What missed calls cost and how to win customers back

Missed calls do not show up in your financial statements. Nobody records the order that never happened because every agent was busy at 9:15 AM, or the office had already closed at 7:05 PM. If you sell and provide service by phone, this may be one of your biggest hidden losses. Here is how to estimate it, where to look for the cause, and how to win some customers back.

Which calls count as missed

There are three kinds of missed calls. Count them separately – each has its own cause and its own fix.

• Unanswered: the call came in during business hours, but no one picked up.
• Abandoned in queue: the customer waited and hung up before reaching an agent.
• Calls after hours, on weekends, and on holidays.

Why a missed call costs more than one sale

• The lost sale: a customer ready to buy does not wait – they call the next company in the search results.
• Lost repeat purchases: once they buy from a competitor, they keep buying there.
• Repeat calls: those who do call back add load to hours that are already busy.
• Wasted ad spend: you paid for the call, and no one answered.
• Frustration: people who could not get through message you, leave reviews, and complain on social media.

How to estimate a year of losses

You do not need precise data – a simple estimate shows the order of magnitude.

1. Missed calls per month = calls per day × share missed × days with calls. If customers call on weekends too, count those days.
2. Lost deals = missed calls × share of calls about a purchase × share of those calls that end in a sale.
3. Monthly loss = lost deals × average order value.
4. Yearly loss = monthly loss × 12 × (1 + number of repeat purchases per customer per year).

An example. A company gets 300 calls a day, and 15% go unanswered – over 22 business days that is 990 missed calls. Half of the calls are about a purchase, and one in ten of those ends in a sale. So the company loses about 50 deals a month on first purchases alone. If a customer buys once more during the year, the yearly loss comes to about 1,190 orders, and half of them are repeat purchases. Multiply by your average order value to get the dollar figure.

Plug your own numbers into the missed calls cost calculator – it estimates the yearly loss and shows how much of it callbacks can win back.

Not every missed call is a lost customer: some people call back on their own. But more often they call another company. The point of the estimate is not a precise number but an answer to whether the problem is worth working on.

Find the hours when calls get lost

Missed calls are not spread evenly. They most often get lost at these times:

• Right after opening in the morning: the queue grows faster than agents get on the line.
• At lunch, when breaks overlap.
• In the evening and on weekends, when no one is working.
• After ads and email campaigns that cause a sudden spike in calls.
• During outages in other systems, when everyone calls at once.

Build an hourly report: how many calls came in, how many were answered, how many were abandoned in queue, and how many fell outside business hours. Look at it by day of the week too – Mondays and Fridays often behave differently.

Why missed calls snowball at peak hours. A queue does not behave linearly: one agent short, and wait times multiply. Here is an Erlang C example with values from the calculator: 120 calls an hour, 4 minutes each. 11 agents answer 81% of calls within 20 seconds, and the average caller waits 20 seconds. One goes on break – now only 65% are answered within 20 seconds, and the average wait is 49 seconds. Two go on break – 40% and 2.5 minutes, and many callers hang up before anyone answers. That is why two overlapping breaks at 11:00 AM cost more than they seem. Adding one agent for two peak hours every business day takes 2 × 22 = 44 paid agent hours a month – compare that with what six lost sales are worth to you.

Automatic callbacks: reach the people who could not get through

Call back as fast as you can. Ten minutes later, the customer is still choosing where to buy; a day later, they already have. With automatic callbacks, the system puts every missed and abandoned call in a queue and dials the customer as soon as an agent is free. No one has to write down numbers and dial them by hand.

• You set the attempt rules: for example, 2-3 attempts 10-15 minutes apart, after which the task closes with the result noted.
• The system does not call people who already got through in the meantime.
• Overnight calls go into the queue too – the system dials those customers in the morning when agents get on the line.
• The result of every callback goes to your CRM, such as Bitrix24 or Kommo (amoCRM), and into reports: whether you reached the customer and how the conversation ended.

In the calculator, the last field covers callbacks – what share of the loss they win back. The default is 20%: out of 10 lost customers you win back two, while the rest have already bought elsewhere by the time you call. In the example above, that is about 240 orders a year, and the loss shrinks to roughly 950 orders.

Customer-requested callbacks: do not keep people on hold

A requested callback is a different thing: the customer chooses a callback instead of waiting. When the queue gets long, the IVR offers a callback instead. The customer hangs up, keeps their place in line, and gets a call when their turn comes. Most people prefer that to listening to hold music, and fewer calls are abandoned. Customers can also request a callback at a convenient time from your website or chat.

An AI voice agent at night and during peaks

Callbacks win back the people you already missed. An AI voice agent helps you miss fewer: it picks up when agents are busy or the office is closed.

• At night and on weekends, it answers frequent questions and takes orders and inquiries.
• During peaks, it handles simple requests and passes complex ones to an agent along with the details it collected.
• On a sales line, it finds out what the customer needs and creates a task for the right person to call back.
• It creates a CRM task with a summary of the conversation – the morning shift starts with a ready list of follow-ups.

An AI agent can replace the evening shift where evening callers ask frequent, simple questions. Complex cases still go to people – in the morning, from tasks already waiting in the CRM. The missed calls cost calculator counts only the money won back through callbacks. To see how much an AI agent saves on agent work, use the AI agent savings calculator.

Fix the root cause

• Use the Erlang C calculator to work out how many agents you need at peak hours, and build shifts around that.
• Move breaks and training out of the busiest hours.
• Agree with marketing that the contact center hears about ads and email campaigns in advance.
• Move frequent simple questions to self-service – the IVR or messaging. We cover how to make a phone menu easy to use in our article on IVR design.
• Reduce average handle time (AHT) – the same team will take more calls. Our article on reducing AHT explains how.

Make one person responsible for missed calls

Missed calls often fall between departments. Marketing drives the calls, the contact center answers them, sales owns revenue – and nobody owns the calls no one answered. Assign one owner and set up a weekly routine.

• One person reviews the missed calls report every week and reports how the numbers are changing.
• Callbacks have a clear target – for example, the first attempt during business hours within 15 minutes.
• Marketing sends the contact center its calendar of ads and campaigns in advance.
• Sales sees which won-back calls turned into orders – the work shows up in revenue.
• Once a quarter, the loss estimate is recalculated in the calculator with fresh numbers.

Once the problem has an owner, it stops being "normal losses" and becomes a metric you can improve month after month.

Your plan for the next month

1. Build an hourly report of missed calls for last month.
2. Estimate your yearly loss in the calculator.
3. Mark the hours where you are short on agents, and use the Erlang C calculator to see what it takes to cover them: one agent for two peak hours a day is about 44 paid hours a month. Compare that with your losses.
4. Turn on automatic callbacks for missed and abandoned calls, and offer customers a callback instead of waiting in the queue.
5. If people call with simple questions at night and during peaks, put an AI voice agent on those hours.
6. After four weeks, compare how many calls were missed, how many customers callbacks reached, and how many of them bought.

CONTACT.WORK logs every call, including overnight ones, so the missed calls report, callbacks, and the AI agent all work from the same data – and you can see how many customers you won back.