Nobody Launches a Catalogue Through a Browser

Adding a film through a web form is fine, and it stops being fine somewhere around the twentieth one. A launch catalogue is not twenty files, it is several hundred arriving from a distributor who has their own delivery process and no interest in learning yours. Browser uploads fail at ninety per cent on the largest file in the batch, they need somebody present, and they do not resume. This is the problem FTP was solving before streaming existed, which is exactly why it is still the right answer. It is unglamorous, it is universally understood by the people who deliver content for a living, and it does not care whether anybody is watching.

Getting two hundred video files onto a platform is a solved problem. Getting two hundred titles, synopses, release years, cast lists, genres and pieces of artwork onto a platform is the part that actually consumes a launch, and it is where projects quietly slip by a month. A file called something like a release group name tells you almost nothing, so the useful question is not how the video arrives but how much of the description arrives with it. Pulling metadata from a catalogue source rather than typing it turns the job from data entry into review, and reviewing two hundred prefilled records is a fundamentally different task from creating them. That side is covered under movies management, and what happens to the media itself under video uploading and transcoding.

The instinct is to treat bulk ingest as a migration tool used once and then forgotten, and operators who do that end up back in a browser every week. Ongoing delivery is the more common case in practice. A distributor sends you a batch every month, a production partner delivers episodes as they finish, a syndicated feed arrives on a schedule. All of those are the same shape as the launch problem at a smaller size, and all of them are better served by a folder than by somebody being emailed a file and uploading it themselves. Because ingested content goes through the same pipeline, everything downstream behaves identically, including the publish window and territory rules described under countdown and premiere and conditional access.

The first is naming. A batch of files named consistently can be matched to metadata almost automatically, and a batch named inconsistently becomes manual work no tool can remove, so agreeing a convention with a distributor before the first delivery is worth more than any feature. The second is that an ingest folder is a door into your platform, and it should be treated like one. Separate credentials per partner, and an expectation that content appears as drafts to be reviewed rather than going live the moment it lands. Nothing published without somebody looking at it is a good rule generally and an essential one when the files came from outside your organisation. The review step itself is where movies management picks up.