What a missed call actually costs

The calculation is not sophisticated, which is part of why it goes unexamined — it looks too simple to be worth doing:

• Calls a week. Every inbound call, not just the new customers.
• The share you miss. After hours, on a job, already on the line.
• Your close rate when you do answer. You know this one.
• Average job value. The typical ticket, not your best month.

Multiply, annualise, and you have the revenue your phone is turning away. Run it on your own figures below — the numbers that matter here are yours, not an industry average.

Whatever you just entered for “percentage you miss” is probably too low, and the model is too generous in two other places as well.

• You are guessing, and the guess is flattering. Owners consistently estimate this from memory, and memory keeps the calls you answered. Your carrier log has the real figure. Pull last month before you trust the slider.
• A missed call is not a delayed call. It is a call that goes to the next name on the list. Treating it as deferred revenue rather than lost revenue is the single most common error here — you are not behind on those jobs, you never had them.
• The calls you miss are worth more than average. Nobody calls at 5:45pm on a Friday about a routine quote. Out-of-hours calls skew toward emergency work, which carries your highest ticket and your least price resistance. The model above prices them at your average.

Ordered by cost, because the expensive one is not always the right one and the cheap one solves more than people expect.

• Voicemail-to-text, with a rule. (Effectively free.) Your carrier already does this. The fix is not the transcript, it is the rule you attach to it: every missed call gets a reply inside the hour, from whoever sees it first. Most of the recoverable money in this article is available at this tier, and it is skipped because it feels too small to count as a solution.
• Routing to a second phone. (Low.) A ring group so calls hunt to a partner or spouse before dying. Breaks the moment both people are on jobs, which is precisely when volume peaks — but it costs almost nothing to try.
• A human answering service. ($1–3 a call, roughly.) Good coverage, genuine warmth, and a real ceiling: they do not know your pricing, your service area, or which jobs you refuse. They take messages competently. They do not book work, and customers can tell within about two sentences that they have reached a call centre.
• An AI receptionist that books into your calendar. (Highest.) Answers every time, knows your hours, your service area and your standard jobs, and writes the appointment straight into the calendar you already use. Worth it only above a certain call volume — below that, the cheaper tiers recover nearly as much for a fraction of the setup.

The honest ranking by return on effort is first, fourth, second, third. The answering service is the one people reach for first and the one that disappoints most often, because it solves the coverage problem without solving the booking problem.

Before you buy anything, get the real inputs. Both of these are free or close to it:

• Export your call log. Every mobile carrier and every VoIP provider exposes inbound calls with duration. Anything at zero seconds was missed. Two weeks is enough to see the shape — you are looking for the hours, not the total.
• Put a tracking number on the highest-intent surfaces. Your Google Business Profile and your site’s contact page. This separates “calls we missed” from “calls we never got,” which are different problems with different fixes, and are routinely treated as one.

The pattern almost always concentrates. Most businesses find the loss is not spread evenly across the week but stacked into a handful of predictable hours — which means the fix can be narrow and cheap rather than broad and expensive.

If you are booked three weeks out and turning work away, answering more calls does not add revenue. It adds a longer waiting list and more people to disappoint. The constraint is capacity, and no amount of phone automation moves a capacity constraint.

The one thing worth doing in that situation is the opposite of what this article describes: answer faster in order to say no faster, and to refer the work out while the referral is still worth something to whoever takes it. Automating the intake of jobs you cannot service is how a business ends up with a queue of angry customers and a very efficient system for producing them.