The ROI of an Answering Service: Inputs, Formula, and a Full Worked Example

First number: your monthly incoming calls. It's in your carrier bill or app — take a normal month, not a vacation month. Second: your missed-call share. Without your own measurement, the known industry range is 25–30%; with one, use your real number — it beats every average in the world.

Third: your average job value. Take your last twenty invoices, average them — it's the most personal number in the calculation, the one that swings the result from single to triple. Fourth: your conversion rate on a live conversation. How many clients you speak with end up signing? Use your real rate, not a hope.

Potential monthly gain = calls × missed share × recovered share × conversion rate × job value. Each factor carries uncertainty, which is normal: the recovered share (what the SMS or AI actually re-engages) is the main unknown. Be prudent: half of missed calls re-engaged is a high hypothesis, a quarter a low one.

ROI then reads simply: (gain − solution cost) ÷ cost. An ROI of 5 means every dollar spent returns five. And break-even reads the other way: how many jobs per month must you recover to cover the subscription? That's often the most telling number for a solo operator.

Let's make the assumptions explicit. 160 calls per month, 28% missed — about 45 lost calls. Prudent recovery hypothesis: a quarter re-engaged, ≈11 conversations. The plumber's observed conversion: one in three signs, so ≈4 jobs. Average job value: $350. Monthly gain ≈ $1,400 against a $149 subscription. Even on low hypotheses, the ratio is unambiguous.

Now redo it with an optimistic hypothesis, to see the range: half of missed calls re-engaged (≈22), one in two signing (≈11 jobs). Monthly gain ≈ $3,900. The truth lives between the two — which is exactly why you measure from the first real month instead of believing either end of the range.

The subscription isn't the only cost. Count setup time — a quarter hour for simple forwarding, more to polish a script. Count the weekly conversation review: ten minutes, a small but real cost. And honestly count the minutes beyond your plan, if your volume explodes because... it works.

There's also the cost of doing nothing, invisible and permanent: the 25–30% of calls going to competitors every month. That cost appears on no invoice, which is why nobody ever computes it — yet it's the comparison term for any ROI. The status quo is never free; it's just unbilled.

Back to the example: at $350 per job, the $149 subscription is covered by one recovered job — a single one. The rest of the month is margin. That's the most useful reasoning for a solo operator: you don't need to believe the optimistic scenario; one more job per month than the status quo pays for the service.

The reading changes for small tickets: a barber at $45 a cut doesn't break even on single appointments — the service pays through the volume of booked slots (and the retention of clients who actually get a slot). Same formula, but the input value becomes "value of the appointment," not "value of the job."

Error 1: a fantasy conversion rate. The rate to use is your real one, not the one you wish you had. Error 2: counting as "recovered" people who would have called back anyway — some clients persist and return; net gain is lower than gross gain, especially for well-known neighborhood businesses.

Error 3: judging on one month. Seasonality, outages, and the first weeks of tuning blur any single month. Run the math over a full quarter, with a measured — not assumed — recovered share. At that point the calculation becomes an honest forecast, which is to say a decision tool, not a sales pitch.