Dialer Ramp Strategy: How to Start Campaigns Without Burning Your List

Why the First 30 Minutes Are the Highest-Risk Window

At campaign start, three things are simultaneously unknown:

1. Real connect rate for this list. Your database may say 28% based on prior campaigns, but every list degrades. The actual connect rate on this session could be 18% if significant time has passed since the last dial, or 35% if the list is freshly sourced warm leads. You do not know until you measure.

1. Real average handle time for this agent set. AHT varies significantly across agent cohorts and by day of week. Monday morning AHTs are typically longer than Thursday afternoon AHTs because contacts are less rushed. You do not have this session's AHT baseline yet.

1. Model calibration state. For predictive mode, the algorithm's statistical model starts empty. For the first 20–30 minutes, it is extrapolating from historical data or defaults — not from actual session data.

Running at full pacing ratio during this window is running at full pacing ratio with three unknowns. The result is usually an abandon rate spike in minutes 5–15 as the model over-dials and the unknowns collapse against each other unfavorably.

The Three-Phase Ramp Structure

A systematic ramp moves through three phases:

Phase 1: Calibration (minutes 0–20). Run in power mode. Strict one-call-per-agent. This is your measurement window. After 20 minutes of power mode, you have a reliable connect rate for this specific list, this specific time of day, and this specific agent set. You know real AHT. You know which agents are performing above and below average on wrap-up time.

Phase 2: Conservative pacing (minutes 20–45). Switch to progressive or low-ratio predictive (1.3–1.5x). You now have data. Set the pacing ratio based on the connect rate you measured in Phase 1. Monitor rolling 5-minute abandon rate. If it stays below 1.5%, move to Phase 3.

Phase 3: Target pacing (minute 45 onward). Run at your campaign's target pacing ratio, informed by Phase 1 data and confirmed by Phase 2 stability. Adjust continuously based on rolling abandon rate as the list penetrates and connect rates shift.

Most campaigns can skip Phase 2 if Phase 1 data matches historical expectations closely (within 3 percentage points on connect rate, within 1 minute on AHT). Phase 2 is a bridge for campaigns where the data does not match — which is more common than most operators expect.

Ramp Strategy for New vs Existing Lists

New lists (no prior campaign history): Always run Phase 1 and Phase 2. The unknowns on a brand-new list are too large to skip calibration. Targeting a known good configuration that was tuned on a different list is a mistake.

Existing lists, same campaign recycle: If you are redialing a list you ran last week and your connect rate then was 24%, start conservative but can shorten Phase 1 to 10 minutes. Use historical AHT as a starting estimate but verify it is still accurate (agents may have shifted since the last session).

Warm leads (inbound web leads same-day): Connect rates on same-day inbound leads can run 40–60% — dramatically higher than cold lists. A pacing ratio calibrated for a 25% connect rate list will under-dial on a 50% connect rate list. Start with power mode for 10 minutes to measure the actual connect rate on this batch before applying any predictive ratio.

The Burned-List Problem

Over-dialing during ramp creates a specific kind of list damage. When the predictive model over-fires in the first 20 minutes, contacts answer and hear silence or a machine hang-up. On a mobile number, this registers as a missed call from your caller ID. If you call the same contact again in the same session (because your retry logic does not distinguish an abandoned call from a failed dial), you are now the caller who called twice and said nothing.

A contact who receives two abandoned calls from the same caller ID in the same morning is unlikely to answer a third call from that number. If your list scrub does not exclude numbers that received an abandoned call, you are progressively degrading the reachability of your own list with each campaign start that uses aggressive pacing.

Exclude numbers that received an abandoned call from retry within the same session. Flag them for callback in the next session with a human-initiated call or preview mode.

Caller ID Ramp Considerations

For campaigns that rotate caller IDs, ramp your caller ID usage with the same caution you apply to pacing ratios. Starting a session with all caller IDs active simultaneously generates a pattern of call bursts that carrier analytics may flag as unusual. A staggered caller ID activation — bring your first set of IDs online with your Phase 1 power mode test, add additional IDs as you move to Phase 2 and Phase 3 — spreads the traffic pattern more naturally.

UnlimCall provisions caller IDs on demand across 33 live markets, not from a shared inventory. If you need additional IDs for a specific campaign, they are provisioned for your account. There is no pool competition with other operators on the same number.

What Happens When You Skip the Ramp

The data from operations that track campaign-level abandon rates versus ramp structure consistently shows the same pattern: campaigns that start at full pacing ratio generate 60–80% of their total session abandoned calls in the first 20 minutes. The same campaigns that start in power mode and ramp generate more total contacts per session despite the slower start, because they are not burning the most reachable segment of the list on abandoned connections.

The throughput argument for skipping the ramp — "we lose 20 minutes of full predictive throughput" — is real but smaller than it appears. Twenty minutes of power-mode calls at 20 agents on a 25% connect rate generates approximately 40–50 connected contacts (20 agents × 2 calls/minute in power mode × 20 minutes × 25%). Starting full predictive and generating 4–8% abandon rate in that same window produces perhaps 60–70 connected contacts, but 4–6 of them were abandoned and a material fraction of the list has now seen an abandoned call from your number.