Before BLU365, I repeatedly saw Brazilian lenders use two or three email templates on days 3, 7, and 14 of delinquency. Managers said, "It works," and some borrowers did pay. They rarely measured who could have been reached but was not, whether from ignoring lender email or from receiving the day 7 message at 10 AM.
These three mistakes recur. They reinforce one another and can be fixed without replacing your collections infrastructure.
Mistake One: One Message for Every Borrower
Overdue portfolios vary widely. Someone five days late on a R$600 installment because payday and the due date missed each other is not in the same position as someone avoiding a debt since origination. Sending both "resolve your overdue balance immediately" ignores that difference.
This affects results as well as fairness. A borrower facing shame or hardship may respond better to "We have options that may work for your current situation." Someone who simply missed a payment may need urgency at the first touch. One template will reduce resolution for at least one group.
You do not need twenty templates. Start with two or three tone profiles based on account signals and prior engagement. Manual teams can segment by balance and days since payment. A platform can add behavior from current-cycle contact attempts. Both approaches beat one universal template.
Mistake Two: One Timing for Everyone
When the second message fires on day 7 at 10 AM, a borrower starting a shift at 9 AM may miss it. A night worker may be asleep. Someone in another time zone may receive it during the evening.
Timing affects contact rate independently of copy. A good message sent at 10 AM on Tuesday can perform very differently from one sent at 7:30 PM to a borrower active on messaging apps in the evening. In BLU365's early-access pilot, matching timing to observed engagement improved contact rate by 14 percentage points when at least two prior contact events showed a pattern.
For new accounts, there may be no preference data. After two or three current-cycle attempts, there is. Opening a day-3 WhatsApp message at 8:15 PM signals an evening window. Use it to time day-7 outreach when your platform records delivery and open events.
If timing is undifferentiated, split the next overdue cohort: half at the default time, half from 4 PM to 8 PM. The contact-rate gap should appear within two weeks.
Mistake Three: Treating Non-Payment as the Only Signal
A borrower who opens the day-7 message, visits the payment link, and stops differs from one who never opens it. The first engaged, but friction blocked resolution. The cause might be a lost login, a wish to discuss payment, or unavailable funds at that moment.
Those outcomes need different follow-up. Someone who clicked but did not pay should not receive only the standard day-14 reminder. Try: "We noticed there may have been an issue with the payment link earlier this week. You can also pay directly via PIX code: [code]. If timing is a challenge, reply and we can look at options." This recognizes engagement and addresses likely friction.
Many programs miss partial engagement because payment-link clicks and dunning records sit in separate systems. Connecting them is a data-plumbing task, usually a one-time integration with continuing recovery benefits.
What Connects These Mistakes
All three reflect programs built around easy configuration, not the borrower's situation: one message, one timing, and non-payment as the only signal. These defaults reflect tool limits, not proof that batch-and-blast is the best recovery method.
You can fix them independently. Add two tone variants, a time-of-day rule for priority segments, and a branch for incomplete payment attempts linked to payment-link clicks. None requires replacing the collections platform.
BLU365 handles all three at scale without manual setup. The same principles work in rules-based programs. Algorithmic channel and timing decisions become more efficient above 1,000 accounts per month, but the logic works at any size.
Rough Numbers
With a 22% contact rate, an 8% first-resolution rate, and 1,000 overdue accounts per month, an email-first program reaches about 220 borrowers. About 80 resolve on first contact. The rest need follow-up, escalation, or write-off.
Timing optimization can move contact rate toward 31%. Tone segmentation and partial-engagement tracking can move first resolution toward 12-14%. At 1,000 accounts, moving from 22% to 31% creates 90 more contact opportunities. Moving first resolution from 8% to 13% adds about 45 resolved accounts monthly. At an average R$600 balance, that is roughly R$27,000 more recovered each month from 3 operational changes, without a new platform or major engineering project.
You do not need an automated collections platform for these improvements. Below 400 accounts monthly, good manual workflows using the three fixes can beat many current programs. As volume grows, measure when the attention required by those workflows costs more than automation.