Every iGaming brand wants streamers. Few have a real system for picking them, paying them, and proving they worked. We run a live network of active streamers on our own casino brand, Spartans. The difference between a program that scales and one that quietly bleeds budget comes down to three things: deal structure, attribution, and vetting. Get any one of those wrong and the other two don't save you.
The three deal structures that actually get used
In practice, almost every streamer deal is some version of these:
- Flat fee — pay per stream or per month, regardless of performance. Lowest risk for the streamer, highest risk for the brand. Works for testing a new partner before committing to anything bigger.
- CPA (cost per acquisition) — a fixed payout per qualified first-time depositor (FTD). Aligns incentives cleanly, but only works if your tracking is airtight — see attribution below.
- Revenue share (revshare) — the streamer earns a percentage of net gaming revenue from the players they bring in, often for the lifetime of that player. Higher long-term cost, but it's the deal streamers with real audiences actually want, because it pays out as their referred players keep playing.
Most durable programs end up hybrid: a smaller flat fee to cover the streamer's time, plus CPA or revshare on top. That protects the streamer from a bad week of variance while still tying most of the payout to real results.
Attribution: the part everyone underbuilds
A deal structure is worthless if you can't prove which players came from which streamer. The baseline setup is a unique promo code or referral link per streamer, tied to a deposit window (typically 7–30 days from click) so late conversions still get credited correctly. Beyond that:
- Track first deposit and depositor retention — a streamer who brings in players that churn in a week isn't actually delivering CPA-grade value even if the raw FTD count looks good.
- Watch for code-sharing outside the streamer's own audience (their code showing up on deal/coupon sites) — it inflates volume without the trust that made the deal worth doing in the first place.
- Reconcile weekly, not monthly. Streamer relationships live or die on how fast you can answer "how did my stream do," and a month-long lag kills momentum on the partnerships that are actually working.
Vetting: audience quality over follower count
Follower count is the least useful number in a streamer pitch. What matters is audience geography (does it match your licensed markets), average concurrent viewers versus follower count (a real signal of an engaged, not bought, audience), and chat activity during actual gameplay versus giveaway spam. A streamer with 8,000 real concurrent viewers in a target geo will outperform one with 200,000 followers and 400 viewers, every time.
What this looks like running it ourselves
On Spartans, our own streamer network runs dozens of active partners at any given time, each on individually negotiated terms based on their audience and track record. Not a one-size-fits-all rate card. New partners start on flat-fee-plus-light-CPA to prove out fit, then graduate to hybrid or revshare once attribution data shows they're worth it. Same structure and process NOVYX runs for client brands — nothing gets simplified for the pitch deck and then run differently in practice.
FAQ
What's the biggest mistake brands make with streamer deals?
Starting new partners on a large flat fee or generous revshare before attribution data proves they're worth it. Start small, prove fit, then scale the deal — not the other way around.
How long should an attribution deposit window be?
Typically 7–30 days from click, depending on your average consideration time before a first deposit. Too short and you undercredit streamers for real conversions that took a few extra days. Too long and you're crediting traffic that has nothing to do with the stream.
Should follower count factor into vetting at all?
Only as a tiebreaker, never as the primary signal. Concurrent viewers relative to follower count and real chat activity during gameplay tell you far more about whether an audience is engaged or just following passively.