Marketing operations gets asked two questions about AI that most stacks cannot answer: what did we actually run, and what did it cost. When generation happens in personal chat windows, the honest answer is a shrug.
HeadshotMarketing treats an AI action like any other governed action. Each one records the model, a summary of the prompt, the inputs referenced, the output, the cost, the actor and the approver. The AI usage ledger rolls this up by month, by team and by tool, and quotas and spend caps per workspace stop a runaway job rather than recording it after the fact. Workspaces can bring their own model keys.
Drafts saved from a generation carry an AI-made tag with the tool and model, so months later it is still possible to ask which assets started as AI output and which run produced them.
This is an illustrative scenario, not a customer story.
Do it yourself
Watch AI usage accumulate as real, inspectable records: month-scoped generation and token counts, a per-run duration and cost, a full execution feed, and an ai-generated tag on everything that came out of it.
Open Content → AI Tools and read the three cards at the top: available tools, generations this month, and tokens this month.
You should see: Workspace AI usage for the current calendar month is visible immediately, derived from real execution records rather than a lifetime counter.
Ready to make this your story?
