Blog

What Missed Calls Cost a Small Business, and Four Ways to Stop It

Pull the call log, run the arithmetic, then compare the real fixes at real 2026 prices: a $19/month auto-text, a $150/month AI answering plan, $250/month of live humans, or removing the call entirely.

7 min read

Nobody knows how many customers called them last month and didn’t get an answer. Everybody has a guess. The guess is always lower than the log.

The log takes about four minutes to pull, and it is the only honest report you have on this. Below is how to read it, the arithmetic that turns it into a dollar figure, and the four ways to fix it with what each one actually costs in 2026. Written for the businesses we work with around Rochester, Winona, Red Wing, Austin, and Owatonna: trades, clinics, job shops, processors, large-animal vets, logistics.

Step one: count. Don’t estimate.

Pull two weeks of missed calls. On a cell phone it’s the Recents list filtered to missed. If you run a hosted phone system — RingCentral, Quo, Ooma, Teams Phone, a VoIP line from your local provider — the dashboard will give you missed calls by day and by hour, which is better.

Then split the count into two columns, because they have different fixes:

  • Missed during business hours. The phone rang while everyone was on a job, under a truck, or already on another call. This is a routing and staffing problem.
  • Missed outside business hours. Evenings, weekends, lunch, the 7:10 a.m. call before anyone is at a desk. This is a coverage problem, and it is usually the bigger and cheaper one to solve.

Mark the ones you recognize as vendors, robocalls, or the same customer calling twice. What’s left is the number that matters.

The arithmetic

Four inputs, all of which you either have or can estimate honestly. Here it is worked with plausible numbers for a small service business — substitute yours.

  1. Missed calls per month: 40. From the log, not from memory.
  2. Share that are new inquiries: 30%. That’s 12. The rest are vendors, repeat customers who will call back, and spam.
  3. Share you never recover: half. So 6. Assume the other six call back or you reach them later. This is generous to doing nothing, on purpose.
  4. Close rate on inquiries you actually talk to: 50%. Three jobs.

At an average invoice of $400, that is $1,200 a month, or $14,400 a year, walking to whoever answered on the second ring. If your average job is $4,000 rather than $400, stop reading and go pull the log.

Now the other direction, because this is the part vendors leave out: if you run the same arithmetic and get a small number, the correct answer is to do nothing. A machine shop with four missed calls a month, all from purchasing agents at three commercial accounts who will always call back, has no missed-call problem. Don’t buy anything below. Spend the money where your actual constraint is.

Do the free thing today

Before any purchase: your voicemail greeting costs nothing to fix and is probably costing you money. Most small-business greetings give the company name and “leave a message.” A caller who is on their third contractor of the morning has no reason to wait through that.

Say when calls get returned, and give a second way in: “You’ve reached us — we’re on jobs until four and we return every call the same day. If it’s urgent, text this same number with your address and what’s going on and we’ll get right back to you.” Fifteen minutes of work, zero dollars, and it makes the two paid options below work better.

Why calling them back doesn’t save you

The plan most owners rely on — see the missed call, ring them back between jobs — fails more often than it feels like it does, and there is good data on why.

In a Pew Research Center survey of 10,211 U.S. adults, only 19% said they generally answer their cellphone when an unknown number calls. Two-thirds (67%) said they don’t answer but will check a voicemail if one is left, and 14% ignore both. That survey was fielded in 2020 and the trend has not reversed since.

Read that from your customer’s side of the phone. When you call back from the shop line, you are the unknown number. The callback is roughly a one-in-five proposition, not a safety net. The voicemail you leave is worth more than the ring — two-thirds of people will at least listen to it.

One more thing in that data worth knowing, because it contradicts the objection we hear most: adults aged 18 to 29 were more likely to pick up an unknown number than older age groups. “Our customers are older, they’ll answer the phone” is backwards. Older customers are the ones least likely to take your callback.

Speed is the whole game, and the bar is low

The canonical study here is a Harvard Business Review audit that submitted test inquiries to 2,241 U.S. companies and timed the response. 37% responded within an hour, 16% within a day, 24% took longer than a day, and 23% never responded at all. Among companies that did respond within 30 days, the average took 42 hours. In a companion study of 1.25 million leads, firms that made contact within the hour were nearly seven times as likely to have a real conversation with a decision maker as those that waited one hour longer, and more than 60 times as likely as those that waited a day.

Two honest caveats: that work is from 2011, and it measured web form inquiries rather than phone calls. We cite it because the direction is not seriously in dispute and it matches what every contractor already knows — the first callback usually wins the job.

The useful part is the low bar. Nearly a quarter of the companies in that audit never answered at all. You are not competing against excellence. You are competing against a voicemail box at a competitor who is also on a roof.

Which sets the real target: some kind of reply inside five minutes, automatically, including at 9 p.m. on a Sunday. No human does that. That is the entire reason this particular problem is worth automating rather than trying harder at.

The cost of answering everything

There is a cost on the other side too, and it is the reason “just pick up the phone every time” is not the answer either. A well-known University of California study on interrupted work found something counterintuitive: people who were interrupted finished their tasks in less time, with no drop in quality — they compensated by working faster. What they reported instead was significantly more stress, frustration, time pressure, and effort.

That explains why the ringing phone never shows up as a problem in your numbers. It doesn’t make the job take longer. It shows up in your foreman, your office manager, and the mistake somebody makes at 4:30. An hour of front-office time in the Rochester metro costs about $37 fully loaded — we worked that number out here.

The four fixes, cheapest first

1. Missed-call auto-text — about $19 a month

The call goes unanswered and the caller gets a text within seconds: “Sorry we missed you. We’re on jobs until four. Reply here with your address and what’s going on and we’ll get you on the schedule.” Quo (formerly OpenPhone) includes auto-replies to missed calls on every plan, and the Starter plan is $19 per user per month month-to-month, or $15 if you pay annually. Budget a little extra for U.S. business-texting registration — a $19.50 one-time carrier review fee plus $1.50 to $3 a month.

For most shops around here this is the highest-return $19 in the business, for a reason that has nothing to do with technology: the customer does nothing new. They called a number and got a text back. Nobody has to download anything or learn a portal.

2. An AI answering service — $0 to $150 a month

Smith.ai’s AI receptionist is free for the first 25 calls a month, then $150 a month for 75 calls ($2.00 per call) on its Pro plan, and $500 a month for 300 calls on the tier where they build and run it for you. Month-to-month.

Where this earns its money: a high volume of similar factual calls. Your hours. Whether you service Plainview. Whether you’re taking new patients. What the current lead time is. Those are the calls that interrupt real work and contain no judgment.

Where we would tell you not to: if a wrong answer is expensive or embarrassing in a town of four thousand people, be careful. And an AI that books appointments into a calendar nobody watches is worse than voicemail, because now the customer thinks they’re on the schedule. Before you go live, call it twenty times yourself and try to break it. If it makes something up about your business, don’t launch it.

3. Live humans — $250 to $800 a month

Real answering services, priced two different ways. Ruby bills by the minute: $250 a month for 50 minutes, $395 for 100, $720 for 200, $1,725 for 500, with 24/7 coverage included. Smith.ai bills by the call on its human plans: $300 a month for 30 calls, $810 for 90, with overage at $11.50 a call at the low tier and $1.50 extra per call to actually book the appointment.

Here is the number to use against every vendor in this category, including us. Ruby’s entry plan works out to $5 a minute — roughly $300 per hour of talk time. That is about eight times what a loaded front-office hour costs you locally. Framed that way, the rule almost writes itself: buy live answering for the hours you cannot staff, not the hours you would rather not. $300 an hour is a bargain for the 6 p.m. Saturday call you would otherwise lose permanently. It is a poor deal for Tuesday-at-10 overflow that better call routing would have caught for free.

4. Remove the call entirely

The last option is the one people skip, and for a certain kind of business it is the only one that makes sense. If you are already booked out — a locker in November, a large-animal practice, an excavator in spring — you do not have a lead capture problem. The phone is not bringing you work you can’t already do. It is costing you the day.

For those operations the fix runs the other direction: online slot booking with a deposit so the calendar fills itself and no-shows drop; a structured intake form so a complicated spec arrives typed instead of shouted across a counter; and automatic outbound texts — “your order is ready,” “we’re on the way” — that kill the status calls before they dial. Outbound first. Again, the customer does nothing new.

Our own version of this is a website with booking at $149 to $249 a month with nothing down, or a fixed-scope build from $1,500 if you already have a site and just need the workflow behind it. And to be direct: if the auto-text at $19 covers your problem, buy that instead and don’t call us.

What we would tell you not to do

  • Don’t put a phone tree in front of a five-person company. The caller wanted a person. A menu adds friction without adding capacity, and everyone presses zero anyway.
  • Don’t pay for answering that can’t see your schedule. Otherwise you’re buying a message you still have to act on, which is what voicemail already was. If a service charges extra to book into your calendar, that add-on is the part you’re actually buying.
  • Don’t buy any of this before you’ve counted for two weeks. The log decides which of the four fixes you need, and often it says you need the free one.
  • Don’t automate the conversation where the quote is the conversation. Complicated custom work still needs you on the phone. Automate getting that call scheduled, not the judgment inside it.

The short version

Count missed calls for two weeks and split them into during-hours and after-hours. Multiply: new inquiries × the share you never recover × your close rate × your average invoice. Then buy the cheapest thing that clears the number — a $19 a month auto-text, an AI plan at $0 to $150, live humans at $250 to $800, or booking and status texts that stop the call from being necessary at all.

And remember which way the phone works now. Only about a fifth of people answer an unknown number, so your callback is not the safety net it feels like — a text is. Meanwhile nearly a quarter of the companies in that HBR audit never responded at all. The bar you are trying to clear is much lower than it feels from inside the truck.

If you pull the log and want a second opinion on what the number means, send it to hello@ramptec.dev. We’ll tell you which of the four we’d do, including the ones that don’t involve hiring us.

Want a straight answer about your operation?

Tell us what eats the most time in your week and we'll tell you plainly whether software would fix it, what it would cost, and when it wouldn't be worth doing. Email hello@ramptec.dev. We're based in Southeast Minnesota and happy to come to you.