iGaming CPA FTDhow to calculate depositor acquisition cost

# iGaming CPA FTD: how to calculate depositor acquisition cost

The media buying call usually starts well: traffic is live, deposits are coming in and CPA FTD looks great. Then someone opens the ad account, tracker and backend — and finds three different numbers. Welcome to the classic affiliate game of “which dashboard says we are profitable?”

Before scaling a number that only looks good on paper, agree on what counts as an FTD, which costs belong in the numerator and which cohort owns the deposit. This guide applies that logic to Meta Ads, in-app and other sources. Our article on iGaming marketing and depositing cohorts covers the wider model; here we stay with CPA FTD.

Decide what an FTD is — not just what looks like a deposit

An FTD is not every `deposit` event that lands in the tracker. It is one unique player whose first successful deposit is confirmed by the product backend. That gives us four rules:

1. a player enters the denominator only once across the lifetime of the account;
2. declined, reversed and test transactions are excluded;
3. later deposits belong to retention and LTV rather than the FTD count;
4. the player is assigned to a source and campaign under a documented attribution rule.

The ad platform needs the event for optimisation and the MMP needs it for attribution. The backend confirms the money. Otherwise, “deposits” may turn out to be tests, duplicates or failed payments.

One player is still one FTD, however enthusiastic they are

A player can retry a payment, switch methods and deposit again later. Great player, still one FTD. Separate `first_deposit_confirmed`, triggered once, from `deposit_confirmed` for every successful payment. Names can vary; the logic and stable `player_id` cannot.

The formula is easy. The cost scope is the trap

The base formula is:

`CPA FTD = cost assigned to the cohort / unique confirmed FTDs in that cohort`

The fun starts with the word “cost”. A buyer may mean media spend; the business may also mean fees, creative and infrastructure. Keep three metrics:

| Metric | Numerator | Decision it supports | | --- | --- | --- | | Media CPA FTD | Actual platform media spend only | Campaign optimisation and media buying comparisons | | Acquisition CPA FTD | Media spend plus network or tracking fees directly tied to acquisition | Like-for-like channel comparison | | Fully loaded CPA FTD | All direct costs: media, fees, creative, infrastructure and agreed agency cost | Financial assessment of the acquisition programme |

Do not compare apples with invoices. Meta Ads on pure media spend versus an in-app bill with fees already has a preselected “winner”.

Worked example

Say a setup brings 80 FTDs on €8,000 media spend. The dashboard-friendly answer is €100. Add €800 in fees and it becomes €110. Add €1,200 in creative cost and it is €125. All three are correct — they answer different questions. These are illustrative figures, not client results.

Do not divide today's spend by last week's players

For performance decisions, group users by click or install date and let FTDs mature in a fixed D7 or D30 window. That shows the quality of traffic actually bought during the period.

A “deposits today” report is useful for operations, but it may contain players acquired last week. Dividing them by today's spend creates a beautiful CPA that collapses during reconciliation. Do not kill a fresh cohort too early either: some players simply have not reached the cashier yet.

Who is right: the ad account, tracker or backend?

Short answer: all three, within their own jobs.

• the ad platform provides spend, campaign, currency and spend date;
• the MMP or tracker joins the source to a post-install event;
• the backend confirms the player, transaction status, amount and first-deposit state.

For reconciliation, retain `player_id`, `first_deposit_id`, FTD time in UTC, status, amount, currency, acquisition date, source and campaign ID. Keep the original timestamp: AppsFlyer and Adjust document that receipt time may be used when it is missing. That is how a late-night deposit quietly moves into tomorrow's report while the buyer wonders where the conversion went.

When one player suddenly becomes two FTDs

Exclude test accounts, failed payments and technical duplicates first. A retried postback must not magically turn one player into two depositors, so use a stable transaction or event ID. Then compare for every campaign:

• attributed registrations in the tracker with backend registrations;
• the MMP FTD event with unique backend FTDs;
• first-deposit amounts and currency conversion;
• the share of FTDs without a source and events without a `player_id`;
• the delay from registration to first deposit.

If the numbers disagree, do not pick the dashboard with the nicer CPA. Keep the variance visible and find the cause: attribution window, late postback, timezone mismatch, reversed payment or missing ID.

A cheap FTD can still be expensive

A setup can deliver cheap first deposits and almost no redeposits. The buyer's column is green while the business is not. Read CPA FTD alongside:

• registration-to-FTD conversion rate;
• average and median first-deposit amount;
• D7/D30 redeposit rate;
• net revenue or another agreed value metric;
• ROAS for the same cohort and maturity window;
• refunds, fraud and excluded test accounts.

That is why our in-app traffic guide for iGaming looks beyond a cheap CPI. Managed acquisition is covered by the In-app Traffic for iGaming service.

What to check before adding budget

1. FTD means one unique, confirmed first depositor.
2. Cost scope is labelled media, acquisition or fully loaded.
3. Currency, acquisition date and D7/D30 window are consistent.
4. Backend and MMP data are reconciled by ID and timestamp.
5. Duplicates, tests and failed payments are excluded.
6. CPA FTD is reviewed with redeposit, revenue and ROAS.

If CPA keeps jumping, the tracker argues with the backend and every source has its own definition of FTD, start with one cohort from spend to redeposit. Explore IK.Agency's iGaming performance marketing service or discuss measurement with the team. More budget rarely fixes a broken report.