Someone has decided the product is launching in a new market. Maybe it's a new language, maybe it's another English-speaking country, and maybe the date is already in a slide deck. Now it's on your desk, and the question is where to start.
The rest of this book is organised by topic, because that's how you look things up. This chapter is organised by order. It's the plan I'd follow to localise an existing app, step by step, with a pointer to the chapter that covers each step in depth. If your product doesn't exist yet, skip to "Starting from scratch", because your job is much easier.
Each phase below has three parts. Do is the work. Done when is how you know you can move on. Read is where the detail lives.
You don't have to finish every phase before starting the next, and on a big app you won't. But the order matters. Fixing the foundations before you extract strings, and testing with fake translations before you pay for real ones, saves you doing the expensive work twice.
I won't guess how long yours will take, because that depends on the codebase more than anything else. But I'd always start the same way: with the audit in phase 1. Search the code for hard-coded strings, hand-rolled date formats and money stored as floats, so you know the size of the job before anyone gives an estimate.