Just to add a little bit to this...
I avoid using the online metadata scrapers like the plague for this very reason. They aren't foolproof, and if you lose your Mezzmo server for some reason, it all has to scrape again, and any fixes/edits you did before the loss will have to be redone. Or if you build a new computer with better functionality. There are any number of reasons you may have to force Mezzmo to re-index all your content.
My current tool of choice is
Tiny Media Manager
I gather all the metadata using this tool, and if I can't find it I write my own (or home movies!). I then store it in a .nfo file, and scrape the fanart.jpg and folder.jpg files, and store them in the same folder as the movie (you're using subfolders for each movie/tv series, right?), then tell Mezzmo to ONLY get art from the media folders, and to ONLY use data from .nfo files (some will have text metadata embedded in them, even).
This way, no matter what, if I have to restore from a backup, or rebuild my Mezzmo database, the metadata is exactly as I have curated it, and never changes.
Once the file has been indexed properly, it's always a good idea to lock it, too. You can still hand edit what's in the database through the GUI, but it won't edit through metadata scans later (and it speeds up scanning a LOT).
Finally, this is one of the reasons I changed over to using metadata editors. When you have multiple series from different years of the same name, scrapers can't always smartly select the right one. With TMM, you can pick from several and apply the correct one, and it works like a charm.