even compatible with Amazon’s current rules (I haven’t checked), as these services claim.

But. Is soliciting reviews at all a good idea? I don’t know! I’ve done freebies for that purpose before, and enjoyed the process, and after all it’s not so dissimilar in concept from Amazon’s own KDP giveaways. And maybe it’s a way to get people to at least consider your book if you have nothing else going on.

Thing is, I want to do a major promotion of Shiver sometime before January, after I have at least one sequel out and have print versions of all my books available. I’d like to use Bookbub and some other services for it. How the hell do I know whether a bunch of reviews that state the reviewer received a free copy from the author will obscure the reviews Shiver already has, and possibly mean it’ll no longer be considered eligible for promotion with the bigger sites? How do I know whether there are currently enough reviews to qualify? And if the reviewers don’t mention that they got a free copy, what are the potential long-term consequences to me for violating Amazon’s Terms of Service? (And, separately, isn’t my timeframe sort of stupid given that lots of people will be competing for slots near Christmas?)

I simply don’t know. I’ll probably just leave stuff alone, except for contacting a few people who’ve previously reviewed my stuff and offering freebies. I kind of want to do that anyway, regardless of whether they post any more reviews, because I’m grateful. But, yeah, I’m hoping.

I ran into some other stuff too. One service contacted me this morning after I submitted Shiver to ’em (they say they want to do a review before agreeing to promote anything). They said they liked the book’s cover, description, and other stuff. But they also said it needed to have “tags” added (which Amazon discontinued years ago) and said there were keyword problems because they tried “eight major keywords” (emphasis theirs) and didn’t find my book thereby. Well…duh. The book’s ranking is currently so low that it wouldn’t show up on anything but a search by author name or title anyway.

But. I got some more email, and discovered they recommended putting keywords into the book’s description, and possibly also into external pages linking to my book’s page. I have some problems with these notions.

First, it’s never to my knowledge been demonstrated that keyword-stuffing is all that helpful in gaming Amazon’s search results. In fact, given that I do have actual conversion-rate data, I’m unwilling to discard my tweaked descriptions based on an unsubstantiated theory–though, yes, I could always write new descriptions and test them too. I might even do better! But I simply won’t act solely on the supposition that it’s “supposed to be” better to do so.

Second, about those external pages? I’ve run into this idea before. But I’ve designed systems before, too. If I were setting up Amazon’s system, I wouldn’t be likely to crawl external pages to find out what sort of keywords were on them. I’d be much more likely to simply track the number of views a particular page receives, and also the number of sales (or borrows), and from that I’d get a conversion rate. I’d be more likely to give a boost to book detail pages that actually convert well than to rely on an algorithm that could be so easily gamed. Though here’s a caveat: maybe, if lots of people are actually clicking through to a detail page from somewhere else? And they’re converting to actual sales at a good rate? Then I might want to give that book a boost for a while just to see if it’ll work. And I might even index that external site’s page to get more keywords just to find an excuse to show the book’s page even more often! But. Unless people are actually clicking on the link to come to my site (because I’m Amazon here), what the hell do I care that somebody posted content that isn’t doing anything useful for me?

So again I’m theorizing in the absence of data. But I’m sort of disinclined to give people like that my money. If I don’t know the answers here, why would I think they do? Given that they want me to pay them, I think I’ll just do something else instead.

Some more stuff I’ve observed:

• Most of the money I’ll make since The Secret‘s release (so far) has come from “borrows” vs. sales. It’s not even close. Maybe that’s because I’m still a newbie author, and people are more willing to take a chance on me if it doesn’t actually cost them money. Or maybe it’s that the people most likely to read stuff by indies have generally subscribed to KU by now?
• My sales in the UK were pretty damn good–much better than Amazon US in every possible way–right up until I increased the price beyond $.99. Then they fell off a cliff and never recovered. (Plus I got a nasty and very short and oddly punctuated one-star review from somebody who has never reviewed anything else…but also a very short four-star and a longer 5-star, so who knows what had which effect?) (Also, if you’re one of those reviewers and you’re reading this? Thank you!)

I have a theory that Amazon UK is still counting sales of low-priced books as equivalent to sales of high-priced books for purposes of determining sales rank, which many people believe is no longer true at Amazon US. If so, maybe–and this is weird–I should actually aim my future efforts at those guys. Amazon’s KDP site strongly suggests using their $2.99 price point for everything. They have this dingus that claims to compare your book to others and give the price that optimizes author income, but AFAICT it always gives the same recommendation. But you know what? That same dingus shows there are clearly more sales at $.99. And my job now is not to maximize income. It’s to attract new readers!

I have become skeptical that people who download freebies are likely to pay for my other stuff. But people who pay, even $.99? Maybe not so much. And of course my heroes are the KU users. Their borrows improve my page rank, and if they actually read the book I get more money than I would if they’d bought it–at the cheaper price points, anyway. (So, putting all this together, I see no reason to leave KDP Select and seek sales outside Amazon at this point…except possibly for Pagan Sex, which I’m considering making “perma-free” for a while, mostly to see if it can gather some legitimate reviews.)

That’s it. It’s all I know. If you have other info, or theories, I’d love to hear about it.

Have fun out there!