ROI · 6 min read
The ROI of an AI voice receptionist for hospitals
Missed calls are invisible on a P&L — there's no line item for the patient who rang, got no answer, and booked elsewhere. Put a number on it and the case for an AI receptionist is straightforward. Here's the math, kept deliberately conservative.
What a ten-hospital network loses
Take a network where each hospital handles 10,000 calls a month — 1,00,000 calls across ten branches. At a conservative 25% peak-hour miss rate, 25,000 calls go unanswered. If just 20% of those callers intended to book, that's 5,000 lost bookings a month.
| Line | Value |
|---|---|
| Calls per hospital, per month | 10,000 |
| Network monthly volume | 1,00,000 |
| Missed at 25% miss rate | 25,000 |
| Would-be bookings at 20% intent | 5,000 |
| Average OPD consult value | ₹750 |
| Monthly leakage | ₹37.5L |
| Annualized | ₹4.5Cr |
What you spend vs. what you capture
Now the other side of the ledger. At ₹5.5/min and an average 2-minute call across 1,00,000 calls, the network spends about ₹11.0L a month. Against that, Aarvox eliminates the bulk of missed calls and recovers bookings that would otherwise have walked.
| Cost | Capture | |
|---|---|---|
| Total minutes | 2,00,000 | — |
| Aarvox rate | ₹5.5 / min | — |
| Monthly investment | ₹11.0L | — |
| Missed calls eliminated | — | 21,250 |
| Recovered bookings @ 20% intent | — | 4,250 |
| Average OPD consult value | — | ₹750 |
| Monthly revenue captured | — | ₹31.9L |
Net monthly capture after cost: ₹20.9L — roughly a 2.9× return, before counting diagnostics, pharmacy, or downstream procedures.
The non-revenue upside
- —~2,400 front-desk hours freed each month — staff redirected to in-person patients.
- —70–90% reduction in missed calls within the first month of going live.
- —15–25% increase in confirmed OPD bookings from callers who would have hung up.
Even halve every assumption and the model still pays for itself. That's the point of a conservative case: the downside is capped, the upside is not.