From a lead list to calls that end in an outcome

How a campaign is built

A campaign is three things: a lead list, an agent, and dialling rules. The list comes from a file or through the API; the agent gets a goal, a script and the fields to capture; the dialling rules decide when calling is allowed, how many attempts are permitted and what happens on each outcome. Once running, the system creates the call, dials over your own telephony with your own caller ID, and the agent speaks first — it does not wait for the other side to open, which is what prevents the awkward silence that gets calls hung up in the first two seconds.

• Lead lists from a file or through the public API
• Your own caller ID and dialling region, per department or per campaign
• The fields the agent must capture, defined up front and stored on the lead
• A full call script — objection handling included — as an editable flow

No answer, busy, voicemail

Most outbound calls are not answered the first time, which makes the retry policy the core of the thing rather than a nice-to-have. The system classifies the dial result — no answer, busy, declined — and returns the lead to the queue as long as it has not exceeded the attempt limit you set, instead of dropping it or dialling it in a loop. Voicemail detection is available, but stated honestly: it is a conservative heuristic that ships switched off, and at real volume you are better served by your carrier's own answering-machine detection. A call that was never answered does not count as minutes — ring time is not billed.

A circuit breaker: the campaign stops itself

When something breaks on the telephony side — a trunk goes down, a number gets blocked, a list turns out to be full of bad numbers — the classic damage is that the dialler keeps running and burns the entire list while no call actually connects. The circuit breaker detects a spike in the dial-failure rate and pauses the campaign automatically, recording the reason, so you can investigate and resume rather than discover it the next morning. It is opt-in, and our view is that it belongs on every campaign running against a list you did not clean yourself.

Handing a live outbound call to a rep

The most valuable moment in a campaign is an interested customer who is on the line right now. Instead of promising that someone will call back, the agent can bridge: the system dials the rep — a specific extension included — joins them to the same live call, and then steps out. The rep gets a live conversation with context, not a row in a list. If no rep is available, the agent falls back to the script — books a meeting, captures a preferred time window, or flags the lead for callback — so the call does not end empty.

What gets measured at the end

Every outbound call ends in a record structurally identical to an inbound one: recording, transcript, summary, sentiment, escalation reason and outcome — and the outcome rolls back onto the lead and the campaign counters. On your side, a signed webhook fires at the end of each call, and a summarised activity is written to Salesforce or HubSpot against the right lead or contact. Practically, this means you can answer "why isn't this campaign working" from the summaries themselves — which objection kept recurring, at which point in the script calls fall apart — and not only from connect rates.

When not to run an outbound campaign

Outbound dialling amplifies what you already have: a good list yields more, a cold and irrelevant one yields more complaints. If your callers have no reasonable reason to expect a call from you, this tool will work against you. Three limits worth stating up front: responsibility for consent, suppression lists and permitted calling hours sits with the business — the system logs every call and lets you mark a lead do-not-contact, but it is not legal advice; sensitive news — medical, legal, employment — should not be delivered by an agent; and at volume, the reputation of your outbound number is an asset to protect rather than spend on one list.