The Missed-Call Ledger

Every argument in this track eventually lands on the same instruction: measure the leak before you buy the fix. This is the training for the first and easiest measurement — the one sitting in your phone system right now.

The front-office anchor post made the strategic case: a missed call is a lead you already paid for walking to a competitor, and AI answers hand the buyer the competitor's number in the same breath. What it didn't do is walk through the arithmetic. Twenty minutes, four steps, one number.

Step 1 — Pull the log

Every phone setup — carrier account, VoIP dashboard, even a personal cell's call history — can produce last month's inbound calls with timestamps and outcomes. Export it or screenshot it; don't summarize from memory. Memory is systematically kind to your answer rate, which is exactly why this exercise exists.

You are counting one thing: inbound calls during your stated business hours that no human answered. Rang out, hit voicemail, or sat unreturned for more than an hour — all count. Log after-hours calls separately; they're a different opportunity (what 24/7 answering would capture) rather than a staffing failure.

Step 2 — Qualify honestly

Not every missed call is a lost customer — some are vendors, robocalls, wrong numbers. Listen to a sample of the voicemails that were left and classify them. Then apply the uncomfortable correction: buyers with urgent problems mostly don't leave voicemails; they dial the next name. If your voicemail sample runs half prospects, your missed-call population almost certainly runs higher.

Be conservative anyway. A ledger you trust at 60 percent confidence beats an inflated one you'll quietly dismiss later. Write down the prospect fraction you can defend.

Step 3 — Price it

Three numbers you already know, one you just computed:

Missed prospect calls per month × booking rate × average ticket = the monthly leak.

Use your real booking rate for answered calls (if you book roughly half the genuine inquiries you speak to, use that). Use your real average ticket — service call or project, whichever mix your inbound calls actually represent. No industry benchmarks, no vendor multipliers: every input comes from your own records, which means nobody in your company can argue with the output. That provenance is the point — a number without a source is as weak as an answer without one.

Step 4 — Make it a ledger, not an audit

A one-time measurement is trivia. The ledger becomes infrastructure when it's weekly: missed calls, after-hours calls, and the running dollar figure, on the same one-page Monday brief that estimate-versus-actual belongs on. Ten minutes to update once the log export is routine — and it's precisely the kind of mechanical, recurring pull that belongs in the connective layer's job description rather than a human's.

Two things happen once the number exists. First, the fix conversation changes shape: missed-call text-back, extended answering, or a built AI receptionist stop being technology decisions and become arithmetic — the monthly leak against the monthly cost, decided in one sitting. Second, the ledger becomes your before-and-after: when you do install a fix, the same weekly number proves whether it worked, which is more than most technology purchases in this industry can ever claim.

The leak is measurable this afternoon. Price it before anyone prices a fix for you.