On-demand recording

On-demand recording means an agent decides when a call gets recorded by clicking a record button, rather than every call being captured automatically. The agent starts the recording when it matters — say, the moment a customer agrees to terms — and can stop it again. It is the opposite of always-on recording, where the Call recording runs for the whole call without anyone touching anything. With on-demand, the file only exists for the part of the conversation the agent chose to keep.

Why choose it? Sometimes you only need proof of one part of the call, like a verbal agreement or a confirmed order, and recording the rest is unnecessary storage and unnecessary risk. On-demand also helps when a customer asks not to be recorded for part of the conversation — the agent can simply stop and start again later. It is closely related to the PCI pause, which pauses recording during payment details; the difference is that on-demand controls the whole recording rather than just a single sensitive window inside it.

The trade-off is that on-demand depends on a person remembering to press record at the right moment. For quality monitoring across every agent, always-on is more reliable because nothing gets missed and no agent can quietly skip a call they would rather not have reviewed. Many centers use a mix: always-on for general calls and on-demand for campaigns where only specific moments need capturing. Tie the choice to your Disposition flow so the agent is reminded at the exact step that matters.

Whatever you do record still ends up in a file, so the same Recording format (WAV/MP3) and Recording retention decisions apply — how big each file is and how long you hold onto it. On-demand simply reduces how many files you create, which can lower both your storage bill and the amount of sensitive audio sitting on disk. If you go this route, train agents on exactly when to start and stop, and check a sample of calls early on to confirm the important moments are actually being captured.