Kindle Forum banner

Status
Not open for further replies.
41 - 60 of 169 Posts

·
Registered
Joined
·
4,646 Posts
Patty, did you notify KDP of the pending BB? Just wondering.

Also wondering where the threshold is. Best I've done so far is 170 in the free store (but mine went up gradually over a few hours), and no problems there. Makes me wonder if the bot is looking for anything which jumps straight into the top 100 or top 50, with no intermediate movement first.

 

·
Registered
Joined
·
4,279 Posts
LilyBLily said:
I imagine she's not feeling in a mood to appreciate anything to do with Amazon at the moment.
Been there, got the T-shirt. Plenty of things to disparage Amazon for, however, without dissing the *people* who have expectations of its algos. Besides, if generic-you doesn't "believe" in algos (that they play any appreciable role in a book's success), then why should a little rank-stripping put you out of sorts?
 

·
Registered
Joined
·
1,433 Posts
MaryMcDonald said:
I am seeing your book at #5 right now, but only on the free list, not on your book's page. Maybe it's coming back?
Showing at 5 in the free list AND book page for me.

Listing at #2 in the Canada Store, list and book page.
 

·
Registered
Joined
·
13,308 Posts
Discussion Starter · #46 ·
PhoenixS said:
Who ever equated algorithms with magic? Companies spend serious money creating and tweaking their algorithms. Interestingly, I too can pinpoint cause-and-algo-effect for our 1.5M+ sales (and I've managed promos for trad-pubbed, award-winning USAT and NYT authors who've written beloved books that have stood the test of time). The same algos that track rank drive current and future visibility. Generic-you can optimize the algo-effect with a great cover, a hooky blurb and a solid story -- of course. Just as algos can optimize word-of-mouth for a good book. One doesn't preclude the other, however; they can work together or individually.

But magic? No. Who's thinking that? Amazon (as well as Nook, iBooks, Google Play and Kobo, btw, just to a lesser extent) *does* market books via its algos. That's demonstrable and replicable. I'm a bit surprised a SF writer doesn't appreciate the science behind marketing and sales.
I don't disagree with this at all.

What I do disagree with is the ultimate level of importance a lot of writers ascribe to the "Amazon magic" (i.e. its unseen recommendation engine). The only time it becomes important is when your book goes sticky. And no one really knows how to do that.

Yes, you can do it with ads, but that just proves my point. It's something the author does.

Interestingly, I see no slowing of downloads beyond what I would expect for this time of day. I'm about to hit the 20K. I suspect this is because the main source of downloads is the Bookbub email, which is rank-agnostic. If this keeps up, maybe tomorrow's downloads might be affected, but from my own email list operations (over 51k subscribers across multiple lists) I know that many many people delay opening emails for days. This is why I usually leave the Bookbub advertised price for a month (until it falls off their website).

Second day downloads are *at most* going to be a quarter of today's. Of course, UK, CA, AU and the others are not affected.

About data science and book sales in general:

I've been reading a lot of stuff recently about this subject. I come from a background of population biology, which gets referred to as "the science of numbers". It's about the proportion of a large population doing x and another doing y. While you can measure x and y and draw interesting conclusions from it, interpreting these conclusions and modelling projected developments becomes a lot more rubbery.

You only have to have run a couple of large simulations to know that when you change a tiny variable, it can potentially throw your entire model out of whack. If you're predicting future behaviour of pretty much anything (animals, people, plants, the stock market, you name it), the more detailed your model becomes, the more it is liable to contain assumptions (even if they're currently borne out by data) that will throw it out of whack.

Modelling is interesting for gaining understanding in how various aspects of processes are related. It goes without saying that the quality of the model is entirely dependent on the quality of the mathematical functions that control each variable.

Take applied finance. The models are impressive. They're scary. The maths is crazy.

Why then did a university lecturer tell a student, after a question whether they used the models to predict the behaviour of the stock market: nope, we use the model to perfect our understanding afterwards.

OK, so to take this back to data science and book sales. Some of the stuff I've read recently on how people use data science for book sales makes my hair stand on end. Not because it's "wrong", not at all, but because the models rely on base assumptions that are (often by necessity, so not blaming anyone) already of poor quality, or that are extremely volatile and a small change could turn upside down the entire resulting prediction. Therefore, to my mind, it's useful to study after the fact and much less useful to predict future events.

And if predicted cause-and-effect coincides with what we already know, or "common sense", the importance of the model becomes much less important to me as author. In practical terms, I've always included on-sales of second and third books in ROI from ads, because that makes sense. I don't know how much stock I'd put in some of the elaborate methods I see mentioned for calculating it. I can do a wet-finger prediction in three seconds flat and spend the hour I saved writing. Maybe I could more accurately calculate my on-sales as 53% vs my estimated 50%, but because I don't really have enough data to make my calculations statistically reliable, I'm going to keep doing the wet-finger stuff until such time that I do have that data.

Make no mistake, I love the AE reports, because they do exactly what this science was meant to do: make sense of past events to allow us to make global decisions about the future.
 

·
Registered
Joined
·
13,308 Posts
Discussion Starter · #47 ·
TimothyEllis said:
Patty, did you notify KDP of the pending BB? Just wondering
I did. But, as I said, KDP is just part of a big company that just rolls on like a huge ship and takes forever to change course. You cannot expect them to take notice of this.

KDP tech Help is a call centre Cape Town, and last time I spoke to them, it sounded like there were a lot of people there. They would need to have to be VERY organised in order for this sort of stuff to be remembered.
 

·
Registered
Joined
·
2,711 Posts
Seneca42 said:
Let's put aside I don't think anyone is denigrating anyone :)

That said, just on point of fact, the two camps are not equal in terms of industry impact. KU authors are commodifying (ie. driving down margins) for the industry as a whole. Wide authors are not (although I guess one could argue that the permafree authors are; but realistically they are only permafreeing one or two books max, so their impact is negligible).

I 100% agree that everyone should do what is best for them. But at the same time, if in doing so you (and others) are contributing to the erosion and instability of the industry, driving down pricing power for authors in the aggregate, empowering zon to become a monopolistic powerhouse, etc. - you should at least understand those who have an issue with that (not with YOU, but with the general trend of kowtowing to zon)... even if, and though, they nonetheless respect your right to do whatever you feel like (and most probably understand why authors feel they have no choice but to play the game the way zon tells them to).
Then make noise to the other channels to provide more tools to improve discoverability. Make noise to them to drop the curated placements. When I can make more from KU in a month than I make on all other platforms combined due to lack of discoverability on those other platforms, then I have a financial decision to make. As much as I hate being exclusive to one channel and as much as I would love to go wide, I can't pay my bills with good intentions.

But if you think telling exclusive authors that they're contributing to the erosion of the industry and accusing them of having Stockholm Syndrome will win you any converts, you're barking up the wrong tree.
 

·
Registered
Joined
·
2,113 Posts
Perry Constantine said:

But if you think telling exclusive authors that they're contributing to the erosion of the industry and accusing them of having Stockholm Syndrome will win you any converts, you're barking up the wrong tree.
Why are you making this about you?

If you want to have a wide vs. KU debate why not start another thread? It's a topic we're almost all interested in and will probably continue to be interested in.

This thread is specifically about Patty having a BB deal de-ranked by Amazon. Yes, there are some not-positive feelings towards Amazon in this thread.

Stockholm Syndrome might be a strong terms, but so is the term that was popular for a long time: Amazon Derangement Syndrome. ADS. We're authors. We like to use lots of descriptive words.

Find another thread that's more pro-KU or stop looking for reasons to be offended.

Lord knows we're not all in lockstep here at KBoards and you'll find many people who agree with you.

I don't see people "barking" at your tree specifically or trying to convert you!
 

·
Administrator
Joined
·
10,097 Posts
Folks, let's leave the KU-vs.-wide argument for other threads. I read Patty's comment about "Amazon Stockholm syndrome" as pertaining not to KU authors but to people who just loooove Amazon and see no problems with the company whatsoever, a (probably small?) group that likely overlaps with KU authors somewhat but is not identical with them. So there is no reason for that argument to blossom here.
 

·
Registered
Joined
·
2,711 Posts
HSh said:
Why are you making this about you?

If you want to have a wide vs. KU debate why not start another thread? It's a topic we're almost all interested in and will probably continue to be interested in.

This thread is specifically about Patty having a BB deal de-ranked by Amazon. Yes, there are some not-positive feelings towards Amazon in this thread.

Stockholm Syndrome might be a strong terms, but so is the term that was popular for a long time: Amazon Derangement Syndrome. ADS. We're authors. We like to use lots of descriptive words.

Find another thread that's more pro-KU or stop looking for reasons to be offended.

Lord knows we're not all in lockstep here at KBoards and you'll find many people who agree with you.

I don't see people "barking" at your tree specifically or trying to convert you!
Thanks for the reminder of why I don't come here as often as I used to.
 
G

·
1) This, EXACTLY:  Maybe the programmers have factored in BB  :p

1b) And if you go 99c instead of free, you have nothing to worry about at all.

Box Set Rank stripping a few weeks ago. Are they less frequent? The 99c rank stripping incidents?

***********************

2) This: It's happening often.  Could it be a stab at BookBub?

It's 1000% a 'stab at Bokbub' and at other promo sites

************

3) This: I don't see it so much on kboards, but then, I don't read every thread or anything.

No, on KBoards authors choose to give the algorithms benefit of the doubt by saying 'it's being done to target scammers' when it's becoming obvious that it has more to do with hurting other marketing venues

***********

4) Here's an example of giving benefit of the doubt

Patty writes:  You cannot expect them to take notice of this.

It's actually the EXACT opposite

first algorithm delays and tweaks were done to hit all the small promotion sites

now they are going after even Bookbub

Indie authors are writing things like

- it's just a false positive
- they don't even notice this
- it is to target scammers

This is a company that built a specialized Cloud Service for the CIA 4 years ago. You expect us to believe they can't stop a handful of botters

The other defence is that they just don't notice. Actually they notice two things very well

A) That promotion sites are growing very fast

B) That if money flows away from promotion sites it flows towards their own marketing avenues
*******************************

 
G

·
I've just suddenly cottoned onto this issue about rank-stripping.

Whenever I've made my book free in the past it has ALWAYS dropped the ranking from my normal daily setting, and re-ranked it under the free book graph. Only when it was set at $2.99, but Amazon was price-matching it to Free did it continue to show in normal ranking. I thought it was perfectly acceptable practice by Amazon. Maybe I was trialling this new 'rank-stripping' before it was rolled out without even knowing about it. Does it happen on NON-FREE books as well?
 

·
Registered
Joined
·
3,651 Posts
Em, for those advancing the ludicrous theory that Amazon is targeting BookBub, please explain why Amazon imprints use BookBub almost every day.
 

·
Registered
Joined
·
3,651 Posts
Patty Jansen said:
What I do disagree with is the ultimate level of importance a lot of writers ascribe to the "Amazon magic" (i.e. its unseen recommendation engine). The only time it becomes important is when your book goes sticky. And no one really knows how to do that.
There have always been authors who have declared Amazon's system unknowable, and there have always been those who have been working away and figuring it out and putting that knowledge to work.

When I started self-publishing in 2011, there were people who claimed Sales Rank was unknowable. Anyone want to still claim that today?

When we started figuring out the Pop List the same kind of people claimed it was unknowable (or irrelevant). We figured that out too.

Quite a large part of the Amazon recommendation engine and the various algorithms feeding into it have been mapped out at this point, by various authors, who have then shared that knowledge publicly - which has then been applied by authors who have optimized their marketing and sold millions and millions of books on Amazon.

Just a small data point.

(And, yes, I can make a book sticky. I've done it plenty of times, as has Phoenix and many others. The last time I made a book sticky it was below 200 for a whole month, below 800 for a month after that at $2.99 too. This led to a KU All Star and close to 7m KU page reads, and an equivalent amount of sales.)
 

·
Registered
Joined
·
3,589 Posts
Um, my website is getting hacking attacks from Russia and other countries. I think that's the global standard now. Unless AMS reaches a lot more people who click and buy/borrow, people will keep using external promotion sites because large lists of people who actually want those types of books is a far better plan of attack than sticking your book somewhere in the ad section and hoping the person has ads unblocked and bothers looking at them.

Sucks about the rank stripping, but at least it didn't stay stripped.
 

·
Registered
Joined
·
3,651 Posts
I keep repeating this:

I've had direct reports of something like 40 different authors who were ranked stripped. Most shared comprehensive details of what promo was involved.

1. Some were in KU.
2. Some were wide.
3. Some used BookBub.
4. Some did not use BookBub.
5. Some were free.
6. Some were paid.
7. Some used other ad sites.
8. Some used no ad sites.
9. Some only used FB ads.
10. Some only used AMS ads.
11. Some only hit their list.

So for anyone who thinks this is a potshot at BookBub, an attack on non-KU authors, Amazon undermining all ad sites, or forcing people to only use AMS, or whatever, the evidence simply does not support those claims.

Please stop repeating them and causing undue panic.

I can't spot any real commonalities in these cases after examining all the information provided, other than perhaps they were all (or almost all) very visible in the free or paid charts when rank-stripped, and almost all cases are clumped together. That to me indicates *possible* manual/deliberate targeting but only Amazon will know for sure. It could also be a malfunctioning fraud detection system. There is a little evidence for both theories (and they aren't necessarily mutually exclusive either). I don't see much evidence for these other claims, and plenty discounting them outright.
 
41 - 60 of 169 Posts
Status
Not open for further replies.
Top