What does a healthy deposit approval rate look like in Brazil?

Question

Open question for anyone running payments on the Brazilian market, answers in Portuguese are equally welcome, the thread reads in both languages.

Approval rate is one of those numbers everyone quotes and almost nobody defines the same way. Before comparing anything, it helps to align:

- Do you count **attempts** or **unique deposit intents**? A cascade retrying three times can make the same failure look like three.
- Do you exclude abandoned sessions where the player never reached the payment step?
- Do you separate PIX from card and from other methods, or report a blended number?

What I would like to hear:

- What approval rate do you consider healthy per method today?
- What is the single biggest cause of decline in your operation right now?
- How much of the gap is provider-side versus something you control (data quality, flow design, limits)?

Concrete numbers are useful, but the definition behind them is what makes them comparable.

André Beltrão · Accepted Answer

Good question, and the definitional part matters more than the number.

**How we count:** unique deposit intents, not attempts. A cascade retrying twice is one intent, one outcome. Counting attempts is the single most common way people report an approval rate that looks worse, or better, than reality, depending on where the retries land.

We also exclude sessions abandoned before the payment step. Those belong to a funnel problem, not a payments problem, and mixing them hides both.

**Never blend methods.** PIX and card behave nothing alike, and reporting one number for both means you will never see either one degrade.

**Biggest cause of decline right now:** in our operation, most of what we lose is not a hard decline at all, it is intent that never completes, mostly timeouts and users dropping at the confirmation step. Actual risk-based refusal is a small slice. That surprised everyone when we first broke it down, and it moved our roadmap from "negotiate with the provider" to "fix our own flow".

**Provider versus us:** roughly speaking, a meaningful part of the gap was ours, data quality on the request, limits configured too tightly, and a confirmation screen that asked for one field too many. Worth breaking that down before assuming the provider is the problem.

Happy to compare definitions with anyone running PIX at volume. Comparing rates without comparing definitions is how the whole market ends up quoting numbers nobody can reproduce.

João Vinícius · Answer

Adding one thing from the acquisition side, since approval rate is usually treated as a payments-only metric.

Approval varies a lot **by traffic source**. Some sources bring users who are simply less likely to complete a first deposit, wrong device, wrong expectations, sometimes the wrong country entirely. When we started reporting approval segmented by source, two campaigns that looked fine on cost per FTD turned out to be dragging the whole payments dashboard down.

So if your approval rate dropped and nothing changed in payments, check what changed in media buying that week. More often than you would expect, that is the answer.

What does a healthy deposit approval rate look like in Brazil?

Respostas2

Compartilhar discussao