# Call answer rate: benchmarks and how to raise it

Source: https://dialmilo.com/hub/call-answer-rate
Updated: 2026-11-17

What a good call answer rate looks like, how to measure it without fooling yourself, and eight practical ways to get more people to pick up your outbound calls.

Call answer rate is the share of your calls a real person picks up. For unknown numbers calling consumers it is low by default, often below one in five per attempt, and published benchmarks are hard to compare because tools define "answered" differently. The biggest gains come from being expected and recognisable: tell people you'll call, call from a consistent local number with a clean reputation, call in their local evening, retry at a different time of day, and make sure callbacks reach someone who knows why you called.

A campaign with a 15% answer rate and one with a 30% answer rate can use the same script, the same voice and the same offer. The difference is almost always upstream of the conversation: who's calling, from what number, at what time, and whether the person had any reason to expect it. That's good news, because those are things you control.

## What "answer rate" means, and why benchmarks disagree

Before comparing your number with anyone else's, pin down what's being counted. Four metrics get mixed up constantly:

| Metric | Usually counts | Watch out for |
| --- | --- | --- |
| **Connect rate** | Any call the network completed: human, voicemail, answering machine | Looks flattering; voicemail inflates it |
| **Answer rate** | A human picked up | Some tools count a pickup followed by instant hang-up |
| **Conversation rate** | A human stayed long enough to talk, often 30+ seconds | Thresholds differ by vendor |
| **Reach rate (per person)** | Share of your list you spoke to after all attempts and callbacks | The one that tells you whether the campaign worked |

Two more questions change the number completely:

- **Per attempt or per person?** A 15% per-attempt answer rate across three attempts can mean 35% of people reached. Vendors quote whichever suits them.
- **Cold or warm?** Calling people who've never heard of you and calling existing customers about their account are different sports.

When you read a benchmark, ask all three questions. If the source doesn't say, treat the figure as a rough signal at best.

## Benchmarks worth knowing

The more careful sources all point the same way: people have learned to ignore numbers they don't recognise.

- **Pew Research Center, US, 2020.** In a survey of 10,211 US adults, just [19% said they generally answer](https://www.pewresearch.org/short-reads/2020/12/14/most-americans-dont-answer-cellphone-calls-from-unknown-numbers/) when an unknown number calls their mobile. 67% said they don't answer but check any voicemail left, and only 14% said they generally ignore voicemails.
- **Hiya, State of the Call 2026.** From a survey of more than 12,000 consumers in six countries, Hiya reports that [86% of unknown calls go unanswered](https://www.hiya.com/state-of-the-call).
- **CallHub, US and Canada, 2020.** Across more than 2.2 million campaign calls, the [average answer rate was 21.17%](https://callhub.io/blog/phone-banking/best-time-to-phone-bank/), ranging from 17.64% at 8–9am to 23.38% at 6–7pm.

So for cold calls to consumers, something in the high teens per attempt is a realistic starting point, and doubling it is unusual. Calls people expect, from a business they already deal with, are a different situation: the "unknown number" problem is exactly what you can remove.

Pew's voicemail figure explains a pattern you'll see in your own data: people who don't answer often listen to the voicemail or look up the number, then call back hours later. If your callback handling is poor, you're losing reach you already earned.

## Eight ways to raise your call answer rate

### 1. Tell people you're going to call

The single strongest lever, because it turns an unknown number into an expected one. An email or SMS the day before ("we'll give you a quick call tomorrow evening about your renewal"), a line in the signup confirmation, or a checkbox on the form ("call me about my quote").

If your journey tool sends that message, it can also trigger the call. With Dial Milo, a Customer.io journey, or any tool that can send a webhook, posts the person to the campaign with an `at` time after the message has gone out. The [CRM-triggered calls guide](/hub/crm-triggered-phone-calls) has the payloads.

### 2. Call from a local, consistent number

People are wary of international and unfamiliar numbers. Calling UK mobiles from a UK number and Dutch mobiles from a Dutch number looks more legitimate, and in Dial Milo the line is priced for the exact number dialled and your caller ID country. You can buy numbers in other countries from the dashboard.

Keep the number consistent across a campaign. A number people saw yesterday, or saved from your last call, is no longer unknown.

### 3. Protect your number's reputation

If networks or phone apps label your number "spam likely", your answer rate collapses regardless of how good the call is. The usual causes are high volumes of short calls, many complaints or a number recycled from a previous owner. Our guides to [STIR/SHAKEN and spam labels](/hub/stir-shaken-spam-likely) and [branded caller ID](/hub/branded-caller-id) cover how labelling works and what you can do about it.

Dial Milo watches for the symptom: it pauses a campaign automatically and emails you if none of the last 20 finished calls connected (voicemail counts as connected), or if at least 15 of 20 failed or were rejected. That stops you burning through a list while the number is being blocked. Once the number is sorted out, press **Resume** and the check starts again from zero.

### 4. Call in their local time, at the right time of day

Calling hours should apply where each person is, not where you are. For consumers on mobiles, early evening tends to beat early morning. Our [best time to call guide](/hub/best-time-to-call-customers) has the data and a method for testing your own hours.

Dial Milo applies your calling hours (default 10:00 to 20:00) in each person's own time zone, worked out from a `timezone` column or from the number itself.

### 5. Retry at a different part of the day

One attempt is rarely enough. But retrying at the same hour the next day often fails for the same reason the first did. Space attempts so they land in different parts of the day: a late-morning first attempt, an early-evening second. In Dial Milo that's attempts per person plus the gap between them, for example 2 tries, 4 hours apart. Voicemails and no-answers are retried automatically.

Don't overdo it. Three or four attempts in a day trains people to ignore you, and for some kinds of calls, such as debt collection, there are legal limits on frequency. Check the rules for your kind of call before you set the attempts.

### 6. Make callbacks count

People who missed your call often ring the number back, sometimes hours later. In an October 2026 pilot with an online casino operator (61 players, 121 dial attempts over two rounds), players called the campaign number back hours after missing the call. If the callback reaches a generic voicemail or a confused agent, that reach is wasted.

Dial Milo recognises a person calling back from a campaign's recent calls, including unanswered ones, says thanks for calling back, confirms who they are, and continues the campaign conversation. People who aren't on a list hear a neutral line. See [what happens when people call back](/help/when-players-call-back).

### 7. Clean the list before you dial

Wrong and badly formatted numbers drag the rate down and add cost. Before uploading:

- Convert numbers to international format (+44…, +1…). Dial Milo normalises common formats but refuses a national number when it can't tell the country.
- Remove numbers you know are dead, landlines you never reach, and people who've asked not to be called.
- Drop anyone called recently. Dial Milo's contact cooldown does this across campaigns: nobody is called twice in N days.

### 8. Don't lose them in the first second

This one affects what happens *after* a pickup, so it shows up as answered calls that last two seconds. If someone answers and hears silence or a delay, they assume it's a robocall and hang up. Keep the pause after pickup under a second, and open with something human and specific: who's calling, that it's an AI, and the person's name. The same pilot found that a short pickup pause and confirming identity before naming the brand both mattered.

Voicemail matters here too. Carrier answering-machine detection mislabels some voicemail greetings as humans, so Dial Milo also listens for voicemail greetings itself and hangs up rather than talking to a machine. Our [voicemail detection guide](/hub/voicemail-detection) explains why.

## Measuring it properly

Answer rate is easy to measure badly. A worked example shows why the per-person view matters.

**Worked example (illustrative figures).** You upload 1,000 people.

| Step | Calls | Answered | Rate |
| --- | --- | --- | --- |
| Attempt 1, late morning | 1,000 | 170 | 17% per attempt |
| Attempt 2, early evening, to the 830 not reached | 830 | 125 | 15% per attempt |
| Callbacks into the campaign number | — | 35 | — |
| **Total people reached** | | **330** | **33% of the list** |

The per-attempt rate never got above 17%, which looks poor next to vendor benchmarks. But a third of the list had a conversation, and that's the number that predicts results. If you'd only measured attempt 1, you'd have concluded the campaign was failing.

Now the costs. Calls that don't connect have no minute charge, and voicemails cost little because Milo hangs up straight away. Outbound runs roughly 14–20¢ a minute all in for US, UK and Dutch numbers. If the 330 conversations average two billed minutes, that's about $92 to $132 for the list. The pilot mentioned above came to $11.53 for 121 attempts, about 9.5¢ an attempt, because most attempts were unanswered or reached voicemail. You can price any number on the [rates page](/rates) and model a list with the spend calculator, which uses your own past answer rates.

### What to track every week

- **Per-attempt answer rate by hour and by caller ID.** Tells you if a slot or number is failing.
- **Per-person reach after all attempts and callbacks.** Tells you if the campaign is working.
- **Share of answered calls under 10 seconds.** A rising share means people are hanging up on the opening, or voicemails are being mistaken for people.
- **Opt-out rate.** If it rises as answer rate rises, you're calling too often or too late.

## When a low answer rate isn't the real problem

Sometimes answer rate is fine and something else is wrong. If people pick up and then decline, the issue is the offer or the script; test alternatives (Dial Milo lets you A/B test up to five scripts side by side). If people agree and then nothing happens, the issue is the follow-up. And if answer rate is low only for one segment, such as players who speak Dutch, the issue may be language rather than timing. Pass each person's language with the list so Milo opens in it.

Answer rate is the first gate, not the goal. Get it to a healthy level for your kind of list, then spend your effort on what happens once people pick up.

If you want to see how your list behaves, Dial Milo is pay as you go from a $10 top-up with no subscription, and the browser demo lets Milo call you before you pay anything. [Create an account](/signup) and start with a small slice of your list.

## Frequently asked

**What is a good call answer rate?**

It depends heavily on whether the person knows you. Calls from unknown numbers to consumers are mostly ignored: Hiya's 2026 report says 86% of unknown calls go unanswered, and a 2020 Pew survey found only 19% of US adults generally answer unknown numbers. A cold list landing around 15 to 20% per attempt is normal; calls to existing customers who expect them should do noticeably better. Your own trend matters more than any benchmark.

**What is the difference between answer rate and connect rate?**

Definitions vary by tool, which is why benchmarks are hard to compare. Answer rate usually means a human picked up. Connect rate often also counts voicemail and answering machines, so it looks higher. Always check whether a figure counts voicemail, and whether it's per dial attempt or per person after all retries.

**Should I measure answer rate per call or per person?**

Both, but decide with the per-person number. Per-attempt answer rate tells you whether a time slot or caller ID is working. Per-person reach, meaning the share of your list you actually spoke to after all retries and callbacks, tells you whether the campaign worked.

**Does a local phone number increase answer rate?**

People are more likely to answer a number that looks familiar and relevant, and a number from their own country avoids the international-call suspicion many people have. A consistent number also lets people recognise you and call back. Combine it with telling people beforehand that you'll call.

**Why did my answer rate suddenly drop?**

The most common cause is phone networks starting to label your number as spam. Other causes are a stale list, a change in calling hours, or calling the same people too often. Dial Milo pauses a campaign automatically and emails you if none of the last 20 calls connected, or if 15 of them failed.
