Outbound campaigns & dialing

The list works the agents, not the other way round.

Campaigns pull from lead lists through a hopper that is filtered, ordered and topped up continuously, so an agent never waits on a list refresh. Recycling rules decide when a no-answer comes back and how many attempts a number gets before it retires — the difference between working a list and burning it.

Campaign — renewals

live
Dialed1,284
Connects318
Sales41
Modepredictive · ratio 2.4pacing
Hopperleads ready to dial2,140
Machine detectedrouted to voicemail drop27%
SuppressedDNC · before dial96
4 modesManual to predictive
Pre-dialDNC suppression
Per campaignRetry rules
Time-zoneAware dialing
What you get

Four ways to dial the same list.

A campaign that works in month three does not work in week one. The mode is a setting, not an architecture, so a list can start careful and end fast without anybody rebuilding it.

👈

Manual and preview

The agent sees the record before the call goes out and decides when to place it. Slower per hour, and the right choice for high-value accounts, complex scripts and a team still learning the product.

Progressive

One call placed per free agent, automatically, the moment they finish wrapping. No idle gap between calls and no abandoned calls, because a person is always waiting on the other end.

📡

Predictive

Multiple calls placed per free agent, with the ratio adjusted continuously against live connect rate and talk time. Occupancy climbs, and the pacing engine holds the abandonment rate under the ceiling you set.

🎵

Answering-machine detection

Detection runs per campaign, with the handling you choose: drop the call, leave a recorded message and move on, or pass it to an agent anyway. Set it against your own tolerance for false positives.

🔄

Recycling and retries

Rules per disposition decide when a no-answer or busy comes back around, at what time of day, and how many attempts a number gets before it retires from the list for good.

🚫

Suppression before dial

DNC and your own suppression lists are checked before the number is dialed, not after. Attempt caps and calling windows apply the same way, per campaign and per destination.

The hopper

An agent should never wait on a list refresh.

Leads are pulled into a hopper that is filtered, ordered by your priority rules and topped up continuously while the campaign runs. Filters are yours — custom fields, region, value band, last outcome — so the same raw list can drive three campaigns with three different orders.

  • Custom fields imported with the list and available to filters
  • Priority ordering by value, recency or your own rule
  • Continuous top-up so the dial queue never empties
  • Multiple campaigns drawing from one underlying list

Hopper — renewals

live
Ready to dialafter filters2,140
Filterregion · south & midlandsactive
Orderby contract value, descendingpriority
Retiredattempt cap reached312
At the agent

The record, the script and the outcome on one screen.

When the call connects, the agent already has the customer record, the script step they are on and the disposition list. Callbacks are scheduled from the same screen and come back to the same agent if that is what you want.

  • Screen-pop with the customer record on connect
  • Scripting with branches driven by the answer given
  • Dispositions that drive the retry rule automatically
  • Scheduled callbacks, agent-specific or open to the team

Agent desktop

ConnectedM. Hart · renewals01:12
Scriptstep 3 of 6objection
Dispositionawaiting selection8 options
Callbackscheduled · Thu 10:00same agent
Compliance controls

The limits are yours to set, and the dialer holds them.

Abandonment ceiling, attempt caps, calling windows by time zone, DNC and internal suppression. Every one of them is a setting you configure against your own regulator, and the dialer enforces it rather than reporting the breach afterwards.

  • Maximum abandonment rate enforced by the pacing engine
  • Per-number and per-campaign attempt caps
  • Calling windows evaluated in the destination's local time
  • DNC and suppression checked pre-dial, with an audit trail
  • Recording pause-and-resume so card details never reach the audio

Controls — renewals

Abandonment capenforced by pacing3.0%
Measured todayrolling2.1%
Calling windowlocal time at destination09:00–20:00
Attempt capper number6
Specification

The details engineers ask about.

No asterisks. If something depends on your carrier, your hardware or your regulator, it says so.

Dialing modes
Manual, preview, progressive and predictive, switchable per campaign without rebuilding the list.
Pacing
Ratio adjusted continuously against connect rate, talk time and available agents, capped by a maximum abandonment rate you configure.
Lead lists
Import with custom fields; filters, priority ordering and multiple campaigns per underlying list.
Answering-machine detection
Per-campaign detection with configurable handling — drop, leave a recorded message, or route to an agent.
Retries
Per-disposition recycling rules with time-of-day control and per-number attempt caps.
Suppression
DNC and internal suppression lists applied before dial, with an audit trail of what was blocked and why.
Calling windows
Evaluated in the destination's local time zone, configurable per campaign and destination.
Agent tools
Screen-pop, branching scripts, dispositions and scheduled callbacks from the same screen.
Carrier routing
Per-campaign carrier selection and least-cost routing across parallel SIP trunks.
Compliance
These controls are configured by you to match your regulator. They are enforcement tools, not a compliance guarantee.
Questions

Answers to what usually comes next.

What is the difference between power and predictive dialing?
Power — what we call progressive — places exactly one call per free agent, so nobody is ever connected to silence. Predictive places more than one, betting on the measured connect rate to have an agent free when someone answers. Predictive raises occupancy and introduces the abandonment risk that the pacing cap exists to control.
Can we start on preview and move to predictive later?
That is the recommended path, and it is a setting rather than a migration. Run preview while agents learn the script and the connect rate is unknown, then switch the same campaign to predictive once there is enough data for the pacing engine to work from.
How do you stop a good list being burned?
Per-disposition recycling rules and a hard attempt cap per number. A no-answer comes back at a different time of day; a wrong number never comes back. When a number hits its cap it retires from the list permanently rather than quietly reappearing next month.
Can we drop a recorded message on answering machines?
Yes, per campaign. Detection decides, and the handling is your choice: drop silently, play a recorded message and release the line, or pass the call to an agent anyway if you would rather they judge it.
Does it handle multiple time zones?
Calling windows are evaluated in the destination's local time, so a single national campaign will not call the west coast at six in the morning. The window is set per campaign and per destination.
Let's talk

Bring a real list for a week.

We will import it with your custom fields, connect your trunks, and run one team through preview into predictive. At the end you will have a measured connect rate and a measured abandonment rate rather than a projection.