Progressive: one call per available agent
Progressive dialing places one outbound call as soon as an agent becomes free, and connects that agent to the call once it is answered. Because the system never dials more numbers than there are available agents, progressive carries close to zero abandon-rate risk — every answered call has an agent ready. The tradeoff is throughput: agent idle time between calls is higher than power or predictive dialing, since the system waits for a live answer before initiating the next call for that agent.
Power: a fixed dial-ahead ratio
Power dialing dials a fixed ratio of numbers ahead of agent availability — for example, 1.5 calls per available agent — rather than waiting for one-to-one availability. This increases throughput over progressive by overlapping some dial time with agent wrap-up, but introduces nonzero abandon risk: if the ratio is set too aggressively for the actual answer rate, more calls connect than there are agents to take them, and some answered calls get abandoned before an agent is available.
Predictive: statistical pacing for maximum throughput
Predictive dialing uses statistical modeling — historical answer rate, average handle time, and current agent availability — to dial ahead of agent availability by a calculated margin that adjusts continuously rather than using a fixed ratio. This delivers the highest agent utilization of the three modes, because the system is actively minimizing agent idle time based on real, current campaign performance. It is also the mode where a pacing miscalculation has the most direct compliance consequence: overly aggressive pacing produces abandoned calls, and abandon rate is the metric regulators and carriers scrutinize most closely for autodialed campaigns.
Matching mode to campaign priorities
Progressive suits campaigns where answer-rate discipline and near-zero abandonment matter more than raw throughput — sensitive outreach, small agent teams, or campaigns under tight regulatory scrutiny. Power suits a middle ground where moderate throughput gains are acceptable with disciplined ratio management. Predictive suits high-volume outbound operations, such as large-scale collections or sales campaigns, where maximizing agent utilization is the priority and the operation has the monitoring discipline to keep abandon rate within limits as conditions shift.
The abandon-rate discipline predictive dialing requires
Because predictive pacing actively trades abandon risk for throughput, it needs active guardrails, not just a one-time configuration: real-time abandon-rate monitoring, automatic de-pacing when the rate approaches a threshold, and a hard stop (kill switch) if a campaign breaches its configured limit. Voz360’s Compliance Engine applies abandon-rate fraud gating with a kill switch specifically because predictive is the mode most likely to need it, alongside the same DNC, TCPA consent, and allowed-hours checks that gate all six dialer modes before a call places.
Can the vendor tell you — in one sentence — which of their AI capabilities are rule-based, which are generative, and which are still roadmap?