What an hour of livestream costs

The cost of a stream hour worked out end to end: the published rate, the per-platform multiplier, and the break-even maths.

Almost every question about live selling turns into the same question once you try to budget it: what does an hour on air cost, and does that hour make more than it costs?

Most vendor pricing pages answer neither. They quote a seat, a session or a "contact us", and a seat is not the thing you consume. An hour is. This is the cost of a stream hour worked out end to end, including the parts that are not on anyone's price list.

The published rate

Genpio prices by the stream hour, and the rate falls as the plan grows.

That is arithmetic you can check rather than a quote: 240 hours at $13 is $3,120. Every plan carries the whole stock avatar catalogue and the whole public voice catalogue, so the hour is the only dial that moves the bill.

Prices are in US dollars. Pay in another currency and you are billed in that one, converted at the day's rate with 3% added for the cost of taking a payment across a border, with a tax-compliant invoice for your country.

The line most budgets miss: hours are counted per platform

This is the one that changes a spreadsheet after it has been approved.

An hour live on TikTok, Shopee and YouTube at the same time spends three stream hours, not one. The session is one session and the work is one setup, but you are buying air time on three destinations and each one is counted.

So the honest planning unit is not hours in the day, it is hours times destinations. A shop that wants six hours a day on two platforms is asking for 360 hours a month, which is Growth with room to spare rather than Starter with a shortfall. Work that multiplication out before you pick a plan, because it is the difference between a plan that fits and a plan that runs out in the third week.

The useful consequence: simulcasting is a choice to price, not a free bonus. Two destinations at once is worth it when the second one sells. It is worth nothing when you added it because the box was there.

The four costs that sit outside the hourly rate

An AI host removes the host's fee and the studio booking. It does not remove these, and a budget that ignores them under-reads the first month.

• Catalogue preparation. Names, prices, variants, stock, shipping times, the ten questions your chat actually asks. This is the real work of the first week, and it is the input that decides whether the host can answer anything. It is also done once and then maintained.
• Product media. Photographs and clips that hold up on screen for an hour at a time. Whatever you have for your listings is usually thin for live.
• Somebody reading the transcripts. Not staffing the stream, reviewing it. An hour of unattended selling generates questions you have no answer for yet, and those belong back in the product records.
• Fulfilment and support downstream. Orders placed at 3am are picked, shipped and occasionally refunded by the same team as the rest. Overnight selling moves the bottleneck rather than removing it, which is the argument of running a livestream channel overnight.

None of those are large after the first month. All of them are real, and all of them are yours whichever vendor you choose.

The arithmetic that decides whether an hour is worth running

Here is the whole calculation, and it takes four numbers you already have.

1. Cost of the hour. The plan rate, times destinations, plus whatever you allocate for the work above.
2. Orders per hour in the slot you are considering. Not your best hour - the slot you are actually adding.
3. Average order value. Your own, not a regional benchmark.
4. Contribution margin per order after goods, shipping, platform fee and your return rate.

An hour is worth running when orders per hour times contribution margin beats the cost of the hour. Written out with placeholder numbers so you can put your own in: at $13 an hour on one destination, an order that contributes $4 after everything needs four orders in the hour to break even. Three orders loses money on that hour. Six pays for the hour and the hour after it.

That is a low bar in some categories and an impossible one in others, and the difference is your catalogue rather than your country. Which is why the only benchmark worth having is the one measured on your own products - the argument we make with the published figures in livestream commerce data in 2026.

Two things fall out of the formula that surprise people:

• A cheap hour does not fix a bad slot. If nobody is watching, a lower hourly rate just loses money more slowly. Test slots against traffic, not against price.
• The marginal hour is what matters, not the average. Your peak hours were always going to be profitable. The question an hourly rate answers is whether hour nine, and the overnight block, and the second language, clear the bar. Those hours were never staffed before, so nobody knows until they are run.

Where an AI hour genuinely wins, and where it does not

It wins on the hours a person cannot be scheduled for at any sensible cost: overnight, weekends, public holidays, and the language you have no one to staff. Those hours are not free, they are simply purchasable in units of one hour instead of units of one employment contract.

It does not win on the hours that already work with a host your audience follows. A human host builds a following, reads a room, improvises an offer and negotiates. If your peak hour converts because of who is on screen, the hour is not a commodity and the cheaper hour is the wrong trade. That is the whole argument of AI host or human host, and the answer is per hour rather than per brand.

What to do with this

Work out your hours times destinations, pick the plan that covers it, then run the slots nobody was covering and measure contribution per hour rather than views. One month of that is worth more than any benchmark, including ours.

There is no trial, free or paid. Book a demo at meet.genpio.com, watch it run on your own catalogue with your own awkward chat questions, then start on the plan that matches the hours you just worked out. The full rate card, the yearly terms and what each plan includes are on pricing.