The listed price of an activation is not the price of a result. Three sources quietly eat into the budget: cancelled activations, repeated attempts on a route with poor deliverability, and an already-rented number that simply sits idle without tasks. Each loss looks small on its own, but over a month it adds up to a noticeable share of spend. Here is how to see it and calculate it — and why the cost of an attempt and the cost of a result almost never match.

Three sources of invisible loss

The first is cancelled activations: a number is ordered, the code does not arrive or arrives too late, the activation is cancelled. The second is repeated attempts on a weak route: the same country and carrier fail to deliver a code again and again, and you pay for another attempt each time. The third is an idle rented number: a long-term number is rented for a month but is actually used on three or four days out of thirty, while the rental is billed for the full paid term regardless of how much work the number actually did.

Cancelled activations: a quiet but real cost

The refund for a code that never arrived is usually automatic: the money returns to the balance if nothing came through within the timeout. But a cancelled activation also costs time — the waiting window spent, reordering a number, retrying registration on the same service. When a route fails systematically rather than once, the habit of "order one more number and see" turns into dozens of cancelled attempts a month, each costing working time even when the amount is refunded in full. If you are unsure whether to cancel or keep waiting for the code, the step-by-step breakdown is in what to do if the code never arrives, not by SMS, not by email.

Repeated attempts on a weak route multiply the cost of success

The metric that decides everything is the delivery rate (success rate) of a specific country-carrier pair for a specific service. If a route's delivery rate is 40%, you need 2.5 attempts on average for one successful verification — the real cost of the result is 2.5 times the price of a single activation. At a 90% delivery rate, the same math gives a factor of roughly 1.1. That is why a route priced at half but delivering at 40% and a route priced double but delivering at 90% land on different outcomes: 0.5 × 2.5 = 1.25 units per success versus 1 × 1.1 = 1.1 — the cheap route ends up costing more per result.

An idle rented number without tasks

A long-term number earns its keep when tasks keep coming to it: several verifications a week, incoming calls, an account with active messaging. If a number is rented "just in case" and sits idle for most of the paid term, the cost of a single code it delivers can end up several times higher than a one-off activation — simply because the paid term is divided across a handful of real uses. An idle number is not saving money, it is deferring the spend to a future use that may never happen.

How to calculate the real cost of a successful verification

The formula is simple: total spend on a route for a period — successful activations plus cancelled ones plus the cost of renting an idle number — divided by the number of codes actually received and used. That is the cost of the result, not the cost of an attempt. The raw data lives in the transaction history: every activation is listed as its own line with a status and an amount, including cancelled and refunded ones. Checking the history over a week or a month makes it easy to spot which route is dragging the budget down, even when a single attempt looks attractive on the OTP activations page. The rule is simple: count the cost of the result, not the attempt, and decide whether to switch routes based on that figure rather than the per-unit price.

Frequently Asked Questions

How do I know a route is worth switching rather than retrying again?

If the delivery rate over the last 5–10 attempts on one route is below 50%, the real cost of a result is already higher than on an alternative route with a higher per-attempt price. You can check this against the transaction history for the same period.

Does the refund for a cancelled activation cover the whole loss?

It covers the amount but not the time: reordering a number, retrying registration and the missed window for a task are not part of the refund. That is why systematic cancellations cost more than the balance alone suggests.

How quickly can I see that a rented number is sitting idle?

Compare the number of days rented against the number of days tasks actually came in — both figures are visible in the transaction history. If the gap exceeds half the term, the rental is being used inefficiently.

Check which routes and numbers actually pay for themselves: the full history of activations, cancellations and rentals is in the transaction history section.