Our ASR settings currently transcribe all section-published media, but we're starting to use Collections more widely, and the ASR service would be useful here too without transcribing all content.