Ask a manager what percentage of inbound calls their team answers and the confident reply is “oh, almost all of them.” Pull the call records and the real number lands in the low 80s. The manager believes the claim, because a missed call is invisible to the person who didn’t answer it: the phone rings somewhere else, or nowhere, and the caller’s bad afternoon stays off the team’s radar.

Answer rate puts those calls on the record. The sections below cover a clean way to measure it and the ranges to expect.

Answer rate vs. abandonment rate

The two get conflated in most conversations about phone metrics, and they answer different questions.

Abandonment rate is queue-centric: of the callers who entered a queue, how many hung up before an agent answered? It’s the right metric for a formal call center with an ACD, and we’ve covered its optimal range in an earlier series.

Answer rate is organization-centric: of all external inbound calls that reached you during business hours, what fraction did a human pick up? It includes calls that rang an unattended desk, calls that dead-ended in an unchecked voicemail box, and calls to a hunt group whose members showed “available” while away from their desks. A company with no queue at all still has an answer rate, and it’s often worse than a call center’s.

Watch abandonment alone and the calls that failed before reaching a queue vanish from view. Answer rate catches them.

Compute it without fooling yourself

The naive version (answered calls divided by total calls) produces a number you can’t trust. Three adjustments fix it:

Business hours. Calls at 2am going to voicemail are working as designed. Including them deflates the rate and hands skeptics an excuse to dismiss the metric. Filter to your published hours, per site if hours differ.

External inbound. Internal extension-to-extension calls follow different patterns (the caller redials in seconds or walks over), so mixing them in adds noise. Use the trunk or direction field in your call records to keep outside calls and drop the rest.

Dedupe rapid retries. A caller who gets no answer and redials twice in two minutes shows up as three missed calls, but it’s one failed contact. Collapse calls from the same number within a short window (10 minutes is a reasonable rule) into a single attempt, counted as answered if any leg was answered. Expect this to move the rate up 2–4 points and make it honest.

The formula: answer rate = answered unique attempts / total unique external attempts, within business hours.

What’s normal

With the clean definition, the ranges from systems we’ve watched over the years cluster like this:

  • 90% and above: a healthy front office. Coverage exists for lunch and breaks, and voicemail is a deliberate choice rather than a default.
  • 80–89%: coverage gaps. Nothing is broken; the schedule leaves certain hours or groups uncovered, and the gap went unmeasured.
  • Below 80%: a structural problem. Bad routing, hunt groups pointing at departed employees, a published number ringing a desk that lost its occupant in a reorg, or far less capacity than peak demand.

One caveat: a 97% answer rate can still hide bad service. If it comes with long ring times or a receptionist who answers and transfers straight into the void, the calls count as “answered” while the callers leave unserved. Treat answer rate as a floor: clear it, then look at ring time and what happens after the transfer.

The usual culprits

Break a mediocre answer rate down by hour and day and the same four gaps show up:

The lunch hour. 12:00–13:00 is the worst hour of the day in most systems we see: the staggered-lunch plan exists on paper while the whole desk eats at noon.

Shift boundaries. The 15 minutes around shift change go uncovered: the morning person has logged off and the afternoon person is settling in.

Friday afternoon. Volume drops, so managers staff down. Volume drops less than staffing does, and the rate sags from about 3pm.

Voicemail black holes. A hunt group’s overflow points to the voicemail box of someone who left the company. The calls “complete” and the messages pile up unheard. Check an organization over 50 extensions and expect to find at least one of these; this analysis is how teams find them.

Fix the top three gaps on the cheap

None of this requires hiring. Pull answer rate by hour-of-day for the last four weeks, find the three worst cells, and:

  1. Stagger lunches with a written rota for whoever covers the main number: two people offset by an hour beat five people “keeping an ear out.”
  2. Overlap shifts by 15 minutes at the worst boundary, or move one person’s start time. Cost: near zero.
  3. Audit each voicemail destination reachable from your main routing tree. Delete or re-point any box belonging to a departed employee, and put a 30-second cap on ring-no-answer before overflow goes somewhere staffed.

These three changes are worth 5–8 points of answer rate within a month in the systems we’ve watched.

Track it week by week

An annual answer-rate review is archaeology. The metric earns its value when you check it each week: a one-page view of answer rate by day and hour, glanced at on Monday, catches a new gap the week it opens, whether the cause is a resignation, a routing change, or a seasonal volume shift. If producing the number means exporting raw call records and fighting a spreadsheet, the cadence dies: you measure once, fix something, and stop looking.

PBXDom exists for this exact problem. It collects call records around the clock from your existing Cisco, Avaya, Mitel, Panasonic, 3CX, or Asterisk/FreePBX system and keeps answered and missed-call dashboards current in real time, with hour-by-hour breakdowns by site and group, so the Monday glance takes one minute. There’s a free 14-day trial if you want to see your real number; details on the pricing page.