We've all heard the occasional indie publisher post about buying certain of their titles, to see if they show up in sales reports for them. In most of these threads, the author comes back and reports that “Yes” their sales showed up. I have seen those however, who said their sales never did show up (usually at the KDP forums). You then begin to wonder if they're being honest about their experiment or not. I actually had one occasion in which a sale didn't show up for me on the former Pubit! platform and it was a very narrow niche title, that no one else bought during that same month either. It would not surprise me if I haven't also been credited for sales that I wasn't suppose to receive or that a return wasn't applied to my reports that should have been. So...it goes both ways. The point is though, that we have this way to test the sales reporting systems at booksellers and we should be doing so, from time to time.
Something me and another indie publisher here at Nook did recently, was to buy one of each other's books. We actually did this at Amazon, not here, but both sales did show up. My suggestion is that more publishers begin doing this, especially when sales trends at a certain bookseller site, seem to be way off. Yes, I realize this is being done already but some of the authors doing this will announce the name of the book they're having a fellow author, friend or family member buy and that's like telling someone you’re playing hide-n-seek with, that “you're going to hide behind the tree by the birdbath”. I also suggest you do not do these type tests by buying your own book because booksellers can instantly see matches for emails and names, that are connected to an author's book.
This is just a suggestion but if you want a little more peace of mind, I highly suggesting doing this experiment with the different bookseller sites, from time to time. A small cost for better confidence.