What a streamer sponsorship costs, and how to budget one

The first question on every brief is what it costs. There is a real benchmark — roughly a dollar per concurrent viewer, three to four times that in the United States — and it will get you to a sane opening position in a minute. What it will not do is explain why one creator quotes $500 and another quotes $2,000 for the same audience on the same night. That gap is about what is in the package, not about the audience.

The benchmark: $1 per concurrent viewer

There is a working rule of thumb, and it is simpler than most briefs expect. A placement tends to price at roughly $1 per average concurrent viewer. A channel averaging 3,000 concurrent viewers is a roughly $3,000 conversation for a placement on one stream.

The multiplier is where the range lives. In the United States, rates for the same delivery frequently run $3 to $4 per concurrent viewer — three to four times the baseline, driven by advertiser demand rather than by anything different happening on screen. Latin America and much of Asia sit at or below the baseline for comparable audiences, which is why region is usually the largest single efficiency lever available to a campaign.

Treat this as an opening position, not a quote. It is the number to sanity-check an ask against — a creator quoting five times the baseline for a plain placement in a non-premium region needs a reason beyond enthusiasm.

Why the same channel can cost four times more

Because “a placement” is not one thing. What lifts a price far above the per-viewer baseline is almost never a bigger audience — it is a longer list of deliverables bundled into one deal.

A single on-stream banner might sit at $500. Add a banner in the stream description, a pinned link and chat commands pointing at the product, spoken mentions to a talk-track, and a QR code routed to the shop, and the same creator on the same night is a $2,000 package. Nothing about the audience changed. The brand simply occupies four more surfaces, and three of them are attributable.

This is why comparing two quotes on price alone is meaningless. Compare them on which placements are actually included — the same headline figure can buy a logo nobody looks at, or a full package with three trackable routes out of the stream.

Marketplaces price differently

Alongside direct deals there are marketplaces that pay creators per thousand impressions rather than per activation. That model is familiar to anyone who buys digital media, it scales without negotiation, and it is genuinely efficient for pure reach.

It also systematically undervalues what live streaming does. An impression model prices a moment; a sponsored stream delivers hours of continuous presence, in a room where the audience is actively discussing what is on screen. Buying live on a CPM basis tends to get you the passive end of the placement ladder — a banner, an overlay — and none of the endorsement that made the channel worth buying. Use marketplaces to top up reach; do not run a consideration brief through one.

Why published rate cards mislead

Most circulating figures take a follower or peak-viewer count and multiply it by a constant. Three things break that. Followers do not watch — a channel with 800,000 followers and 2,000 concurrent viewers is a 2,000-viewer channel. Peak is one instant, not the session. And a single constant assumes every category and every region prices the same, which is the opposite of true.

The practical consequence is that a rate card sets an expectation the market does not honour, and the negotiation starts with both sides disagreeing about reality.

What actually sets the price

• Average concurrent viewers

  Weight

  The base the price is built from — but ACV over a recent window, not the best night the creator screenshots.

• Category

  Weight

  The same audience size costs very differently in a competitive shooter, a casino category and a variety channel. Advertiser demand per category sets the floor.

• Region and language

  Weight

  Latin American and parts of Asia price well below North America for comparable delivery. Often the single biggest efficiency lever available.

• Share of the stream

  Weight

  A banner for a session costs a fraction of a stream built entirely around the product.

• Exclusivity

  Weight

  Locking a creator out of competitors for a period is priced as lost future revenue, and can exceed the activation fee itself.

• Usage rights

  Weight

  Reusing clips in your own paid media is a separate licence. Frequently forgotten at negotiation and expensive to add afterwards.

• Deliverable count

  Weight

  Multi-stream packages usually price below the sum of their parts; a single one-off carries a premium.

• Lead time

  Weight

  Short-notice bookings cost more, because the creator is rescheduling around an existing calendar.

Budget backwards instead

You do not need to know rates to build a budget. You need to know what an outcome is worth to you. The sequence is the same every time:

• Start from the outcome. Installs, signups, or a reach target against a defined audience. Attach a value you would happily pay for one.
• Convert it to an acceptable CPM — cost per thousand hours watched, or per thousand relevant viewers if the brief is reach. This is your ceiling, and it is the only number you control.
• Estimate delivery from measured history, not from a media kit. What has this creator actually averaged, in this category, over a recent window?
• Divide the asking price by the estimate. Anything under the ceiling is a candidate. Anything over it needs a reason beyond enthusiasm.
• Hold back a contingency. Ten to fifteen per cent covers reschedules, a creator underdelivering, and the extra deliverable you will want once something works.

This also settles the argument that consumes most first campaigns — big channel or several small ones. Run both through the same division and the answer stops being a matter of taste. Frequently the small, category-specific channels win, and by a wide margin.

Where budgets go wrong

Paying for followers

Follower count is a historical artefact. It records everyone who ever clicked follow, including during a viral moment four years ago. Concurrent viewers over a recent window is the number that predicts delivery.

Buying peak, receiving average

A creator quotes on the strength of a 40,000-viewer night. The brand is on screen for a session that averages 4,000. Both numbers are true; only one describes what was bought. Insist on average concurrent viewers over a stated window, and check it against independently measured history.

Forgetting usage rights

The clip performs, someone wants it in a paid social campaign, and the licence was never negotiated. Retrofitting it costs more than including it, because at that point the brand has revealed how much it wants it.

Treating exclusivity as free

Asking a creator not to work with competitors for six months is asking them to forgo revenue. Priced properly it can exceed the activation fee. Ask for it only where a competitor appearing would genuinely damage the campaign.

A realistic first budget

For a brand entering live streaming, the useful shape of a first budget is not one large activation. It is several small ones across different categories and creator tiers, sized so that each can fail without spoiling the test, with reporting good enough to tell you which succeeded and why. The output of a first campaign is not reach — it is a rate benchmark for your own category that no published rate card can give you.

Once you have that, the second campaign can be concentrated with confidence. Which format to concentrate it into is covered in the nine formats.