Skip to main content
Every number on every dashboard comes from one path through the system. If you know the path, you know what each figure is counting — and, just as important, what it is not counting.
This page is the shared vocabulary for the whole product. The Dialer Performance dashboard, the reports, and your AI agent all use the definitions below. When a number looks wrong, start here.

​
🗺️ The path of one dial

Only the green box is a conversation. Everything else answered in some sense, which is exactly why “answered” and “connected” are different words here.
Answered, no rep is the one to watch. A real person picked up and reached nobody. Regulators call this an abandoned call, and the limit is 3% of live human answers. Our engine stops itself well before that — see Predictive pacing.

​
📖 What each outcome means

A person answered and reached one of your reps. This is the only outcome that can become a sale.Counted when: a human answered, the call bridged to a rep, and there was real audio. A machine verdict, a voicemail verdict, or a screening verdict can never be counted here, no matter how long the call lasted.
Somebody picked up the phone and never got a human. Usually every rep was already talking when the call landed.What to do: lower your lines per rep, or put more reps on the floor. This is the number the pacing engine exists to hold down.
An answering machine picked up. The system detects this by listening to the greeting, then stops the call from reaching a rep.Why it matters: a voicemail is a dial that cost you money and reached nobody. It is counted in your dials, never in your connects.
The lead’s phone answered on their behalf — iPhone Call Screening, Google Call Screen, or a carrier screening service. A recorded voice asks who is calling.Why it is its own category: a screened call is not a machine and not a person. The lead may well call you back. Merging it into “voicemail” hides that, and counting it as a connect inflates your rate with conversations that never happened. See iPhone screening.
It rang out. Nobody home.
They saw the call and actively rejected it. This is a signal about your list or your caller ID, not a missed opportunity.
Your side cancelled before the lead could answer. Almost always pacing: too many lines for the reps available.
The number does not exist. You are paying to dial it every time it comes round in the list. Worth cleaning.
The call failed on the network. Watch the trend, not the individual call.

​
🧮 The formulas, in plain words

One reconciliation rule. On the dashboard, the outcome breakdown always adds up to your dial count exactly. If the bars sum to a different number than the tile above them, something is wrong — tell your agent, or see Troubleshooting.

​
🌍 Time zones change your numbers

Every figure is bucketed by your day, not the server’s. “Today” means today where you are.
1

Set your time zone once

Open your profile in the top right, choose your time zone, and save. See Profile settings.
2

Check the dashboard header

The Dialer Performance header shows the time zone the numbers are using. If it does not match where your reps sit, the day boundaries will be wrong.
If your time zone is wrong, an evening call can land on the next day’s report. The numbers are still correct — they are just filed under a day you did not expect.

​
➡️ Where to go next

Dialer Performance dashboard

Read your numbers and know which ones to act on.

Predictive pacing

How the engine decides how many lines to dial.

iPhone screening

What screening is and how we detect it.

Ask your AI agent

Get these answers without opening the dashboard.