Missed Call Cost India: Your 5 Numbers, ₹2/min

The number you are looking for — "Indian businesses lose ₹X to missed calls" — would need four things to be worth quoting: a defined population of businesses, a sample size, a stated method for deciding a call was lost revenue rather than a wrong number, and a date. We looked for a figure carrying all four and did not find one.

What we found instead falls into three groups. We are describing the groups rather than repeating the numbers in them.

Percentages about caller behaviour with nothing Indian behind them. Figures about how many callers never try a second business, or never leave a voicemail, are quoted widely on Indian pages. We could not trace any of them to a measurement of Indian callers — no sample, no year, no method. Several read as vendor marketing copy that has been restated often enough to look like a finding.

Round national loss figures with no anchor. Crore-scale numbers appear with no population, no sample and no year. A number that cannot be traced to a measurement is a marketing device, not evidence.

Vendor outcome claims with no source line. Outcome percentages and customer counts, published with nothing behind them. We have deliberately not reproduced any of those here, not even as an illustration, because we have not verified the pages they sit on — and an unverified number does not become safe by being quoted disapprovingly.

One structural fact about India is worth stating on its own, because it changes what a missed call even means here. Incoming calls in India are free to the receiver — the calling party pays, and has since the early 2000s — so a missed call is itself a message. It has been used to mean "call me back" for as long as most of us have had mobile phones. Voicemail is the opposite: on most Indian networks it is an opt-in service, and whether it is available to you at all, what it costs and how you retrieve it depend on your operator and your plan. Check yours rather than assuming.

We are not putting a percentage on any of that, because we could not find an Indian measurement of it we would stand behind. What it means commercially does not need a statistic: an unanswered call with no message leaves you with a phone number and nothing else. You call back cold, at a time that suits you rather than the caller, with no idea what they wanted. An answered call leaves you with a name, a requirement and a time. That difference is real whether or not anyone has ever measured it nationally.

Your own call log has all four of the things the industry figure is missing — a defined population of one business, a complete sample, a method you choose, and a date. It takes about ninety seconds to open. That is why the rest of this page is about your numbers rather than anyone's.

Five inputs, all of which you can get from your own phone and your own memory of the last month. None of them come from us.

The formula:

Monthly loss = M × G × C × V × (1 − R)

Read the last term carefully, because it is the one every hype page leaves out. If half your missed callers are recovered on a callback, half your headline loss is not a loss at all. Leaving `(1 − R)` out is the single largest reason the numbers on the internet are too big.

This is a model, not a measurement. It will not be exactly right. It will be roughly right in a way no national average can be, because every one of the five inputs is yours. If you want the shortest possible version: the only figure that matters is what one recovered job is worth, multiplied by how many of them you are actually losing.

Three failure modes to avoid while you do this.

Do not count every missed call as a lost customer. That is the arithmetic that produces crore-scale headlines. In most Indian businesses a large part of the missed-call log is not new demand at all.

Do not use your best job as your average job. Use the median of what you actually invoiced last month.

Do not count a call twice. If the same number rang three times in an hour, that is one enquiry, not three.

Five worked examples. Every input below is illustrative and made up for the arithmetic — none of it is data about these industries, and you should replace all of it with your own figures. They exist to show you the shape of the sum and the size of the gap between losing a call and answering it.

The last column is our own arithmetic on our own published rate: missed calls × 2 billed minutes × ₹2/min. Two billed minutes per call is a modelling assumption, not a measurement — billing is whole-minute, so a 65-second call bills as 2 minutes. It is talk-time only; the full monthly outflow, including the number recharge and the Care Plan, is in the next section and is much larger than this column.

The bottom row is the most useful one on this page, and it is the row that argues against buying from us. At 30 missed calls a month, an estimated ₹3,600 of loss does not cover a ₹7,600 first-quarter monthly cost. That business should not buy an AI agent from anyone this year.

Notice also how differently the four viable rows behave. The real-estate row has the largest loss and the fewest closed jobs behind it — one brokerage. The restaurant row has the largest number of missed calls and a low recovery rate, because nobody calls a restaurant back. The AC service row is the most typical small-business shape. No single industry figure could describe all four, which is the whole argument of this page.

What none of these rows contain is the cost of the callback you are doing today: the time you spend ringing unknown numbers back in the evening, most of which do not answer. That is a real cost and we have not attempted to price it, because we would be inventing a value for your hour.

Take the AC service row — 150 missed calls a month, 300 billed minutes, ₹600 of talk-time. Here is the full monthly outflow, not just the per-minute rate.

And the monthly total at 300 minutes:

The third column assumes you route an existing number of your own to the agent instead of taking one from us. Whether you can do that, and how, depends on what your own operator supports on your specific connection — treat it as a question to ask them, not as a saving you already have.

Read the bottom row rather than the headline rate. At 300 minutes a month our advertised ₹2/min is functionally ₹25.33/min for the first three months. Nothing about that is a trick — it is what a fixed monthly cost does to a small number of minutes, and it happens with any vendor that has one, ours included — but you should see it before you buy rather than after.

Credit runway. The first ₹10,000 load covers month one (₹2,000 number + ₹5,000 Care Plan + ₹600 usage = ₹7,600), leaving about ₹2,400. At this volume expect to top up during month two. Top-ups start at ₹500, credits never expire, and at zero balance everything pauses gracefully — there are no surprise bills.

Break-even, in jobs rather than rupees. At an illustrative ₹1,200 average job, ₹7,600 is 6.3 recovered jobs — so seven clears it in months one to three — and ₹2,600 is 2.2 jobs, so three clears it from month four. Against an estimated ₹34,650 of monthly loss in that row it clears easily, but that estimate was built on inputs we invented, and yours will be different. Run yours before you decide.

The founding-client trade, stated plainly and not as fine print. The ₹40,000 build is free in return for two things: the Care Plan for the first three months at the founding rate, and an honest testimonial plus a short case study after go-live. That is a real commitment on your side, and it is exactly why months one to three cost ₹7,600 rather than ₹2,600. Every figure above is also on /pricing with nothing held back for a sales call. All Dvaarik prices are quoted exclusive of GST; 18% applies when you load credits.

Every competitor figure below was read on that vendor's own page on 21 Jul 2026, with the qualifier they published beside it. Nothing here comes from a search snippet, an AI summary or a rival's comparison page — each of those was tried during collection and each produced at least one false number.

The right-hand column is our arithmetic on their published rate, not a price any of them quotes: what one two-minute call costs at the rate shown. It is there because a per-minute rate is the wrong unit for this question, and a per-call figure is the right one.

Three things to take from that table.

First, the lowest advertised numbers in this market are attached to volumes a small business will never do. ₹0.40/min sits on a 10,000 minutes/month minimum with a three-month commitment. Agni's ₹2/min is an Enterprise rate at 10,000+ minutes. HuskyVoice's ₹4/min is an enterprise floor. The AC service business above does 300 minutes.

Second, of the sixteen vendors checked on 21 Jul 2026, exactly one — SquadStack — publishes no price of any kind, and at least six require a monthly subscription before you can place a single call.

Third, the claim we make is about terms rather than about being lowest, and we will not make the other one: ₹2/min from the very first minute — no subscription, no minimum volume, no lock-in on the rate. The one commitment that does exist is the three-month Care Plan traded against the free build, and it is in the table above rather than buried. A fuller market breakdown lives at /blog/ai-calling-agent-price-india-2026.

Four situations where the honest answer is that this is not worth buying, from us or from anyone. I would rather lose the enquiry now than have you work it out in month two.

1. Your estimated loss is under about ₹8,000 a month. Look at the small-clinic row: ₹3,600 of estimated loss against ₹7,600 of monthly cost in the first quarter. That is not a marginal call, it is a clear no. Below roughly 200 minutes a month our own fixed costs dominate — about ₹7,000 of number and Care Plan against ₹400 of talk-time — and a better outgoing message plus one fixed hour each evening for callbacks solves the same problem for nothing. If you still want an agent at that volume, look at the vendors in the table above that publish no monthly commitment and no setup fee, Trikon among them, and remember to add the telephony their own page says your carrier bills separately.

2. Your recovery rate is already high. If you are a real-estate agent or an insurance advisor who chases every enquiry until they answer, your `(1 − R)` term is small and most of what looks like a loss is a delay rather than a loss. The value of answering for you is speed and consistency, not recovery — which is a real benefit, but a smaller one than the headline arithmetic suggests. /blog/speed-to-lead-ai-calling-india is the honest version of that case.

3. Most of your missed calls are not enquiries. If your `G` is 15% because your log is mostly suppliers, couriers and existing customers, you are not losing revenue — you are losing convenience. Worth fixing, not worth ₹7,600 a month.

4. Your calls need judgement more than they need answering. If most of your inbound is existing clients with complicated or high-stakes situations — a lawyer mid-case, a doctor with a worried patient — an agent that takes a message is worth something, but you should be clear about which half of your calls it would actually be handling before you buy.

And one thing an AI agent is genuinely worse at than a person. It does not read distress well and it should not try to. Anything messy escalates to you. A human beats any AI at deciding to bend a rule for a customer who is upset; an AI beats any human at picking up on the second ring at 9pm on a Sunday. If your missed-call problem is the first kind, hire someone. /blog/should-i-hire-a-receptionist-or-use-ai-india runs that comparison properly.

Do this before you speak to any vendor, us included. It makes every sales conversation shorter and it is the only way to know whether the answer is yes.

Step 1 — count the misses. Open your call log, last 30 days. Count calls that rang out and were never connected. Collapse repeat calls from the same number within an hour into one. That is M.

Step 2 — sample twenty of them. Take twenty of those numbers. How many were a new customer wanting to buy? Not a supplier, not a courier, not a wrong number, not someone selling you something. That share is G, and it is the number that most often surprises people.

Step 3 — use the close rate you already know. Of enquiries you actually speak to, what share become jobs? That is C. Do not optimise it upward.

Step 4 — take the median invoice. Last month's bills, middle value, not your best month. That is V.

Step 5 — be honest about recovery. How many of those missed callers did you or they eventually reach? That is R.

Then: `M × G × C × V × (1 − R)`. Compare it against the ₹7,600 and ₹2,600 figures in the cost table, adjusted for your own minutes. If it does not clear the ₹7,600 comfortably, wait.

Two honest notes on our own calculator. /tools/missed-call-calculator runs a version of this arithmetic in a form, and it is quicker than doing it on paper. But it carries built-in default assumptions for conversion and no-show impact that you cannot edit, and it has no recovery-rate input at all, so treat its output as a starting estimate rather than your answer. Where your own close rate or recovery rate differs from a generic default — and it will — the formula above is the one to trust.

Then argue with the thing. The live demo is at app.dvaarik.com/dvaarik-ai. Call it and speak the way your customers actually speak: half Telugu, half English, background noise, a place name it has never heard, interrupting halfway. Five minutes of that is worth more than this entire page. All 22 scheduled Indian languages are included at the ₹2/min Bharat Standard rate with no language surcharge, including code-mixed speech.

One rule that is not negotiable per client, since it comes up whenever missed calls are discussed: if we call your missed callers back, those calls go only to your own leads and enquiries — missed calls on your own number, your own form fills, your own past customers. We do not dial purchased or scraped lists at any volume, for any price. Calling hours are respected and the agent identifies your business at the start of every call. /blog/is-ai-calling-legal-in-india-2026 covers the TRAI and DLT side of that.

For the record, and because this page will be quoted: if a properly sourced Indian study existed — defined population, stated sample, published method, dated — we would cite it here with all four of those attributes visible, and we would link it. We would rather have that number than not have it. It would make this page shorter and easier to sell from.

We did not find one, so there is none.

The reason we are telling you this rather than quietly picking a plausible figure is that the alternative is worse for both of us. A statistic you cannot check is a statistic you cannot use to make a decision, and a vendor who will invent one for a blog post will invent one for a sales call. The three numbers on this page that are ours — ₹2 per minute, ₹2,000 per 30 days for the number, ₹5,000 a month for the founding Care Plan in months one to three — are all published, all checkable at /pricing, and all stay true if you quote any one of them on its own.

If you do have a sourced Indian figure for missed-call revenue loss, send it to us on WhatsApp. We will read it, and if it holds up we will put it on this page with the source attached.