Skip to content
KPIShield
Book a demo
Book a demoStart free trialSign in

What a finding looks like

Seven patterns, and what each one does to your numbers.

Every finding opens with what was logged next to what the record shows, the other explanations ruled out first, and the metrics it distorts. Drawn here with synthetic data from the generator the checks are tested on.

Book a demoStart free trial

From a real audit · Sep 18 – Oct 2, 2026

4.6% of 41,656 logged calls had no real chance of reaching anyone.

Read the case study

01of 07

Ringing hang-ups

A dial ended two to five seconds into ringing counts as a dial, but nobody could have picked up. A few happen to everyone. Dozens a day add activity that never had a chance.

Logged as
120 dials today
The record
Many ended within seconds of ringing

The check: Very short dial attempts

One rep's morning · seconds each dial rang30 dialsSynthetic data
  • No real chance of reaching anyone
  • Rang long enough to be answered
  • Answered (minutes connected)
8 of 30 dials ended within five seconds of ringing, each logged as a call; nobody could have answered that fast. The rest rang out, reached voicemail or were answered.

What clears it

  • Busy signals and dead numbers: the network ended those, not the rep
  • Outage windows and spam-flagged caller IDs
  • Voicemail pickups and call screeners
  • Compared only with reps on the same team and dialer mode
  • An accidental dial redialed at once into a real call

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials (distorted)
  • Talk time
  • Conversations
  • Connect rate (distorted)
  • Pipeline forecast
  • Coaching
  • Lead burn (distorted)

Dials go up while connect rate goes down, so the scorecard says the list is bad. Leads get marked as attempted without ever ringing long enough to be answered.

See what this looks like on your team.

Book a demo

02of 07

Burst and rapid‑redial runs

Runs of very short dials packed into a few minutes, or the same number dialed three or more times within ten minutes with no answer. Each attempt is logged as its own call.

Logged as
36 calls before 10:30
The record
A 30-minute run, nearly every dial cut while ringing

The check: Burst dialing · Rapid redial of the same number

A 30-minute run · seconds each dial rang36 dialsSynthetic data

One number, four attempts, nobody answered

36 dials in 30 minutes, almost all ended within five seconds of ringing, each logged as its own call. Below: the same morning, four attempts to one number, one to two and a half minutes apart.

What clears it

  • Power, parallel and predictive dialer pacing
  • A run of voicemail hang-ups in a row
  • A requested callback or a dropped call
  • Retrying a busy line
  • One owner behind several accounts on a shared main line

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials (distorted)
  • Talk time
  • Conversations
  • Connect rate
  • Pipeline forecast
  • Coaching (distorted)
  • Lead burn (distorted)

One prospect becomes several counted touches and a cadence step reads as done. Managers coach to a dial count the runs produced.

See what this looks like on your team.

Book a demo

03of 07

Conversations the record contradicts

A call dispositioned as a connected conversation while the carrier record shows no answer, a busy line or a hang-up during ringing, or the recording holds a voicemail greeting or a phone menu.

Logged as
Disposition: Connected
The record
No answer on the network

The check: Human-conversation disposition mismatch

Logged as · what the record shows4 callsSynthetic data
Calls logged as conversations, with what the call record shows
CallLogged asWhat the carrier record showsRecording
Tue 9:14 AMConnectedNo answer · rang 31 s—
Tue 9:52 AMConnectedBusy signal—
Tue 11:07 AMConversationAnswered · 34 sVoicemail greeting, no message left
Tue 2:31 PMConnectedRep hung up after 3 s of ringing—
Each of these counts as a conversation in the activity reports. The carrier record (and, for the third, the recording) says nobody spoke with the rep.

What clears it

  • Your own approved meaning for each disposition: "Connected" alone never counts
  • An exact match between the disposition and the call
  • A person reached on an adjacent call within 10 minutes
  • A voicemail message actually left
  • A person who joined later in the call

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials
  • Talk time
  • Conversations (distorted)
  • Connect rate (distorted)
  • Pipeline forecast (distorted)
  • Coaching (distorted)
  • Lead burn

Each one is a conversation in the funnel that never happened. Conversion looks worse, the forecast looks better, and coaching time goes to calls nobody had.

See what this looks like on your team.

Book a demo

04of 07

Malformed or dead numbers, dialed again

A number with a missing digit, or one the network already reported as disconnected, fails in zero seconds and can still be logged as a call. The first attempt is a list problem. Dialing it again is not.

Logged as
4 calls to one prospect
The record
Failed in 0 s every time: 9 digits

The check: Dials to numbers already reported disconnected

A number that cannot ring4 attemptsSynthetic data
Number as stored(312) 555-0199 digits

Each attempt failed in 0 s and was logged “No Answer”: 4 dials on the rep's day.

The first failure is a list problem: the number is missing a digit. Every redial after it is a dial that could not reach anyone, still counted on the day.

What clears it

  • The first attempt to a number: only dials after it failed count
  • A number that later reached a person
  • Outage windows, which never mark a number dead

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials (distorted)
  • Talk time
  • Conversations
  • Connect rate
  • Pipeline forecast
  • Coaching
  • Lead burn (distorted)

Dials that could not ring sit in the activity totals, and the bad record stays in the list for the next rep to dial.

See what this looks like on your team.

Book a demo

05of 07

Calls logged by hand with no dialer record

A call logged by hand in the CRM, with talk time and a disposition, and no dialer call to that number anywhere near that time.

Logged as
Conversation · 14 min
The record
No dialer call to that number within 10 minutes

The check: Calls logged by hand with no dialer record

Dialer calls · calls logged by hand8:00 – 12:00Synthetic data
The shaded window is ten minutes either side of the logged call. The dialer placed no call to that number inside it, or anywhere that morning.

What clears it

  • A dialer call to the same number within the match window (a double log)
  • Dialer-outage days, when many reps log by hand
  • Approved mobile or desk-phone calling, compared with peers on the same policy

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials
  • Talk time (distorted)
  • Conversations (distorted)
  • Connect rate
  • Pipeline forecast (distorted)
  • Coaching
  • Lead burn

Talk time and conversations nobody can open or hear. They count toward targets and the forecast like any other call.

See what this looks like on your team.

Book a demo

06of 07

Long holds and phone menus counted as talk time

A call that stayed connected for minutes, all of it a phone menu or a hold queue, with no live person on the other side, logged as talk time or a conversation.

Logged as
Conversation · 7 min 59 s
The record
Hold queue the whole call; no live person

The check: Extended automated connection

One call · 7 min 59 s connectedLogged: ConversationSynthetic data
The transcript covers the whole call: hold announcements every 20 to 45 seconds, nothing from the rep, no live person. Gaps between announcements stay unknown time; they are never counted as silence.

What clears it

  • A person who answered after the automated part
  • A destination several reps reach (a switchboard): it goes to an operations queue instead
  • A scheduled callback or later human contact
  • Expected phone trees, suppressed with an owner, a reason and an expiry

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials
  • Talk time (distorted)
  • Conversations (distorted)
  • Connect rate
  • Pipeline forecast
  • Coaching (distorted)
  • Lead burn

Talk-time targets are met on hold music. The rep's day looks full, and a manager reviewing talk time sees effort that never reached a buyer.

See what this looks like on your team.

Book a demo

07of 07

Scripted, metronome pacing

A person's gaps between dials vary with every call. When they land on the same value again and again, far more often than peers' do, something other than a person is setting the pace.

Logged as
69 dials an hour
The record
A new dial every 52 seconds, within half a second

The check: Automated or scripted dialing

Gaps between dials · ten minutesSame team, same dialer modeSynthetic data
A person's gaps vary with every call: a voicemail, a note, a sip of coffee. These land on 52 seconds, within half a second, dial after dial.

What clears it

  • Power, parallel and predictive dialer pacing
  • A dialer auto-advance setting: peers on it would show the same rhythm
  • A steady wrap-up after real conversations
  • A steady rhythm of dials that rang long enough to be answered

Each is checked before anything reaches a reviewer. What the data can't settle is listed as not checked: a question to ask, not a point against the rep.

What it does to your numbers

  • Dials (distorted)
  • Talk time
  • Conversations
  • Connect rate
  • Pipeline forecast
  • Coaching
  • Lead burn (distorted)

Dials per hour climb with no one behind them. A quiet afternoon reads as the busiest of the week.

See what this looks like on your team.

Book a demo

Case study · Customer name withheld

Eleven days, 51 reps, every dial checked against the record.

A B2B software company's outbound sales team. 4.6% of logged calls had no real chance of reaching anyone; 61 logged conversations were contradicted by the record. Measured cost: $63,000 a year, for 10 reps.

Read the case study

Book a demo

See these patterns on your calls.

Tell us about your team. We'll walk you through the product on demo data, and what your dialer's data lets us check.

  1. The review queue and one finding's evidence, end to end
  2. What your source supports: the capability report
  3. Trial, Integrity Audit or a plan: what fits your team

Rather try it yourself? Start a 14-day free trial.

You get a confirmation from KPI Shield right away, and our team replies to it. We use your details to reply, and keep the page and campaign you came from to see which pages help. Privacy