OK, it's time to get something out in the open.
I've noticed that all the pod-caster novelists and POD-ers have this obsession with Amazon rankings, which I can understand. That's why they all encourage people to buy the book on opening day, so it skyrockets the rankings to the top 10, if even for 1/10 of a second, but still, from then on in the promo materials you can say, "the top 10 Amazon hit..."
But does this really work? Do publishers care about gimmicky marketing tactics like this? Can you show me a case where it backfired?
I have to admit that this approach looks more respectable than begging for some 'agent' who may or may not even respond to you to take 15% of whatever you may or may not make. Seems like the new breed of action adventure authors (and it does seems to be mostly in those genres), are eschewing agents altogether, or at least until the asgents solicit them.
First I will point out that mainstream authors can be as obsessed with their Amazon rankings as POD authors. I know I am, checking it whenever I can while my book is in preorder because I know a spike equals a sale and I get all excited. POD people have even more reason to be obsessed because they are, for the most part, not in stores and rely on Amazon to sell their book.
People used to be more dismissive of Amazon. As of 2007, someone told me all internet sales were only 7% of the book market, not that big. Then in early 2008 I heard 17% from someone else, then 23%. The point is, the internet sales are climbing in proportion to stores going down, and while they won't replace stores, they can no longer be dismissed. So if your book is selling well on Amazon, good for you. Be happy about it. It's no small thing anymore.
As for the rankings system, I've been messing with it a bit myself as of late so I can talk about it a little. There have been waves of people trying to analyze the ranking system and figure out Amazon's algorithm, which is a closely-guarded secret. The reason for the waves is that Amazon occasionally changes the algorithm based on people trying to manipulate it or for it to more efficiently reflect sales. In 2005, someone wrote that she got her book into the Top 25 list by buying a book an hour (it is recalculated every hour for high-ranked books, and mostly once a day for above 100,000 numbered books). I tried that, and it didn't work. She wrote that she only bought one copy because she read somewhere else that for an individual buyer, multiple copies are reduced to one for the purposes of ranking to prevent the manipulation of rank by authors. Come to think of it, I don't know why Amazon would care - they just want to sell books, and they're selling books to you, and they're ranking the books by what's sold, so who buys them shouldn't be a huge issue, be it one person buying 100 or 100 people buying one. On the other hand, 100 people buying one indicates popularity over numbers, so again, it's complicated.
I experimented a little (and got flagged by Amazon, who called me and asked me if I wanted a corporate account, as I seemed to be buying so many books that if I had any intention of actually buying them and not canceling my pre-orders, it would be financially cost-efficient to have a corporate account and not just an Amazon credit card like I do) and here's what I found:
(1) Numbers do matter. Buying 100 copies over buying 1 multiple times over a long period of time will artificially raise your rank faster and higher.
(2) Amazon does calculate the ranks about once an hour, but it puts the calculations in effect on about the 40 minute part of the hour. I don't know when it actually does the calculations, but all I know is that the spike would always be around :35 or :40. Otherwise, my book would just slowly depreciate.
(3) Massive canceling of massive orders will result in the rank going back down (meaning, the ranking system takes cancellations into account and doesn't just track orders. It tracks sales)
(4) It is ridiculously, stupidly hard to get your book above 2000. There are 4 million books listed on Amazon, so that shouldn't be a surprise, but it seems to cap at 2000 for some reason. Despite massive orders in a short span of a few hours, I could not get my rank to reflect that.
(5) Manipulating your rank is probably a little unethical, though it's a victim-less crime.
To answer your questions specifically, I have seen ads for marketing, but and I gave them some thought but decided against them. If you have a ton of friends, you can get them to buy the book all at the same time, but it won't do as much to your rank as the ads promise. Also some people are big on "email blasts" where you email people you barely know or random people you got from a list you paid for to buy your book. Having never bought a book from an email blast (and I get a few of them every couple months at my Rejector email), I cannot say this is an effective measure except if you've written a specific book for a specific community, in which case I would just call that marketing and be done with it.
I confess I have paid a service $3/month to track my sales based on my rank, because I am curious, and also because I think it has access to Bookscan and can actually look up sales once the book goes on sale and tell me how many sold. The rank itself does not totally reflect sales - it reflects your rank relative to the other 4 million books on Amazon, so if other books aren't doing as well, your rank will not depreciate as slowly.
As for the agent thing, that's another discussion entirely, but I think that anyone with more than one book should immediately get an agent even if they already have a book contract. That 15% is well-earned, and I say that not just because I work for an agent but also because I have my own agent for my books, and she is severely underpaid for all the work she does for me in my opinion. If you have one book and have an offer, get an agent. She'll take her 15% percent for looking over the contract, but she's not there for that. She's there to sell your second, third, and fourth books, and so on.
Lastly, some people have begun to mention in query letters that they were "an Amazon bestseller." We don't pay any attention to this. It could mean they manipulated their rank to be a bestseller, and it could mean they were a bestseller in a specific category. I imagine it's not all that hard to be a bestseller in the Books ‹Outdoors & Nature ‹Hunting & Fishing ‹Fishing ‹Ice Fishing category. Or maybe I'm just underestimating the number of books about ice fishing.