What a support ticket actually costs you (the math nobody does)

A support ticket does not cost you the two minutes it takes to answer. Once you load in salary, context-switching, and the revenue lost to slow replies, a single resolution costs most small businesses somewhere between ₹40 and ₹120 — call it $1 to $4 — and the ones that feel cheapest are usually hiding the most. This post does the math nobody does, for three real-shaped businesses, and hands you the formula to compute your own number.

The reason nobody runs this math is that support does not arrive as an invoice. It arrives as a slow leak of hours and half-lost sales, spread across a month, easy to feel and hard to count. Support is the biggest cost in most small businesses that never appears as a line item. So let us give it one.

The cost of a single resolution

Start with the honest unit cost of one human-answered ticket. It has three parts, and the first is the only one people usually count.

Handling time is the obvious part: the minutes actually spent reading and replying. For a typical support ticket that is 5 to 15 minutes. Take the loaded hourly cost of whoever answers — salary plus overhead, not just take-home — and multiply. A support agent in India runs a loaded cost of roughly ₹200 to ₹400 an hour; in the US or Europe, more like $20 to $40. At 10 minutes a ticket, that alone is ₹35 to ₹65, or about $3 to $7, before anything hidden.

The context-switch is the part people forget, and it is often bigger than the handling time. A ticket does not just cost the ten minutes of the ticket — it costs the several minutes of lost focus on either side, as the person pulls out of what they were doing and later climbs back in. A ten-minute ticket is rarely a ten-minute interruption; it is a twenty-minute hole in someone's afternoon. For a maker — a founder writing code, a creator filming — that broken focus is the real tax, and it does not show up in any handling-time report.

The third part is who is answering. A ticket answered by a founder is not priced at an agent's rate — it is priced at the founder's, which is the most expensive labor in the company. When a solo creator answers a WISMO message, the cost is not ₹40 of support time. It is a slice of the one person whose hours should be going into the product, the launch, the thing only they can do.

Three businesses, three bills

Here is the monthly math for three common shapes. Numbers are rounded and directional — plug in your own rates, but the structure holds.

The solo creator: ~300 tickets a month. That is about 10 a day, mostly access questions, refunds, and "does the course cover X." At 10 minutes handling plus context-switch, call it real cost of 15 to 20 minutes each. Three hundred tickets is roughly 75 to 100 hours a month — half a working month, gone. Priced at agent rates that is maybe ₹20,000 to ₹35,000. But it is not agent rates, because there is no agent; it is the founder, and the true cost is the launch that slipped because the mornings went to the inbox. The solo creator's support bill is not paid in rupees. It is paid in the product that didn't get built.

The 3-person team: ~1,500 tickets a month. About 50 a day, shared across three people who also do other jobs. At a loaded ₹250 to ₹350 an hour and ~12 minutes real time per ticket, that is roughly 300 hours a month of support labor — the equivalent of nearly two full-time people, ₹75,000 to ₹1,00,000 in loaded cost, smeared across three calendars so nobody notices it adding up. This is the stage where support quietly eats the team's ability to do anything else, and where the case for a dedicated hire starts forming — usually a year before the volume actually justifies one.

The 10-person team: ~8,000 tickets a month. Now support is a real function with real headcount. At this scale the per-ticket cost is lower (specialists are faster, macros exist) but the totals are large: 8,000 tickets at ~10 minutes is well over 1,300 hours, several full-time roles, comfortably ₹4,00,000-plus a month in loaded cost, or tens of thousands of dollars. Here the hidden cost shifts from founder-hours to churn: at this volume, slow answers are quietly canceling subscriptions faster than anyone is tracking.

The hidden costs that dwarf the salary

Two costs never make it into the spreadsheet, and both are usually bigger than the labor.

The first is churn from slow answers. A customer who waits a day for a reply is measurably more likely to leave, and a lost customer costs you their entire remaining lifetime value, not one ticket. If even a small share of your slow-answered tickets nudge a customer out the door, that leakage alone can exceed your whole support salary line. The most expensive support ticket is the one that was answered too late to keep the customer.

The second is founder time priced honestly. Founders routinely value their own hour at zero when it goes to support, because it does not feel like spending money. It is the most expensive spending you do. An hour a founder spends on WISMO is an hour not spent on the one or two things that actually move the business — and at seed stage, that trade is close to the whole game.

The deflection math

Now the good news, because there is a lot of it. The knowledge half of your queue — the repetitive how-do-I and policy questions — can be absorbed by a help center plus an AI that answers from it, without a human ever touching them. In practice, a decently seeded help center lets automation resolve 40 to 60 percent of total volume. Not the interesting tickets — the boring, repetitive majority.

Run that against the three businesses. For the solo creator, deflecting half of 300 tickets returns roughly 40 to 50 hours a month — most of a working week, handed back to the product. For the 3-person team, absorbing half of 1,500 is ~150 hours a month, close to a full-time person you did not have to hire. For the 10-person team, deflecting even 40 percent of 8,000 tickets is over 500 hours — multiple headcount worth of capacity, or the difference between hiring three more people and hiring none. Every point of deflection is either an hour returned or a hire deferred — and at these volumes, it compounds fast.

The clean way to see the value: a resolution that used to cost you ₹40 to ₹120 of loaded human time now costs a fraction of that when the AI handles it. Multiply the gap by your deflected volume and that is real money, every month, appearing not as a new expense but as a cost that stopped.

Compute your own number

Here is the formula. Fill in five values and you will know your real support bill:

1. Tickets per month (T).
2. Real minutes per ticket, handling plus context-switch (M) — be honest; it is more than the handling time alone.
3. Loaded hourly cost of whoever answers (R) — salary plus overhead, and use the founder's rate if it is the founder.
4. Monthly cost = T × (M ÷ 60) × R.
5. Deflection value = that monthly cost × 0.4 to 0.6.

Step 4 is the number you are quietly paying now. Step 5 is what a help center plus grounded AI is worth to you every month — the portion of the bill that can simply stop. Most businesses find step 4 is far bigger than they guessed, and step 5 pays for the tooling many times over. If you want the inputs measured rather than estimated, a good support tool will show you your actual ticket volume, resolution split, and response medians directly.

That is exactly what we built Iris to do: answer the repetitive majority from your own published help center — instantly, at any hour, and silently handing the genuinely human tickets to your team — while Reports shows you the volume, the deflection, and the response times so the math above stops being a guess and becomes a dashboard. The cost of a support ticket is real whether or not you measure it. The businesses that thrive are simply the ones who measured it, then made the boring half answer itself.