Pay-Per-Minute Video Calls: A History of the Billing Minute

September 20, 2026·11 min read

Almost every platform that sells a video call sells it by the minute. A creator posts a per-minute rate, a fan buys minutes, and a call that runs sixty-one seconds is billed as two.

That convention is older than video calling, older than the consumer internet, and older than the computers that do the billing. It is a seventy-five-year-old accounting compromise that almost nobody has revisited since the constraint which produced it disappeared.

This is where the minute came from, how it survived three generations of technology that no longer needed it, and what happens to paid video calls when you finally take it out.

The minute was an accounting limit, not a unit of conversation

Before calls were dialed, they were written down. A long-distance call in the 1940s was set up by an operator who marked a paper ticket with a “rate step” for the route and read the price off a table — in terms of the first three minutes and each additional minute. The duration of every toll call was recorded by hand.

The minute was not chosen because a minute is a meaningful amount of conversation. It was chosen because it was the smallest unit a human being with a pencil, a ticket and a switchboard full of other calls could record accurately, thousands of times a day. It was a limit of the billing apparatus, and in 1949 the billing apparatus was a person.

1951: the machines start counting, and the minute stays anyway

On 10 November 1951 the mayor of Englewood, New Jersey dialed the mayor of Alameda, California with no operator in between — the first customer-dialed long-distance call, and the beginning of the end for manual ticketing. It connected in eighteen seconds. (A few towns from where Seconds is based, as it happens.)

Behind it sat automatic message accounting, first deployed by AT&T in 1948: equipment that punched each call’s calling number, answer time and disconnect time into a roll of paper tape, which a technician cut daily and sent to an accounting centre.

Read that list of fields again. Answer time and disconnect time. From 1948 onward the machinery was recording both ends of every call, which means it knew the length of one far more precisely than a minute. The minute had stopped being a measurement limit and become a formatting convention in the accounting system downstream.

Formatting conventions are much harder to kill than constraints, because nothing breaks when you keep them. A constraint announces itself the moment it is lifted. A convention just sits there, collecting the benefit of the doubt, for another seventy years.

1980: the minute becomes the product

Area code 900 was installed in 1971, and its first known use was a call-in programme — “Ask President Carter”, in March 1977. In 1980 AT&T restructured the 900 prefix as a premium-rate code: callers paid above-normal per-minute charges, and the difference went to whoever was on the other end.

The application that made it famous was a poll. After the Reagan–Carter debate in 1980, ABC’s Nightline invited viewers to call one of two numbers, at fifty cents each, to say who they thought had won. Within a few years the same mechanism was carrying advice lines, sports scores, horoscopes and technical support: the first mass market in which the conversation itself was the product, sold by time.

That is the direct ancestor of the paid video call. Not the video — the transaction. A person makes themselves reachable, a caller pays for the time, a platform meters it and takes a cut. Everything since has been the same deal with better cameras.

1993: the minute gets written into the rules

Pay-per-call grew faster than its own disclosure practices, and Congress answered with the Telephone Disclosure and Dispute Resolution Act of 1992. The FTC’s resulting 900-Number Rule took effect on 1 November 1993 and required that both the advertisement and the call’s spoken preamble state the cost per minute, any minimum charge, and the maximum charge where the length was known in advance.

It also said this:

The Commission recognizes that time-sensitive billing is done in one-minute increments and that any portion of a minute is billed as a full minute. This manner of billing will not violate the rule.

That sentence is the minute’s high-water mark. A consumer-protection rule written specifically to stop people being surprised by the price of a phone call looked directly at rounding every partial minute up to a whole one and concluded — correctly, for 1993 — that this was simply how billing worked. The regulator was not blessing the round-up. It was describing the equipment.

The effect was to freeze the convention in place. Rounding up was now not merely normal but explicitly compliant, and an industry with a compliant default has one fewer reason to ever look at it again.

The internet arrived, and the minute survived it

Skype launched in August 2003. Calls between Skype users were free, because the marginal cost of carrying one over the internet was approximately nothing. Calls out to the telephone network were not free — and they were priced, inevitably, per minute, at roughly two cents a minute to a US landline.

This is the moment the unit came loose from its cause. There was no operator with a ticket and no tape perforator. There was software that could timestamp both ends of a session to the millisecond. The minute survived for two much weaker reasons: it was the number customers already knew how to compare, and it was the number the billing systems already emitted.

Twenty years on, those are still the only two reasons. When a per-minute video call platform prices a creator’s time today, it is not making a pricing decision. It is inheriting a default set by a paper ticket in 1949 and ratified by a regulation written before the web had images.

Everywhere else, the minute has already lost

Here is the part that makes per-minute video calls look less like an industry standard and more like an oversight: every other metered industry has already abandoned the minute, and in each case the company that moved first used the smaller unit as the entire argument.

Mobile voice, 2009. Tata DoCoMo launched in India on 24 June 2009 priced at one paisa per second. Until then the smallest unit in Indian mobile tariffs was a minute, so a five-second call cost a full one. Ratan Tata called per-second billing a paradigm shift at the launch, and the rest of the market followed. Hundreds of millions of subscribers stopped paying for time they had not used.

Cloud computing, 2017. Google moved Compute Engine to per-second billing on 26 September 2017. AWS had announced per-second billing for EC2 eight days earlier, effective 2 October. AWS had been billing by the hour and Google by the minute; both kept a one-minute minimum, and both gave the same reason: charging for capacity nobody used is a tax on short workloads, and short workloads had become the normal case.

The pattern is consistent enough to be a rule. When metering is expensive, the unit is coarse. When metering gets cheap, the coarse unit stops being a constraint and becomes a charge for nothing — and someone shrinks it, and says so out loud, and takes the market’s attention with them.

It happened to the hour. It has happened to the minute twice already. It has not yet happened to the paid video call.

What the minute actually costs on a video call

Rounding matters more here than it ever did on a phone call, because the calls are shorter. At the $0.05/sec default rate on Seconds ($3.00/min), the arithmetic is unkind:

• A 65-second call costs $3.25 in total billed by the second, and $6.00 once it is rounded to two minutes — 85% more for running five seconds long.
• A 20-second hello costs $1.00 billed by the second, and $3.00 rounded up to the minute: 200% more, three times the price of the thing that actually happened.

Averaged over calls that end whenever conversations end, per-minute rounding bills about thirty seconds that nobody spent. On an hour-long consultation that is a rounding error. On the ninety-second call this whole category exists to make possible, it is a third of the bill.

The money is the smaller half of it. The larger half is what a rounding step does to the person on the call: it puts a cliff in the conversation, so that at 2:52 you are no longer listening but calculating. We took that argument apart in Pay-per-second video calls: why the billing unit is the product.

What the second changes for creators

A minimum billable minute quietly decides what a creator is allowed to sell. If the smallest unit is a minute, the shortest sellable call is a minute — too long to be a hello, too short to be a consultation — so the platform pushes everyone toward longer booked sessions, because that is the only shape the unit supports. The quick question and the ten-second “yes, ship it” become calendar invites, or they never happen.

It also makes the creator complicit in the round-up. Every short call overcharges their own audience, on their own page, with their name on it. Most creators deal with that by not taking short calls, which removes the highest-margin minute they have.

Per second, one rate serves a 15-second reaction and a 45-minute coaching session with no tiers, no packages and no minimums, because the only thing that varies is how long the meter runs. Rates on Seconds run from $0.01/sec to $10.00/sec for exactly that reason: the unit is small enough that the range on top of it can be enormous. Creators keep 70% of the time on every completed call; earnings sit in a pending balance for 7 days and are then withdrawn through our payout partner, Trolley.

Seconds is built on the second

Seconds is a pay-per-second video call platform, and the name is a specification rather than a flourish. The rate a creator sets is a per-second rate. The meter counts seconds. The charge at hang-up is seconds × rate, plus a flat $0.75 service fee and any applicable tax. There is no minimum, no rounding step and no minute anywhere in the billing path — not even the one-minute floor the cloud providers kept.

Most platforms in this category still quote a per-minute rate. We did not want to build a product on a unit we would then have to explain away, so the unit went in the name, and the rest of the product followed from it: a fan requests approval, the creator approves them and sets that fan’s per-second rate, and from then on the fan can call whenever the creator is online. The fan sets a maximum spend before connecting, so the ceiling is known in advance, and pays for the seconds actually used. Calls are end-to-end encrypted, with keys negotiated between the two devices and never sent to Seconds; there is no recording feature and nothing stored to replay.

Seventy-five years after the machines started recording both ends of every call and the bill kept arriving in whole minutes anyway, the round-up is the last part of that system still standing. It does not have to be.

Getting started

If you already have an audience, a per-second rate is the lowest-friction thing you can offer it — nothing to produce, nothing to schedule. Claim your username and set a rate, or read Setting your per-second rate for how the ladder works. Placing a call instead? How to call a creator covers the fan side, and Paid video calls: how creators get paid for their time walks through the category as a whole.

Frequently asked questions