An operator that picks a team on cost per deposit is almost guaranteed to overpay. The cost of an FTD is what you pay. The cohort is what you get.
What a cohort is
A cohort is a group of players who arrived in the same period. You look at it not at a single moment but by lifetime days: what percentage stays active on day seven, day thirty and day ninety. That curve is what shows whether the buy pays off; the number of registrations shows nothing.
What normal looks like
For online casino on live traffic the curve usually looks like this:
- D7 — about 50%. Half drop off in the first week, and that is normal.
- D30 — about 25%. By month one a quarter of the cohort is left.
- D90 — 12–15%. This is the core that actually brings the money.
If a team shows 44% active at D30, the number is either from a different vertical or counted by a different definition of activity. Ask what counts as activity: a login, a bet or a deposit. The difference between these three definitions is several times over.
Three signs of a dead cohort
A sharp drop between D1 and D7
If 20% remain on day seven instead of 50%, the traffic was incentivized: people were paid to register and deposit, and they never intended to play.
High share of minimum deposits
When 80% of deposits land exactly on the lower qualification limit, that is threshold hunting, not interest in the product.
Deposits cluster in time
A hundred FTDs in two hours overnight is almost always one person with an account farm, not a lucky campaign setup.
What to ask for in the report
The minimum set any team owes you: a breakdown by cohort lifetime days D1 / D7 / D30 / D90, the redeposit rate, the average size of the second and third deposit, the share of rejected FTDs with reasons, and a breakdown by campaign setup, not only by geo.
If instead you get a single line saying "we brought N deposits", you will not be able to tell good traffic from bad until you lose money.
Why a cheap deposit costs more in the end
What you have to count is not the FTD price but the price of an active player on day thirty. Take two offers: the first is 120 dollars per deposit at 25% active on D30, the second is 80 dollars at 8%. The second looks a third cheaper. In fact an active player costs 480 dollars in the first case and 1,000 in the second. Twice the difference, and it only shows up a month after the invoice is paid.
So a team selling "the cheapest FTD on the market" is usually selling incentivized traffic. It is only cheaper at the moment of purchase.
What to do if the cohort drops
First, do not switch the whole channel off. A drop is almost always local: a specific campaign setup, creative or payment method. A breakdown by campaign setup will show where exactly; a breakdown by geo alone will not.
Second, check the payment step. Some of the "bad traffic" is in fact good traffic that ran into a form nobody uses in that geo. In Brazil we ran into exactly this: deposit conversion sat at eleven percent until PIX appeared in the form.
Third, compare against a baseline. If you do not have your own historical cohort yet, use the benchmarks above, but remember: poker and slots have different curves, and comparing them with each other is pointless.
How these numbers look on real launches — in case breakdowns: the same page has the retention curve we report against.
