How would I go about writing a history book? | AskHistorians Archive Observer

Here's a step-by-step how-to guide from someone who's done this entire process once, is in the midst of doing it again, and, like an academic Sisyphus, is condemned to push this same boulder up the mountain forever.

Find an original topic that's interesting enough to devote an entire book to and which hasn't already been done better by someone else (good luck with that).
Identify the primary and secondary sources you'll need to do the research for your book and what archives/libraries they're held in (assuming you have all the requisite language skills); then figure out how you're going to find the time and money to travel to those archives and do weeks of research.
Conduct your research using the appropriate methods of source criticism and interpretation; develop your hypotheses based on the information you've gleaned from those documents, then continue working and find out that those hypotheses were wrong, and go back to the drawing board on that phase of your work.
Once you've conducted all of your research, figure out how to order your notes and how you're going to structure all of the information you've gathered into a format that's useful to the reader but doesn't bore them with the tedious minutiae of your research; realize that half of the notes you gathered from your archival research aren't going to make it into the manuscript (that's probably an optimistic estimate, the rate for most books is probably worse).
Spend several months (or more likely a couple of years) writing your manuscript in whatever free time you have from your real job. Factor in enough time to account for the times you decide that what you've written sucks and you scrap it and start over. If you're really dedicated to the subject and have tons of time to work on it, you might finish the manuscript someday.
Write your book proposal and send it to academic publishers (or trade presses if you're still in denial about how interesting/original your work is). Collect rejection letters and pray for an editor to be interested enough to choose your proposal over the hundreds of others they receive in a given year. If you don't have a Ph.D. and aren't employed by some type of academic institution, this is going to be a tough sell. If you don't have a Ph.D., for the love of God do not get one.
Submit your manuscript for peer review. One of your reviewers will probably be pretty nice about it and give you some positive feedback, while the other will tell you that you are the worst mistake that God hath wrought upon mankind and that the best thing you can do with your manuscript is use it to light your own funeral pyre. Hopefully the first review is nice enough that the editor will allow you to resubmit the manuscript with some changes rather than simply deciding that it's not good enough and telling you to go back to step 6.
Finish your edited manuscript and get to the production process of the book. This means at least one round of copyedits that will require you to pore over every letter and punctuation mark of the manuscript searching for errors. Miss an obvious one that also gets past the copyeditor; you'll see it immediately on the first page that you open the finished book to. Spend the money to hire an indexer or take the time to index the book yourself (for the love of God do not do this). Track down the copyright owners of all your photographs and other media featured in the book and email them twice a week for eight weeks until you finally get a response that no, you can't use that picture, go away.
Wait an inordinate amount of time for the finished book to come out and be delivered. Wait even longer because the supply chain melts down and the ship that was bringing your books over from the Chinese printer the publisher outsourced the production of your book to tipped over, spilling the cargo container into the Pacific Ocean (this literally happened to a project I was working on, I swear to God every word of this is true). Finally get the finished product in your hand, open it up, and find the error referenced in step 8.
Wait a few months for the scathing reviews of your books to start pouring in from the academic journals in your field; even if your book is good, most academics are too insecure to admit that and would rather eviscerate you for your failings than acknowledge you've made an original and interesting contribution to the field. Wait until the next tax period to get your statement from the press informing you that you sold 100 copies, mostly to libraries which automatically buy books from a list, and that those purchases have netted you a cool $150 in royalties. Go back to step 1 and repeat until you go blind from staring at microfilm readers or die.

Obviously this is somewhat cynical and tongue-in-cheek, but most of the bad things in this answer actually did happen to me. For all my kvetching, I actually fared alright. My book was published by a press with a good reputation in my field within about four years of me finishing my Ph.D. Thankfully, unlike most Ph.D.s of my generation, I was already gainfully employed in my field by that time and wasn't desperately waving the book in the face of any potential employer who would look at it.

Basically, the point I'm trying to get across is that writing and publishing a history book is a very difficult, time-consuming, and frustrating process that requires a number of different skills (and a lot of stubbornness and good luck). If you're not a credentialed historian of some description, it's going to be hard for you to do this because that's just how academic publishing works. It's an arcane ivory-tower world, and even those of us on the inside struggle to break through. For those on the outside, it's very unlikely to work out well. Of course, if you decide to go through with it, we'll be here to answer your questions. Good luck.