Progressive Dialing: The Underrated Mode Between Power and Predictive

Where Progressive Fits on the Spectrum

Power dialing: one call per agent, fired when the agent is confirmed available. No calls in flight while the agent is in conversation.

Predictive dialing: multiple calls per agent, fired based on a statistical model of when agents will become available. Calls in flight during agent conversations.

Progressive dialing: one call per agent, fired slightly before the agent becomes available based on expected wrap-up time. No multiple simultaneous calls per agent, but the next call is dialing while the current call is winding down.

The practical effect: in power mode, there is a gap between when an agent ends a call and when the next call connects. That gap — 5 to 15 seconds depending on dial time and ring time before answer — is dead time. In progressive mode, the next call is already ringing before the agent wraps up, so the agent transitions directly from one conversation to the next.

The elimination of that inter-call gap is the entire value proposition of progressive mode. No abandoned calls (still one-call-per-agent), but 10–20% more conversations per agent per session compared to strict power mode.

The Wrap-Up Prediction Problem

Progressive mode fires the next call based on an estimate of when the agent will be available. If the estimate is wrong — if the agent wraps up faster than expected or the call runs long — one of two things happens:

Agent wraps up faster than predicted: The next call is still ringing. The agent is available but no call is connected yet. There is a brief gap. The benefit of progressive mode disappears temporarily.

Call runs longer than predicted: The next call connects and nobody is available to take it. The call either has to queue (if your platform supports micro-queuing) or is abandoned. This is the compliance risk in progressive mode.

The quality of the progressive dialing implementation depends almost entirely on how accurately the system predicts individual agent wrap-up time. A system that uses a static "average wrap-up time minus 15 seconds" trigger fires calls at the wrong moment for agents who deviate from average. A system that tracks per-agent wrap-up time and adjusts the trigger per-agent performs significantly better.

Minimum Agent Count for Progressive Mode

Unlike predictive mode, progressive mode does not require a large agent count to stabilize. It works for 2-agent operations through 200-agent operations because it is not running a statistical model across the pool — it is making a per-agent prediction.

This makes progressive mode the correct upgrade from power mode for any team that has optimized all the power-mode variables (list quality, wrap-up management, retry logic) and is still seeing 10–15% inter-call dead time. It is also the mode to reach for when you want more throughput than power mode but cannot reliably run predictive mode due to agent count or list quality constraints.

When Progressive Mode Beats Predictive

Progressive mode is preferable to predictive in three specific scenarios:

1. Teams of 5–14 agents. Below the predictive mode minimum viable team size, progressive delivers the throughput gain without the model instability. See the analysis on predictive dialing for small teams for the numbers in detail.

2. Lists with high handle time variance. When your list includes segments with wildly different conversation lengths — short appointment reminders mixed with complex sales conversations — predictive mode generates frequent abandoned calls because the model cannot handle the variance. Progressive mode does not over-fire; it only fires one call per agent, so handle time variance causes dead gaps, not abandoned calls.

3. Blended inbound/outbound operations. As covered in the blending inbound and outbound article, predictive mode in a blended environment generates abandoned calls when agents are pulled to inbound mid-session. Progressive mode does not have this problem — it only fires the next call when the specific agent is near wrap-up, and if the agent is pulled to inbound, the system simply delays the next outbound call for that seat.

Wrap-Up Timer Configuration

In progressive mode, the wrap-up timer (the window after a call ends, before the agent is considered available for the next outbound call) directly controls inter-call gap. If your wrap-up timer is set to 90 seconds and the progressive trigger fires at 75 seconds into wrap-up, the next call is ringing 15 seconds before the agent confirms ready.

If the progressive trigger fires at 60 seconds into a 90-second wrap-up, the next call has a 30-second ring time to work with before the agent confirms ready — plenty of time to reach a live contact on most lists. If the progressive trigger fires at 85 seconds, the next call has only 5 seconds — not enough ring time to reach most contacts, resulting in the next call not connecting until after the agent has already been available for several seconds.

The optimal trigger point is: (wrap-up timer duration) - (expected ring-to-answer time for your list) - (5-second buffer). On a 90-second wrap-up timer with a typical 15-second ring-to-answer time on your list, fire at 70 seconds into wrap-up.

Caller ID Rotation in Progressive Mode

Progressive mode fires calls at predictable, metered intervals — one per agent, sequenced. This makes it easier to rotate caller IDs in a defined pattern, which is valuable for campaigns where a single caller ID being called multiple times in a day will generate blocking or "spam likely" labeling.

UnlimCall provisions caller IDs on demand across 33 markets. For progressive-mode campaigns that rotate caller IDs across a pool, IDs are provisioned per request — not drawn from a shared inventory. Each caller ID is associated with your account.

Cost on Per-Minute vs Flat-Rate Networks

Progressive mode, like power mode, generates fewer connected minutes per session than predictive mode (because of the inter-call gaps, even if smaller than pure power mode). On per-minute platforms, this means slightly lower per-session termination costs than predictive — but also lower throughput.

On UnlimCall's flat-rate model, mode selection does not affect pricing. A 20-seat operation pays $1,980/month (US/CA rate) whether running power, progressive, or predictive mode. The choice is made entirely on throughput and compliance grounds, not cost grounds.