Fintech marketing: from sign-up to repeat payments

A payment product in Almaty uses two advertising channels. Each spent 200,000 ₸. Channel A produced 100 registrations and B only 50, so the marketer proposes moving the budget to A. A product colleague opens the next stage: 20 people from A and 30 from B made a first purchase payment. Cheap registration has not yet settled the channel choice.

We will use a debit payment product. All figures below are illustrative, including the bonus and financial results. A loan, deposit or investment service would need different outcome and cost definitions; this calculation concerns paying for purchases.

Name the action the person came to perform

Registration creates a system record. Identity confirmation is a document check, and without it the product will not let the person pay. A first completed purchase payment shows that someone has begun using the payment product. These events happen at different times and may need help from different employees.

In our report, a first payment is a completed merchant purchase within 30 days of registration, confirmed by the transaction record. Test and cancelled transactions are excluded, and another notification about the same transaction does not create a new one. Count each person once regardless of purchase count.

Everyone in both groups has completed the full 30 days. The picture makes it easy to see how the result changes after registration.

The user needs the "payment completed" screen, but the analyst reconciles the event with the transaction record. If a refund arrives, recalculate under the documented rule. The report needs one confirmation source, or the app screen and payment records may produce different totals.

Use the same cost base

Registration in A is half the price, but a first payment costs more. That describes the two observed groups. Moving the entire budget to B still requires information about the available audience, subsequent usage and product costs.

The table includes only advertising so far. If one partner also received a commission or a paid team supported the other channel, show those costs alongside it. Build full CAC using a cost definition agreed in advance. Comparing an all-in amount for one channel with another's media budget would obscure the choice.

Find the obstacle between registration and payment

An unfinished journey can have several causes: the person did not understand the terms, could not upload a document, is awaiting verification or simply has no immediate need to pay this way. A single "inactive" status cannot tell the team what to fix.

The marketer walks through the journey with the product team using a test account. Is the next step visible after registration? Is entered information retained on return? Does an error explain what to do? Can someone get help without repeating the entire story?

The company's specialist teams make customer verification decisions. Discuss form usability separately from the checks themselves. FATF guidance considers a risk-proportionate approach alongside access to financial services. For this task, distinguish a technical difficulty from a verification outcome: fixing a document upload does not require promising everyone will pass.

A credit product would add a separate lending decision. Confirming an identity cannot be called loan approval. The marketer establishes what each status means with the process owner before writing about it in advertising.

A first purchase and continued use

Someone may qualify for a bonus and stop using the product. Track repeated activity over an equal window after the first payment.

Add this observation: everyone who made a first payment has completed the following 30 days from that payment. In A, 12 of 20 made another completed purchase; in B, 18 of 30 did. Both channels have a 60% repeat rate among those who started using the product. Against all registrations, the rates differ: 12 / 100 = 12% and 18 / 50 = 36%.

The repeat window begins at the first payment, not registration. The last person, making their first payment on day 30, needs another 30 days of observation. Mark their result as immature until that window closes. Keep these retention analysis rules consistent between channels.

Even matching 60% rates can describe different needs. One person needs daily payments, another an occasional purchase. Choose a useful form of repetition for the product, then compare it with support enquiries and income.

Include the bonus in acquisition cost

In B, each of the 30 users with a first payment was awarded and paid 3,000 ₸. Bonuses cost 90,000 ₸, making 290,000 ₸ together with advertising. Per user with a first payment, that is 290,000 / 30 ≈ 9,667 ₸.

The same amount divided by 18 repeat users gives about 16,111 ₸. Changing the denominator did not increase spending; it changed the outcome receiving that cost. Bonus terms should make clear in advance which action qualifies, when the reward becomes available and where it can be used. Kazakhstan's law On Advertising (as amended up to 17 May 2026) prohibits hiding material contract terms that an ad mentions (Art. 14, subpara. 5), and a licensed advertiser states its licence number and the issuing authority (Art. 6, para. 4).

The bonus has already been paid in this example. If it has only been promised, track future payouts alongside cash spent. Until the bonus is paid, the campaign only looks cheap: the obligation outlives the advertising report.

Purchase volume and product income

The value of purchases paid by card passes through the product. It is not equal to the product's income. An account top-up does not become the organisation's revenue either. Finance defines which income and expenses belong to the product.

For a separate calculation, suppose finance confirms 100,000 ₸ income and 40,000 ₸ variable transaction and servicing costs for group B over the chosen completed period. Those costs exclude advertising and the promotional bonus. Contribution before those two items is 60,000 ₸.

After 200,000 ₸ advertising and 90,000 ₸ bonuses, the period's result is 60,000 - 290,000 = minus 230,000 ₸ before other costs. Future transactions may change the result, but forecast them separately and later compare the forecast with actuals. The first-payment count does not fill that financial gap.

For a product with several services, allocate income and shared expenses under an agreed rule. The same receipts cannot pay back several campaigns in full at once.

Reporting data and customer data

The marketer needs the source, agreed event, time, processing status and aggregate group results. Document copies and reasons for individual decisions stay within the process responsible for them. Keep them out of advertising parameters and free-text event labels.

Before integration, data owners agree the fields and permitted recipients, and the developer works through test transactions, cancellations and duplicate notifications. Delayed or rejected delivery must appear as a data error in the report. An absent advertising event cannot establish that the customer did not use the product.

Exercise: justify the next budget

The owner proposes doubling channel B because a first payment costs about 6,667 ₸. Prepare a short calculation including the bonus and repeat purchases, then name the information needed for expansion.

The calculation contains 290,000 ₸ of advertising and bonuses, about 9,667 ₸ per user with a first payment, and about 16,111 ₸ per user with a repeat purchase. Under the period economics given, minus 230,000 ₸ remains. The next budget requires subsequent income, servicing costs, available channel volume and an acceptable duration for funding this group.

The table shows where B outperforms A in observed usage. Doubling requires an estimate for the next volume and the money the company is willing to invest before future receipts. Record separately when the team will revisit that estimate and which actual figures should be available by then.