The Crafty Poet II: A Portable Workshop. That experiment is now over. Here are the results.
It was easy to set up the Amazon Giveaway and it went into effect immediately as Goodreads now does also (used to be a 7-day wait period). While a Goodreads Giveaway allows the user to select the length of time the Giveaway will run, there is a 7-day time limit on the length of the Amazon Giveaway, but that time will be cut off once a winner has been selected.
There are several options for how a winner is selected. My Giveaway was over within hours of its start time. I selected that there would be one book given and that each entrant had a 1 in 100 chance of winning. I would increase the 100 if I were to do another Amazon Giveaway as that would extend the time.
Amazon provided me with a Giveaway page code, but I never used it as the time was up so fast. They quickly sent me statistics. I had 424 Hits (people who looked at the Giveway), 175 Entrants (people who entered the Giveway), 14 Page Visits (people who went from the Giveway page to the book page).
So the exposure for my book with an Amazon Giveaway was far less than with past giveaways I ran at Goodreads, but I could increase the exposure if I changed the odds.
I was given the name of the winner as I was with Goodreads, but with Goodreads I had to mail out the book while with Amazon they mailed out the book. Before Goodreads turned bad, the only cost I incurred was the cost of one book, envelope, and postage. Amazon charged me a "setup cost" of $27.09 and later refunded $.06. The price for my book at Amazon is now $18.64 discounted from $21.99. So I was charged $8.39 for postage and handling. It would cost me less if I mailed a copy from my own stash and paid the postage.
Conclusion: I doubt I'll do another Amazon Giveway as I don't see any particular benefit to it. It's far less costly than a Goodreads Giveaway, but had no apparent effect on sales.
But just to continue this experiment one step further, I'm going to try a Giveaway at Facebook.