The Credit Offer That Arrives at the Wrong Moment

The same offer can read as help on one day and as pressure on another. In financial products, timing is part of the product.

Published 2026-10-22 ยท 5 min read

The Credit Offer That Arrives at the Wrong Moment

Eligible Is Not The Same As Ready

A fintech decides who can receive a credit limit increase. The decision is careful: income, history, repayment behaviour, risk models. The result is a list of eligible customers.

Then the offer goes out to that list on the day the campaign is scheduled.

Eligibility answers one question: can this person have this product? It does not answer a second one: is this a good moment to bring it up? For that question the app usually has information it never uses.

Two Customers, Same Offer

Two customers are equally eligible for the same limit increase.

The first has opened the app several times this week, visited the limits screen twice, and looked at instalment options for a planned purchase. An offer here answers something she was already looking into.

The second has been checking his balance many times a day, more than usual, and opened the screen for a bill he has not yet paid. The same offer, on the same day, lands in a different situation. It can read as an invitation to take on debt at the moment he is most worried about money.

The risk model treats them the same. The week they are having is not the same.

What The App Already Sees

None of this requires new data collection. It is the pattern of use the app already records:

  • Rhythm of balance checks. A sudden rise compared with the person's own baseline is one of the clearer signs of financial worry.
  • Where they go. Repeated visits to limits, instalments, or simulators suggest active interest. Repeated visits to overdue bills suggest pressure.
  • What they start and leave. A simulation opened and abandoned three times is a decision in progress.
  • Change from their normal. The comparison that matters is with the person's own usual behaviour, not with an average user.

    • No single signal is proof of anything. Together, read with a confidence level, they describe whether this is a moment of interest, a moment of strain, or neither.

    Three Possible Decisions

    With that reading, the offer stops being a single action and becomes a choice:

  • Send now. The person is actively exploring the subject. The offer is relevant and timely.
  • Wait. Signs point to strain. Holding the offer for a calmer period costs little, and sending it now may cost trust.
  • Offer something else. For someone under pressure, information about renegotiation or payment dates can be more useful than more credit.

    • The second and third decisions are the ones a calendar-driven campaign cannot make.

    The Line That Must Not Be Crossed

    There is an obvious risk in this subject, and it should be stated plainly. The same signals that show financial worry could be used to target people when they are most vulnerable. That would be exploitation.

    The principle that keeps this honest is direction: behavioural context should be used to hold back or soften an offer in a moment of strain, and to make help easier to reach. It should not be used to press harder because someone looks desperate.

    It also matters what is read. Fluence works with behavioural patterns only, never with the content of messages or transactions, and is designed to comply with LGPD and GDPR.

    Measuring It Properly

    Acceptance rate is the usual measure of an offer, and it is not enough. A timing decision should also be judged by what follows: repayment behaviour, complaints, whether the customer keeps using the product, and how many offers were held back.

    An institution that sends fewer offers and sees better outcomes on those has improved something real. In financial services, an offer made at a bad moment is remembered longer than one that was never made.

    See how we approach measurement and limits.