Agents change the math: they expand capacity, so quotas should be set on pipeline and meetings the human team can actually work, not on activity the agent generates. Base targets on conversion reality and ramp them as trust in the agent grows.
Quota the human bottleneck
Agents can generate far more top-of-funnel than a team can convert. Set quotas around the constraint — AE meeting capacity and pipeline quality — rather than around agent output, or you'll flood the funnel and drop the balls that matter.
Base targets on real conversion
Work backwards from your actual reply-to-meeting and meeting-to-pipeline rates (see Conversion rate analysis) to set achievable numbers. Quotas built on optimistic assumptions demoralize; quotas built on your real funnel math hold up.
Ramp as trust grows
Early on, keep more in human review and set conservative targets. As the agent proves itself and you release more to auto, capacity rises and quotas can climb. Increase deliberately alongside agent autonomy, not ahead of it.
Measure the right unit
Shift the metric from activity (emails sent) to outcomes (qualified meetings, pipeline created) — the agent handles activity now. Reps are measured on what only they can do: working conversations and closing.
