Where 30% of influencer traffic quietly disappears
Postbacks fire, dashboards agree, and a third of the deposits still never reach the affiliate's ledger. We audited eleven programmes across Tier-1 to find where attribution breaks — and what the fix actually costs.
A Tier-1 casino programme paid out on 1,402 first-time deposits last quarter. The affiliate who sent that traffic counted 2,061. Neither side was lying, and neither side could explain the gap for six weeks — which is roughly how long it takes for a profitable campaign to be switched off on suspicion.
We asked eleven programmes to hand over raw click, registration and deposit logs for the same set of campaigns their affiliates were running, then reconciled them line by line. In nine of the eleven, the loss sat between 19% and 34%. In the two clean ones, the difference was under 3% — and both had done the same unglamorous thing.
The click never had an owner
Influencer traffic arrives without a referrer. A viewer sees a code on stream, opens a browser three hours later and types the brand name. There is no click to attribute, so the deposit lands in the operator's direct bucket and stays there. Every programme in our sample called this "organic uplift" and none of them paid for it.
The second failure is more mechanical: cross-device. A link opens in the TikTok in-app browser, the user registers there, then deposits from a desktop the next morning. Without a durable identifier the two sessions never join up.
"We were not being robbed. We were being rounded down, every single day, by systems nobody owned."
Head of affiliates, Tier-1 casino brand
What the numbers look like
Averaged across the nine programmes with a measurable gap, here is where the missing volume went:
The fix nobody wants to sell you
Both clean programmes were doing promo-code-first attribution: the code, not the link, is the primary key, and it is captured at registration by the operator's own form. The tracker reconciles against it nightly. It is boring, it requires the operator's product team to care, and it recovered 22% and 29% of previously unattributed deposits respectively.
The cost, in both cases, was two sprints of engineering time and a weekly reconciliation call. Set against the volume in dispute, that is the cheapest thing in this article.
None of the eleven programmes asked to be named in this piece. Four have since published reconciliation terms in their affiliate agreements — which is, in the end, the only version of this story that changes anyone's payout.