During lender onboarding for BLU365's early-access program, we saw a clear pattern: collections managers wanted recovery, not to mistreat borrowers. High-urgency, threat-led, deadline-first language came from their systems, not their preference. They lacked a tested alternative.
Our pilot shows that empathetic framing works better for early-to-mid-cycle delinquency. The difference is measurable, not philosophical.
What Empathy Means in Collections
Empathy is not softness or forgiveness of legitimate debt. The borrower owes a real amount, and collections must state that clearly and lawfully, including under CDC Art. 42, which bars humiliating, threatening, or harassing collection practices.
Operationally, empathy means recognizing context, offering a realistic path forward, and framing outreach as problem-solving. Compare: "Your account is 10 days past due. Resolve immediately to avoid further penalties and potential credit bureau action." Or: "Your payment did not process this cycle. We have options for this week, including a short extension or split payment."
Both messages say payment is expected. The first makes the lender an adversary; the second makes it a partner in resolving the problem. For borrowers avoiding contact because of shame or uncertainty, the second drives engagement. Our contact-rate data shows this.
Contact Rates When Tone Fits Context
In BLU365's early-access program, 3 lender clients and approximately 2,400 accounts from Q4 2025 through Q1 2026, we compared WhatsApp messages on delinquency days three and seven. Resolution-framed messages, with options and mid-level urgency, produced an 18 percentage point higher contact rate than deadline-first messages in the same lender portfolio, controlling for channel and timing.
Among borrowers who engaged with the resolution sequence, 67% gave a specific reason for missed payment, such as salary timing, medical expense, or job change. That disclosure enables action: a seven-day extension for a delayed paycheck, split payment after a medical expense, or a hardship-program referral after job loss.
Deadline-first messages produced a disclosure rate below 20%. Borrowers either paid without contact or went silent for the rest of the recovery cycle. Urgency produced quick compliance in low-friction cases or avoidance in the rest, not better visibility into borrower circumstances.
Why Adversarial Framing Backfires
Financial stress and shame can make borrowers withdraw from a creditor instead of addressing the problem. This is not unique to Brazil and can intensify when formal credit collections are unfamiliar or the interaction feels automated and impersonal.
A Brazilian borrower who receives a legal-sounding WhatsApp message from an unfamiliar number may block it or ignore later messages. The lender loses contact even when the borrower can and wants to pay, because the style triggered avoidance first.
The lender loses both recovery and relationship value. Borrowers who face adversarial collection are less likely to take another product from that lender. With acquisition costing five to eight times retention, aggressive framing costs more than the overdue balance.
Where Empathy Falls Short
Our pilot found that resolution-first framing is not best for every account or delinquency stage. Some accounts respond better to direct, higher-urgency language, so programs need both.
Accounts with no opens, responses, or payment-link clicks across prior attempts showed marginally better response to consequence-led messages at day-14 or day-21. A borrower who received three resolution-framed messages without responding is unengaged for another reason: incapacity, avoidance, or a channel mismatch.
For these accounts, stating the next step, credit bureau reporting, referral to a collections partner, or legal recovery, can be appropriate and raise late-stage contact rates for some. It signals that the resolution window is closing.
Empathy-first is the right starting point, not a rule for every touchpoint. Start with resolution framing, then use engagement signals to guide urgency rather than leading with it from day three.
What This Looks Like in Practice
Empathy at scale needs a calibrated tone library and escalation logic that uses current-cycle engagement history to choose the next tone.
The library groups templates from resolution-first to mid-urgency to direct-consequence, with WhatsApp, SMS, and email versions. Tone reads differently by channel: formal in email, aggressive and personal on WhatsApp. Select channel and tone together.
The routing rule starts new accounts with resolution-first tone, moves engaged but unresolved accounts to mid-urgency, and sends accounts with zero engagement after two touchpoints to direct-consequence framing at the right stage. Basic rules need no AI and can run manually. At portfolio scale, automation routes messages without account-by-account judgment.
Our pilot outcome data suggests better recovery on accounts worth recovering: borrowers able to pay who need a clear, non-threatening path. The customers page gives lenders the early-access program's scope and outcomes.
The collections managers we worked with already believed borrowers respond better when treated as people, not liabilities. The data helps them make that commercial case to leadership and compliance. Ethical and commercial arguments point the same way.