My understanding is that in a traditionally published situation, with the first print run of a book, the publisher sets the number of copies to print, and if it sells out, then a new print run of a potentially higher number may follow, because of the demand. If you’re an established author with a good sales history, your next book will likely require a bigger first print-run of your next book, as compared to a first-time author’s first print-run will. This is obviously because that author has a sales history that can be used to gauge how well successive books will sell. Whereas in the first-time author’s case, there is no sales history.
A POD platform is a “print-on-demand” facility, and that’s all it is. It may offer other services parallel to what a traditional publisher would do, but you pay for those services above the costs to print your book. A publisher invests in you, whereas with POD you are putting up all the capital and therefore absorbing all the risk.
Now, when I order 30 copies of Centipede Dragon, the POD will print 30 copies. When someone off the Barnes and Noble or Amazon website orders a single copy, one is printed. The up-side is that copies from a “print-run” will not need to be stored and scheduled for distribution. There’s no inventory, in other words.
As you see from this graphic, taken from the Independent Publishing Magazine’s website, CreateSpace was ranked as the number one POD platform. That, coupled with its being the least expensive-costing option to print, is what my decision was primarily based. Et voila! My choice was made.